Bridge, MFT, achromats, dSLR, primes - a journey of exploration

Have you tried anything to get more insects? Or do you not have the space? Apple trees are very good - whilst in bloom. Bramble is great as well, full of insects.
My garden is quite small and when neither the apple trees nor ivy are not flowering, I don't get very many insects.
I have planted some hanging baskets with bee friendly plants (in May) but despite them flowering nothing has come to them yet.
Obviously letting grass grow and leaving some small pieces of wood around are great ways to find things. I probably would have a lot less things to photograph without that.
All of these advice can cause marital problems - and you do so at your own risk.

Thanks for the helpful suggestions. Unfortunately it is out of my hands. I am the gardener's assistant; the gardener decides. I mention it from time to time, but to no avail. It's a bit frustrating as we have plenty of space. Perhaps it's time to try again.
 
Nick can you elaborate on the 'specific' PP adjustments on one or two of the images posted or have I missed this information?
 
Thanks for the helpful suggestions. Unfortunately it is out of my hands. I am the gardener's assistant; the gardener decides. I mention it from time to time, but to no avail. It's a bit frustrating as we have plenty of space. Perhaps it's time to try again.
That is a shame, it can make a real difference. When I was a child I never ever saw grasshoppers in my parent's garden, my front garden is about 1% of their back garden and I often see them there (not this year sadly, but I do now have 2 species of bees there), but I never mow.
 
Nick can you elaborate on the 'specific' PP adjustments on one or two of the images posted or have I missed this information?

No, you haven't missed it.

Let's put it in context.

I first do a very fast longlisting cull, using FastPictureViewer, taking 1-2 seconds per shot and marking up all that don't have some immediately obvious problem that makes them unusable. If in doubt, I include. I may end up with a lot of "repeats" of a shot, but that is fine. I make no attempt at this stage to decide which is the best of a series. I just take each shot on its own merits: include or drop. That way I can get through this stage for say 600 shots from a session in 10 to 20 minutes.

At any stage in what follows I may drop images (e.g. all but one of a set of "repeats", or because on closer examination I don't want to use a shot). Or I may backtrack when an image looks basically usable, but I want to redo something in an earlier stage.

The longlisted raw files are processed with a preset in PhotoLab. This does geometry correction, applies a white balance calibrated for the flash/diffuser setup I am using and applies a DCP camera profile (I use Adobe DNG Profile Editor to create these DNG profiles from reference shots of a ColorChecker Passport captured with the camera in question). It also balances the lightness somewhat, enhances microcontrast somewhat, applies lens-specific sharpening and applies noise reduction. I used to use presets which were specific to a particular camera and a particular one stop ISO range (e.g. ISO 100 to 160) with the noise reduction increased for each higher ISO band. Recently I haven't been using ISO-specific presets, instead just applying the base ISO preset for a particular camera. This is because I now have confidence in the Topaz noise reduction in the final stages being able to handle any noise, and this lets me reduce any loss of detail from applying heavier noise reduction earlier in the process.

PhotoLab produces DNG files which are then processed with a preset in Silkypix. It is the same preset for all the DNG files, irrespective of camera, lens, subject matter etc. It does just one thing, which is to pull down the highlights, a lot. I think Silkypix is particularly good at this, for example conserving colours in highlights as it brings the highlights down, and also giving me flatter files which facilitate highlight adjustments.

Silkypix produces TIFF files, which I then give image-specific adjustments in Lightroom. I'll come back to this in a moment.

When I've finished in Lightroom I output 1300 pixel high JPEGs, to which I apply noise reduction/sharpening using either Topaz DeNoise AI or (and sometimes and) Topaz Sharpen AI. DeNoise AI has two methods and Sharpen AI has three methods. I typically use just one of these five methods, but occasionally two. I often use these with their default parameter values. I often do all the images that remain at this stage in a single batch operation. Some people worry about altering and resaving a JPEG file but when saving with a high JPEG quality I have never been able to see much difference. Normally none actually, when I have taken the time to do a detailed comparison.

Now back to the image-specific adjustments.

When a TIFF files come into Lightroom it looks flat, as for example on the left below, but because of what has been done to it in PhotoLab and Silkypix the file is now in a good condition to work on. What this one looked like at the end of the Lightroom processing is on the right.


01 Lightroom start and finish
by gardenersassistant, on Flickr

I start in Lightroom with something that isn't image-specific: I apply mild adjustments to Texture (to enhance the visibility of detail a little), Clarity and Dehaze. I use the same three values for all the images so this can be done in one operation.

Then for an individual image I typically adjust the overall lightness to get it looking roughly right. I then push up the Whites until the top end clipping indicators appear. These are channel-specific and I generally stop as soon as any of them clip, because going further will unbalance the colours. I then pull down the Blacks until the bottom end clipping indicators appear. This is essentially setting the white and black points so I'm using all the available output dynamic range without damaging anything. I then adjust the Highlights and Shadows to taste. If I have a series of very similar images I'll make these adjustments to one of them and (in one move) copy the adjustments to all of the others.

I crop and possibly rotate the image. Sometimes I'll copy the crop to give me a good start with some similar images.

I may then do some local adjustments, which might be cloning or making local adjustments to lightness, highlights etc, using a brush, radial filters (oval areas) or graduated filters (either side of a straight line). Very occasionally I may use HSL colour adjustments (this is more likely with my small sensor cameras). Here is a silly example of the sort of things you can do, although obviously you need to be rather more subtle about it!


02 HSL adjustment, silly example
by gardenersassistant, on Flickr


Sometimes there is some cloning that is too difficult (or too tedious) to do in Lightroom, and so I may loop out to Photoshop to do that and then back to Lightroom. One other reason I very occasionally go out to Photoshop is when the composition doesn't quite work and I want to stretch or contract one side of the image to get the subject positioned where I want it in the frame. This is generally with flowers. I don't know that I've ever done this with invertebrates, although come to think of it that might be useful sometimes.

And that's about it for image-specific adjustments. [EDIT>] So most of the processing is done by batch processes, freeing up my time to do other things. My processing may well sound horribly complicated, but in practice it lets me deal with large numbers of images with relatively little effort on my part.
 
Last edited:
As described in this earlier post, I discovered that by using a macro lens and teleconverter on interchangeable lens cameras I could get deeper depth of field than with my usual setups of close-up lenses on telezoom lenses, and that despite the heavy diffraction blurring, with careful post processing I could produce images that to my eye looked acceptably detailed and sharp for my usual 1300 pixel high images. Since I greatly value larger depth of field for my invertebrate images I decided to explore this approach further.

I did some more tests in the garden, using full frame, APS-C and micro four thirds, and it quickly became apparent to me that the full frame Sony A7ii setup was the one I felt most comfortable using out in the field. However, to get the range of magnifications I wanted this would mean using two setups, either changing lenses out in the field (between the Sigma 105 macro with 2X teleconverter and the Laowa 25 macro), carrying around two cameras each set up with (different) flash arrangements, or carrying just one setup and restricting myself to the range of scene sizes which it covered.

None of these alternatives was appealing. The very small aperture technique is highly vulnerable to dust etc on the sensor. Changing lenses out in the field increases the risk of this substantially, especially if changing lenses a lot. Carrying around two cameras both rigged for flash is something for which I haven't yet found a practical solution. Restricting the range of scene sizes I could tackle in a session would mean missing good opportunities.

A timely discussion I was following at dpreview came up with a possibility that might avoid these problems by using a single camera with a single lens to cover all the scene sizes I wanted to cover. This would be to use a 2X teleconverter with one of the Laowa 2X macro lenses. There are two models, 65mm and 100mm. These are fully manual lenses, so no autofocus and no communication between the lens and the camera, so no Exif data from the lens, and no way to adjust the aperture using a camera dial. Also, focusing has to be done with the aperture stopped down to the capture aperture, which means a dark viewfinder if using a dSLR with the sort of small apertures I use. If using mirrorless the EVF/rear screen will gain up but is likely to be very noisy and unresponsive. There is one exception to this however; the EF-mount version of the 100mm 2X is chipped, providing communication with the camera and focusing with the aperture fully open, just closing down when the image is captured.

I would have to use an EF to FE adapter. There was no guarantee that this would work. Adaptation can be a bit hit and miss. My EF to FE adapter is a Sigma MC-11, and Sigma guarantees it will work only with some of its own lenses. A further potential problem was that I would be putting a (third party) teleconverter between the lens and the camera. However, my Sigma 105 (not one of Sigma's guaranteed to work lenses) works with this 2X configuration. And even if the Laowa 100 turned out not to work with the A7ii, I would be able to use it on my (EF-mount) Canon 70D and possibly on micro four thirds using my Commlite mFT to FE adapter.

So I bought a Laowa 100mm 2X macro lens.

It works fine on the A7ii with Kenko TelePlus Pro 2X teleconverter.

I did some basic measurements on working distances, shown below in green. From 1:1 to 4:1 I could use an effective aperture of at least f/90, to give twice the depth of field I can get with my close-up lens setups with their f/45 full frame equivalent minimum effective aperture, and progressively more than twice the DOF as the magnification increased (with progressively worse diffraction softening to the point of unusability, the limit of which I have yet to establish). The working distances looked acceptable too, from 175mm at 1:1 down to 75mm at 4:1 (compared for example to around 50mm at 4:1 with the MPE-65 and around 40mm with the Laowa 25).


1653 Illustration 10 - Working Distances
by gardenersassistant, on Flickr

At 4:1 the minimum scene size would be 9mm x 6mm. I would not be able to get to the 5X magnification available with the MPE-65 or Laowa 25 (7mm x 5mm). However, in my testing of the Laowa 25 for small flies on the pond I had chosen to use 2.5X to 3.5X even though I had magnification up to 5X available, so I think 4X will be sufficient for the vast majority of my purposes. For a 1300 pixel high image, at 4:1 a 100% crop of the 24 megapixel A7ii sensor would cover a scene of around 3mm x 2mm, so for example a 1mm mite would only fill half the frame height, but I very, very rarely photograph such small subjects and even filling half the frame with a tolerably detailed image would be an achievement for me.

This setup would give me from infinity focus down to scenes 9mm x 6mm, with a single camera and no lens changing. I would be able to handle invertebrates using flash, but also natural light shots of larger sunny day invertebrates such as butterflies and dragonflies (although it is years since I have seen a dragonfly), and even the odd pheasant or fox, and flowers. It would lack the flexibility and functionality that I make use of with my normal flower setup, but should be good enough for occasional use, for example out at a nature reserve (assuming I ever get out to a nature reserve again). This all means that with spare batteries for the camera and flash and a ColorChecker Passport all fitting into pockets, I would not have to take a bag with me.

This was looking good on paper. But would it work in practice?

Short answer, it looks like it will. In a test session out in the garden, with some invertebrates and also a few flowers, it proved practical to use and let me produce 1300 pixel high images of a quality which I thought fit for my purposes. I have posted some examples in this thread in the forum.

Here is the setup I used for the test session. As well as the Laowa 100, I was testing a new flash setup (see next post), with a new type of diffuser on the Yongnuo YN24EX flash heads.


1653 Illustration 01 A7ii+2X TC+Laowa 100 2X, Yongnuo YN24EX+DivDiff
by gardenersassistant, on Flickr

The LCD on the A7ii is hinged rather than fully articulated. This is ok for invertebrates as I never use portrait mode for invertebrates. I use portrait mode a lot for flowers, but this setup will only be used for flowers occasionally. As with my articulated screens I am using an LCD hood to help viewing the screen on bright days. We can also see the flash unit controls. They are simple enough to use, but for my purposes I won't really be using them (see next post).


1653 Illustration 02 YN24EX controls, LCD hood
by gardenersassistant, on Flickr

The lens does not extend as the magnification increases, but it does retract further into the lens barrel as the magnification decreases. Here is the lens as positioned for 2:1.

1653 Illustration 04 Laowa 100, 2to1, lens forward by gardenersassistant, on Flickr

Here it is retracted for infinity focus.


1653 Illustration 03 Laowa 100, infinity focus, lens retracted
by gardenersassistant, on Flickr


The rig weighs 2.4 kg. I would prefer it lighter, but it is balanced ok for carrying around with the left hand under the TC and back end of the Laowa. In use, it works fine with the left hand under the lens and the thumb and index finger on opposite sides of the focus/magnification ring. The focus/magnification ring is firm but smooth and can be moved by extremely small amounts. The ring rotates about 120 degrees from infinity to 2:1, and around 90 degrees from 1:2 to 2:1. The small amount of travel is good for locating subjects: carry the camera with the focus at say 1:2, use that to locate the subject and then move the camera towards the subject while rotating the ring until the desired magnification/framing is reached. This is much easier and faster than with the MPE-65, which requires a lot of turning to move between magnifications, and won't go out beyond 1:1 anyway, or the Laowa 25 with its fixed and short 40mm working distance, a stiff ring that feels like hard work to move and an even more restricted range of magnifications. I am in the early stages of developing the muscle memory for this, but I get the impression that once I have done that this is going to be at least as fast as the zoom in/out method I use with close-up lenses on telezooms.

I have often written that one of the reasons I like close-up lenses on telezooms is that it gives me autofocus at all the magnifications I use, and that lets me respond quickly and place the centre of focus exactly where I want it, and do this for moving subjects. It may seem surprising therefore that I am finding I'm ok using manual focus with this setup. I think this is partly to do with the A7ii giving me a fairly good view of the scene on the LCD and partly because I'm finding that focus peaking is working for small subjects; only just, but that is enough. Even with the small flies I was getting a slight focus peaking signal, and it turns out that was all I needed. Pressing the shutter button as soon as the focus peaking signal appeared produced a higher hit rate than I am used to with small subjects. And with larger subjects, independent of any focus peaking, I could see which part of the subject was in focus, and so by moving in and out slightly I could judge when to press the shutter button as I passed through the plane where I wanted focus to be centred. I suspect this is going to work as well as close-up autofocus for moving subjects, and may turn out to be better for small subjects that are moving around. Manual focusing does not seem to be as stressful on my eyes as I feared it might be.

On the "Manual focus is working surprisingly well" front, it is I think instructive to compare it with what happens with autofocus, especially for small subjects. I do get autofocus with close-up+telezoom setups, and very useful it is too. However, with close-up lenses autofocus only works within a given range of working distance. For more powerful close-up lenses this range of usable working distance gets to be quite small. In order to use autofocus you have first to get the working distance within range, and the way I do that is using autofocus on a trial and error basis. If it doesn't work I move towards or away from the subject and try again. If you are not within the usable range straight away then you lose time. The more failed attempts you have, the longer it is until you can capture an image, and the more likely it is that the subject will have left the scene. The focus peaking approach is more direct and may well be quicker, especially for small subjects where the working distance "window" for close-up lenses is very small and may be difficult and time-consuming to find.

So, capturing looks doable. Now on to post processing.

My current basic processing products and workflow work ok, but I am exploring some adjustments to the workflow to make it better suited to the raw files this approach produces. There are two major issues.

First, the very small apertures produce images that are extremely soft "out of the camera". In fact they are so soft that it can be difficult to tell if an image is going to be usable or not. I quickly realised that if I used my usual visual standards for producing an initial longlist then I was throwing away a lot of perfectly usable images. Topaz DeNoise AI provides batch processing and I am exploring its use early in the pipeline to produce throw-away sharpened versions to reveal the potential of the images for longlisting purposes.

The other side of that equation is that I'm finding it quite extraordinary what DeNoise AI, particularly its legacy AI Clear method, can bring into visibility. As far as I can tell this is not by and large invented pseudo detail, although the AI methods can do that. I believe it is detail that is in the image but with such low contrast that I for one can't see it.

The other issue is dust spots and other spots. With very small apertures any dust on the sensor shows up clearly and sometimes, less clearly, dust on the lens/teleconverter. Having cleaned the A7ii sensor I was still getting spots, but these were smaller and darker than how dust spots usually show up, pretty much black. I have seen small black spots before - dust under the IR or whatever other filter is on top of the sensor. These are bad news; a sensor replacement is the only cure. It turns out though that these new dark spots were being caused by the rather strong adjustments I use sharpening tiny bits of detritus on foliage in the scene. I have instigated a spot-checking stage for the images included in the final shortlist. For these I look systematically across a whole image at 100% and clone out any problem spots.

In the next post I will discuss the flash arrangements for this setup.
 
Last edited:
Continued from previous post.


In the test sessions I did before getting the Laowa 100mm 2X macro lens, one of the subjects was an ant with a very reflective body. I got horrible, completely blown flash-reflection highlights. And I still got those blown highlights even if I radically under-exposed the images. It is a problem I have had for a long time with highly reflective animal bodies. While I was awaiting delivery of the Laowa 100 I built a new pair of diffusers for the Yongnuo YN24EX in yet another attempt to get smoother light.

The new diffusers are made from 6 inch tin foil cooking trays, two layers of "plastic paper" and lots of hook and loop pads. The pads are self-adhesive but don't adhere well to tin foil, especially crinkly/crumpled tin foil. I have previously lined the outside of my diffuser boxes with duct tape, to which the hook and loop pads will adhere well, and the tape also helps to stop the tin foil tearing at vulnerable points, especially the flaps used to hold the diffuser boxes on to the flash heads. This time I used a drop or two of superglue on any of the pads which came off. Not using duct tape made the diffuser boxes much easier and quicker to construct. I'll have to wait and see if tearing is a problem. Hopefully it won't be because this design doesn't have the delicate little flaps.

Here are the new diffusers seen from the front. The flash heads can be rotated sideways and up/down and this can be used to close the gap between the diffusers.

Two things I was attempting to achieve were to get the light source as close as practical to the subject to conserve power and speed up recycling for the next shot, and to make the area from which the light is emitted as large as practical from the point of view of the subject.


1653 Illustration 05 DivDiff front view
by gardenersassistant, on Flickr

The side view illustrates another thing I have learnt to do for better diffusion, which is to use multiple thin layers of diffusion material with as large a space as practical between the layers. There are only two layers in this case.


1653 Illustration 06 DivDiff side view
by gardenersassistant, on Flickr

Each flash head has two loop pads on it top and one on each side. Each tin foil box has four loop pads in matching positions. Each pair of loop pads is joined with a folded hook pad which catches on to both loop pads.

The outer diffusion layers are held in place by a loop band which fixes to hook pads on the outside of the outer diffusion layers.


1653 Illustration 07 DivDiff attachment straps
by gardenersassistant, on Flickr

Next we see the inner diffusion layers. They are attached to the inside of the tin foil trays with matching hook and loop pads. They are lifted a centimetre or so from the base of the tray.


1653 Illustration 08 DivDiff inner diffusion layer
by gardenersassistant, on Flickr


Finally, here we see one of the inner diffusion layers removed. The hole through which the flash head shines its light has been cut on three sides, leaving a flap angled down towards the inner diffusion layer. This flap blocks the light from shining directly on to the scene, diverting it to reflect off of the inside of the tray, which has been pushed into a somewhat curved shape the better to spread the light around. This arrangement is less efficient that letting the light shine directly on to the scene, but the light loss is acceptable at around 2.5 stops compared to not using any diffusion and having the light shine directly at the subject. (The best I have achieved previously is a little under two stops, so this doesn't seem to me to be an excessive price to pay for possibly better light.)


1653 Illustration 09 DivDiff diverter flap
by gardenersassistant, on Flickr

It remains to be seen how good or otherwise the quality of light is from this setup, but one practical issue that I have with using the Yongnuo is that I can't use it in ETTL mode with the A7ii. It is true that I have been using the manual Venus Optics twin flash for several years and so I am familiar with using manual flash. However, the Yongnuo is a Canon-compatible flash and during the earlier testing I used the Yongnuo in ETTL mode on the 70D. I was struck by how much quicker this let me react rather than having to fiddle around with buttons, initially guessing what power would be suitable and then adjusting it (if the subject was still around at that point). Unfortunately though I can only find one similar flash unit that works with Sony cameras, and that appears to be a rather low-powered unit. That would be no good for the amount of power I need when using very small apertures.

The way I am using the power at the moment is to leave the flash power on 1/4, which provides fast enough recycling times for my purposes. I then adjust the brightness of the image by adjusting ISO. I can do this quickly and easily with a conveniently placed button and wheel.
 
Here are the new diffusers seen from the front. The flash heads can be rotated sideways and up/down and this can be used to close the gap between the diffusers.

Two things I was attempting to achieve were to get the light source as close as practical to the subject to conserve power and speed up recycling for the next shot, and to make the area from which the light is emitted as large as practical from the point of view of the subject.


1653 Illustration 05 DivDiff front view
by gardenersassistant, on Flickr

Well, that didn't work. Diverting the light simply moved the hot area from directly in front of the flash head to the outer edge of the tin foil tray. Also, light from the flash would also sometimes bounce back into the lens barrel causing big contrast reduction ("milky/foggy").

I've gone back to an earlier design.


1657 1
by gardenersassistant, on Flickr


1657 2
by gardenersassistant, on Flickr

I've added a thin layer of expanded polystyrene directly on the front of the flash window.


1657 3
by gardenersassistant, on Flickr

Given that diffusion gets better with separation, this may not work too well, and it eats up a fair amount of light. Still, the total light loss from all three layers is around 2 1/3 stops, and I can live with that. I did a quick test run with long-legged flies, and they are quite shiny and difficult. I thought they came out (see here) about as good as the best I've achieved previously, possibly slightly better; I often have to "massage" them locally, typically on their backs, but I didn't have to do that for these. There are colour changes in the areas where the flash falls brightest, but I've got a feeling those are inevitable (I'd like to be proved wrong about that of course, and shown how to avoid it!)

Having worked out what was going on with the light bouncing back into the lens with the previous version, I realised that a milder version of the same thing was happening with this version, so I've also tightened up the edges nearest to the lens opening to try to cure that.

I've got the flash heads aligned so that the left hand one is nearer vertical and points down towards the closest focus region, with the other one around more towards the side, pointing further forward, roughly towards the 1:2 focus region. That combination may let me vary the magnification, as I like to do even for a single subject, without having to adjust the heads.


1657 4
by gardenersassistant, on Flickr

As the flash is only ETTL with Canon, I have to use it in manual mode, so I do have to change the flash power if I make a big change to magnification. I'm still wondering about the Meike twin flash that comes in a Sony version, despite it's apparently low guide number; you can attach more than two flash heads, so that might be a way forward if two were not powerful enough. Arranging diffusion for that might be a bit interesting.
 
Last edited:
In response to one of my videos at You Tube about the Raynox MSN-505 You Tube user Siegfried asked the following question " Hi, I have raynox 205 and my camera is the model Powershot SX 70 HS and i´d like to improve to get more detail photos, do you think i can buy Raynox 505 or can you recommend me another option? thank you.". My thoughts turned out to be more complicated than would fit into a You Tube response so I decided to write it up here as I think it expands on some earlier thoughts about my own Journey.

I have kept the image sizes within this site's posting guidance. There are larger versions of these illustrations in this album at Flickr and you may need to look at those versions to see some of what I'm talking about.

I'll start with the Raynox MSN-202, which is more powerful than the Raynox 250, but less powerful than the 505.

The first illustration shows how much closer the Raynox 202 lets you get compared to the 250 (in this case at full zoom with my Panasonic FZ330, which is similar to Siegfried's Canon SX70 bridge camera, except that the SX70 goes out to 1365 mm full frame equivalent compared to 600mm with the FZ330, Both images were captured with the same, minimum, aperture of f/8 (which is equivalent to f/45 on full frame), so both have the maximum possible depth of field for this kit and focal length/magnification. Minimum aperture also means a great loss of detail/image softening caused by diffraction.

At around 30mm, the working distance with the Raynox 202 is very small, which can make it difficult to get light on to the scene and easy to frighten off animal subjects. The scene width at 600mm equivalent is between 2.5 and 3mm, and would be around half that with the SX70 at full zoom.

It can be very difficult to find the subject with such small distances, especially if it is moving around, but even if it isn't.

The depth of field is very narrow and placing it accurately is difficult, whether using manual focus or autofocus. (Autofocus may or may not work. It does with my FZ200/FZ330s.)


1665 1 Raynox 250 vs 202 coverage comparison
by gardenersassistant, on Flickr

The depth of field is deeper with the less powerful Raynox 250, but which image has the better detail? To examine this closer I cropped the 250 image so it covered the same field of view as the 202 image. Here is the crop that was needed. FZ330 images are 4000 x 3000. This crop took it down to only 1100 pixels high.


1665 2 250 crop screenshot
by gardenersassistant, on Flickr

If we compare this crop with the uncropped 202 image we can see that the depth of field is greater in the cropped image.


1665 3 250 crop vs 202
by gardenersassistant, on Flickr

However, if we look closely we can see that the in-focus areas are much sharper/more detailed in the uncropped 202 image. On the left we are looking at the 250 image which was cropped down to 1105 pixels high to give the same field of view as with the 202, and then upsized from 1105 to 3000 pixels high to match the size of the 202 image on the right.


1665 4 250 cropped and upsized vs 202 at 100pc
by gardenersassistant, on Flickr

Pixel peeping is all very well to get a clear view of the differences, but what would these differences mean in practical terms? Images are often viewed at rather less than 100%. For example, I prepare my images for viewing at 1300 pixels high. Here is the difference between the 250 crop upsized to 1300 pixels high and the 202 image downsized to 1300 pixels high.


1665 5 Cropped 250 vs 202, 1300h, 100pc
by gardenersassistant, on Flickr

Which of these would I use (assuming it was of a more interesting subject!)?

With the upsized 250 crop we essentially have a greater than 100% crop. I think that is too much to ask of a camera with a very small, 1/2.3" sensor like the FZ330 and SX70, at least for my purposes. Actually, I don't know that I have ever used a 100% crop from my small sensor bridge cameras (including a Canon SX10 and before that a Canon SX3is), although I think I have come quite close to 100% occasionally.

So I would have no choice but to use the 202 image. The depth of field is extremely narrow and so I would probably frame my shots so as to allow for a fair amount of cropping (not down to 100% of course). That way I have found that I can get enough depth of field to make just about acceptable (<for my taste) images of small subjects like springtails, barkflies, mites and tics. (For example like the first 14 images in this album at Flickr.) I do think though that this is working at about the edge of what is feasible (for me at least) with a small sensor camera in terms of image quality and because of that post processing takes on even more importance than usual, and that in turn reinforces (to my way of thinking at least) the importance of shooting raw (because I can do more with raw files than out of the camera JPEG images).

Another thing to be aware of is that with the 202 you may get more chromatic aberration than with the 250, especially if you use a lot of zoom/magnification. This may cause more or less of a problem depending on how bad it is (the example below is something of a worst case test scenario, with deliberately high contrast edges - I generally haven't found chromatic aberration much of a problem with the 202). How problematic any chromatic aberration/fringing is may also depend on what processing software you use.


1665 6 202 CA
by gardenersassistant, on Flickr

One thing to bear in mind is that these examples were captured using minimum aperture of f/8 (f/45 full frame equivalent). This means that the images have lost a lot of detail because of diffraction. You could use a larger aperture, which would have less diffraction and therefore a bit better detail/sharpness, but you would have even less depth of field.

Now the 505. It is more powerful than the 202. You get more magnification, with a scene width of just under 2mm at full zoom on my FZ330, so it would be less than 1mm at full zoom on the SX70. The working distance is smaller at 18mm, which is even more difficult to work with. The depth of field is smaller. The chromatic aberration may be worse.


1665 7 202 vs 505
by gardenersassistant, on Flickr

Given all that, I doubt the 505 is a practical proposition for use out in the field, at least until you have become competent with the 202 at full stretch. I have seen terrific images which used a 505, but these were done using focus bracketing with a larger sensor camera, with the 505 on a sharp macro len using "sweet spot" (almost maximum) apertures so diffraction doesn't destroy the details. However, the in-focus "slice" that is captured with each shot is very thin indeed and, for my purposes at least, much too thin to use for single-capture (i.e. non-stacked) images.

For higher magnification with an SX70, rather than trying a 505 it may be better to try two 250s stacked together, or a 250 and a 150 stacked together - I have used both of those combinations. Or try a 202.

As it happens, and why I am so interested in this, is that I have tried various kit over the past few years but kept coming back to a bridge camera and close-up lenses. What has finally got me experimenting seriously with interchangeable lens cameras, especially for small subjects, is that I can get greater depth of field with them than I can with my bridge cameras. (For example see this post above).
 
It has been another month of experimentation, with hardware and software. Long story short, It looks increasingly likely to me that after all this time I will be moving to a full frame, macro lens based approach for invertebrates rather than using close-up lenses as I have been doing for over a decade. Flowers etc will continue with a macro lens on micro four thirds.

Now for some of the gritty details.

As discussed previously, depth of field is at the core of this. I can get more of it with interchangeable lens cameras. I have a lens setup (Laowa 100 plus 2X TC) that gives me sufficient depth of field for my (current) purposes, and I can use it on my full frame Sony A7ii, my APS-C Canon 70D and my Panasonic micro four thirds cameras. From an operational point of view, the A7ii works best, mainly because it gives me stronger focus peaking signals than those of my Panasonic mFT cameras that provide focus peaking (and like some of my mFT Panasonics, the 70D doesn't provide focus peaking at all). I can get greater depth of field with the 70D than the A7ii, and more again with the Panasonic mFT cameras. However, the A7ii provides enough depth of field for my purposes and, coupled with the relatively high focusing success rate, that makes the A7ii my current kit of choice.

I have tried the Meike Sony-version twin flash, but it didn't work out well. It does work in ETTL mode with the (Canon EF mount) Laowa 100 and 2X TC, but the ETTL turned out to be neither powerful enough nor consistent enough for my purposes, and fiddly (hence slow) to adjust out in the field, which is also bad news. I am happy using the Yongnuo 24EX in manual mode (after all I have been using the Venus Optics KX800 in manual mode for several years), and (like the KX800) it is easy to adjust, although for the most part I have been running it at a constant 1/4 power and using ISO to adjust lightness. It turns out that if I turn the aperture-assigned wheel by a given number of detents I can roughly balance the lightness by applying the same number of detents on the ISO-assigned wheel. (I can do this very quickly, and approximately, with a "feel" for the number of detents to match, and the A7ii gives me plenty enough latitude for a mismatch of a click or two).

Coupled with the short throw on the Laowa 100 (which I suspect makes it not very good for general purpose use), this speedy adjustability of aperture and flash power means I can be quite responsive to changes in circumstances such as new, different sized subjects turning up, and also when doing the variable magnification sequences that I like to do for a subject when it gives me the time to do it.

I tried a number of diffuser configurations and eventually burnt out on the attempt to solve the shiny body problem. I have learnt that the extent of flash reflections can vary considerably depending on the angle of attack, sometimes with rather small changes in the angle of attack, but you don't always get to choose the angle of attack and with some subjects it doesn't make a lot of difference. I have therefore put that on hold, at least for now, and have settled on using the Yongnuo flash heads front-mounted on the lens, with round fairly shallow diffusers, each with one layer of vellum paper and two of silk.

This A7ii plus 2X TC covers pretty much all the range of magnifications I want to use without needing to change anything other than the angle of the flash heads for the lowest magnifications I am likely to use with flash (and I can simply turn the flash off and carry on for larger, bright day subjects such as dragonflies and butterflies). So no changing around close-up lenses, extension tubes or anything else, with one exception. It is possible that I might want a bit more magnification with the A7ii setup. If so I have worked out that by using some glue dots to attach a couple of step rings to the front of the Yongnuo flash head frame I can attach a Raynox close-up lens using its fast to use clip on adapter. The Raynox 250 takes the magnification up from 4X to around 6X, and I think that is probably about as much as I would want to use hand-held out in the field, although I could add a 202 or a 505 if I was prepared to live with the extremely short working distances involved. The A7ii does have fairly good cropping potential as well, which is a more likely option. (I might be prepared to go right down to a 100% crop with the A7ii, which I'm not, quite, with my other cameras. This is a surmise though - I haven't tested that yet.)

My post processing experiments have resulted in my dropping DXO PhotoLab as the raw converter. Instead I am now using Silkypix Developer Pro 10 for the first stage of the processing, using its new, processor intensive noise reduction option. I am mainly using Silkypix in this role as it is giving me colours I prefer compared to PhotoLab, even though I can't use DCP camera profiles. Using Silkypix first means that I can't use PhotoLab's PRIME raw only noise reduction, but in comparing the two workflows it turns out that the results using Silkypix first rather than PhotoLab seem to be just as good in terms of noise (and also, importantly, and slightly surprisingly to me, sharpness/detail).

I am using PhotoLab on TIFF files produced by Silkypix (which won't produce DNG files as PhotoLab does). PhotoLab does some lightness balancing and microcontrast enhancement.

The Silkypix and PhotoLab stages use presets. This is simplified compared to the previous workflow as Silkypix automatically handles different ISOs very well, and so I don't have to use ISO-specific presets as with PhotoLab. And after the Silkypix stage, there is no reason to use ISO-specific presets in PhotoLab.

TIFF files from PhotoLab then go into Lightroom for image-specific adjustments, which seem more straightforward than previously, and sometimes are little more (and occasionally no more) than cropping. There may (very occasionally) be a round trip from Lightroom to Photoshop and back.

Lightroom produces 1300 pixel high JPEGs and these are processed twice in Topaz DnNoise AI, once using the DeNoise method with auto settings, and once using the AI Clear method with minimum settings. I then choose which of the versions to use.

I am now using Fast Raw Viewer rather than Fast Picture Viewer for initial longlisting. I used Fast Picture Viewer previously because it had a macro facility which let me automate the placement of selected files into ISO-specific folders for the initial PhotoLab processing. This is no longer needed as I'm not using ISO-specific presets and Fast Raw Viewer gives me a better view of the raw files, showing what they are capable of, rather than Fast Picture Viewer showing the embedded JPEGs, which especially with my exposing for highlights can often be rather dim and difficult to "read".

The images in this post used the latest processing workflow. They are taken from this album at Flickr, which is a rework of a set captured using the latest hardware configuration, except using two layers of vellum paper for diffusion rather than my currently preferred one layer of vellum paper and two of silk.

I previously used different workflows for flowers (because I didn't like PhotoLab's handling of colours and wider tonality for flowers). I have tried the new workflow for flowers and it seems to work ok (see this album at Flickr), so it looks like I may be able to simplify my processing, using the same workflow for everything.

While doing these experiments I tackled some of the same/similar subjects using on of my bridge cameras with close-up lenses, the configuration to which I have kept on returning over the past several years. This convinced me that I really couldn't get the same results with the bridge/close-up setup as with the A7ii/macro lens setup. And given that the biggest difference was the depth of field, and I can't get more depth of field with close-up lenses on mFT or APS-C (or only very slightly with APS-C) as I can with the bridge setups, there was no reason to think that mFT or APS-C with close-up lenses would be preferable to the macro lens+TC approach.

So, as it stands it looks like I will be using G9 with 60mm macro for flowers and other botanical subjects, and the A7ii with Laowa 100 and 2X TC for invertebrates, possibly with occasional use of a Raynox 250.
 
I have for some years believed that close-up lenses don't "lose light". So, adding a close-up lens would not change the exposure. It turns out I was wrong.

What is true is that close-up lenses do not "lose light" as magnification increases in the way that macro lenses, extension tubes etc do. For macro lenses, extension tubes, teleconverters, reversed lenses and bellows this effect can be approximated with the formula

Effective f-number = Nominal f-number * ( 1 + magnification )

Where "Nominal f-number" is the f-number you set on the camera/lens, and "Effective f-number" is the f-number you are actually using.

(Note: this does not apply to Nikon cameras which, with appropriate lenses, report the effective f-number.)

For example, with the lens f-number set to f/8, when shooting at 1:1 the f-number you are actually using is 8 * (1+1) = f/16. At 2:1 f/8 becomes f/(8*(1+2)) = f/24.

So, with macro lenses etc, if the f-number on the camera/lens is unchanged, as the magnification goes up the lightness of the image goes down. The setup "loses light".

This is not the case for close-up lenses mounted on zoom lenses. With a given close-up lens on a given camera and camera lens, as the magnification goes up the lightness of the image does not change.

I wrongly assumed from this that "close-up lenses do not lose light". A couple of days ago I did some measurements which showed that adding a close-up lens can, and for the kit I tested usually does, lose some light, typically around 1/3 stop, but less often 2/3, or none. This table summarises my findings.


1686 1 Close-up lens light loss measurements
by gardenersassistant, on Flickr
 
So, as it stands it looks like I will be using G9 with 60mm macro for flowers and other botanical subjects, and the A7ii with Laowa 100 and 2X TC for invertebrates, possibly with occasional use of a Raynox 250.

But Laowa have now announced a 50mm 2:1 macro lens for micro four thirds. Without needing the use of a teleconverter, this should be very similar to my A7ii + 2X teleconverter + Laowa 100mm 2:1 macro lens, in terms of the scene sizes it can cover and quite similar in terms of the smallest effective apertures it can use. It is very small and light, compared to the 100mm setup which is big and is heavy enough to give me (so far slightly) sore wrists. However, the working distances may be significantly smaller and one of the reviews I have seen said that some images looked a bit "milky", suggesting reflections inside the lens.

It is not in stock yet, but I am very tempted to try it when it does become available.
 
It has been another month of experimentation, with hardware and software. Long story short, It looks increasingly likely to me that after all this time I will be moving to a full frame, macro lens based approach for invertebrates rather than using close-up lenses as I have been doing for over a decade. Flowers etc will continue with a macro lens on micro four thirds.

Now for some of the gritty details.

As discussed previously, depth of field is at the core of this. I can get more of it with interchangeable lens cameras. I have a lens setup (Laowa 100 plus 2X TC) that gives me sufficient depth of field for my (current) purposes, and I can use it on my full frame Sony A7ii, my APS-C Canon 70D and my Panasonic micro four thirds cameras. From an operational point of view, the A7ii works best, mainly because it gives me stronger focus peaking signals than those of my Panasonic mFT cameras that provide focus peaking (and like some of my mFT Panasonics, the 70D doesn't provide focus peaking at all). I can get greater depth of field with the 70D than the A7ii, and more again with the Panasonic mFT cameras. However, the A7ii provides enough depth of field for my purposes and, coupled with the relatively high focusing success rate, that makes the A7ii my current kit of choice.

I have tried the Meike Sony-version twin flash, but it didn't work out well. It does work in ETTL mode with the (Canon EF mount) Laowa 100 and 2X TC, but the ETTL turned out to be neither powerful enough nor consistent enough for my purposes, and fiddly (hence slow) to adjust out in the field, which is also bad news. I am happy using the Yongnuo 24EX in manual mode (after all I have been using the Venus Optics KX800 in manual mode for several years), and (like the KX800) it is easy to adjust, although for the most part I have been running it at a constant 1/4 power and using ISO to adjust lightness. It turns out that if I turn the aperture-assigned wheel by a given number of detents I can roughly balance the lightness by applying the same number of detents on the ISO-assigned wheel. (I can do this very quickly, and approximately, with a "feel" for the number of detents to match, and the A7ii gives me plenty enough latitude for a mismatch of a click or two).

Coupled with the short throw on the Laowa 100 (which I suspect makes it not very good for general purpose use), this speedy adjustability of aperture and flash power means I can be quite responsive to changes in circumstances such as new, different sized subjects turning up, and also when doing the variable magnification sequences that I like to do for a subject when it gives me the time to do it.

I tried a number of diffuser configurations and eventually burnt out on the attempt to solve the shiny body problem. I have learnt that the extent of flash reflections can vary considerably depending on the angle of attack, sometimes with rather small changes in the angle of attack, but you don't always get to choose the angle of attack and with some subjects it doesn't make a lot of difference. I have therefore put that on hold, at least for now, and have settled on using the Yongnuo flash heads front-mounted on the lens, with round fairly shallow diffusers, each with one layer of vellum paper and two of silk.

This A7ii plus 2X TC covers pretty much all the range of magnifications I want to use without needing to change anything other than the angle of the flash heads for the lowest magnifications I am likely to use with flash (and I can simply turn the flash off and carry on for larger, bright day subjects such as dragonflies and butterflies). So no changing around close-up lenses, extension tubes or anything else, with one exception. It is possible that I might want a bit more magnification with the A7ii setup. If so I have worked out that by using some glue dots to attach a couple of step rings to the front of the Yongnuo flash head frame I can attach a Raynox close-up lens using its fast to use clip on adapter. The Raynox 250 takes the magnification up from 4X to around 6X, and I think that is probably about as much as I would want to use hand-held out in the field, although I could add a 202 or a 505 if I was prepared to live with the extremely short working distances involved. The A7ii does have fairly good cropping potential as well, which is a more likely option. (I might be prepared to go right down to a 100% crop with the A7ii, which I'm not, quite, with my other cameras. This is a surmise though - I haven't tested that yet.)

My post processing experiments have resulted in my dropping DXO PhotoLab as the raw converter. Instead I am now using Silkypix Developer Pro 10 for the first stage of the processing, using its new, processor intensive noise reduction option. I am mainly using Silkypix in this role as it is giving me colours I prefer compared to PhotoLab, even though I can't use DCP camera profiles. Using Silkypix first means that I can't use PhotoLab's PRIME raw only noise reduction, but in comparing the two workflows it turns out that the results using Silkypix first rather than PhotoLab seem to be just as good in terms of noise (and also, importantly, and slightly surprisingly to me, sharpness/detail).

I am using PhotoLab on TIFF files produced by Silkypix (which won't produce DNG files as PhotoLab does). PhotoLab does some lightness balancing and microcontrast enhancement.

The Silkypix and PhotoLab stages use presets. This is simplified compared to the previous workflow as Silkypix automatically handles different ISOs very well, and so I don't have to use ISO-specific presets as with PhotoLab. And after the Silkypix stage, there is no reason to use ISO-specific presets in PhotoLab.

TIFF files from PhotoLab then go into Lightroom for image-specific adjustments, which seem more straightforward than previously, and sometimes are little more (and occasionally no more) than cropping. There may (very occasionally) be a round trip from Lightroom to Photoshop and back.

Lightroom produces 1300 pixel high JPEGs and these are processed twice in Topaz DnNoise AI, once using the DeNoise method with auto settings, and once using the AI Clear method with minimum settings. I then choose which of the versions to use.

I am now using Fast Raw Viewer rather than Fast Picture Viewer for initial longlisting. I used Fast Picture Viewer previously because it had a macro facility which let me automate the placement of selected files into ISO-specific folders for the initial PhotoLab processing. This is no longer needed as I'm not using ISO-specific presets and Fast Raw Viewer gives me a better view of the raw files, showing what they are capable of, rather than Fast Picture Viewer showing the embedded JPEGs, which especially with my exposing for highlights can often be rather dim and difficult to "read".

The images in this post used the latest processing workflow. They are taken from this album at Flickr, which is a rework of a set captured using the latest hardware configuration, except using two layers of vellum paper for diffusion rather than my currently preferred one layer of vellum paper and two of silk.

I previously used different workflows for flowers (because I didn't like PhotoLab's handling of colours and wider tonality for flowers). I have tried the new workflow for flowers and it seems to work ok (see this album at Flickr), so it looks like I may be able to simplify my processing, using the same workflow for everything.

While doing these experiments I tackled some of the same/similar subjects using on of my bridge cameras with close-up lenses, the configuration to which I have kept on returning over the past several years. This convinced me that I really couldn't get the same results with the bridge/close-up setup as with the A7ii/macro lens setup. And given that the biggest difference was the depth of field, and I can't get more depth of field with close-up lenses on mFT or APS-C (or only very slightly with APS-C) as I can with the bridge setups, there was no reason to think that mFT or APS-C with close-up lenses would be preferable to the macro lens+TC approach.

So, as it stands it looks like I will be using G9 with 60mm macro for flowers and other botanical subjects, and the A7ii with Laowa 100 and 2X TC for invertebrates, possibly with occasional use of a Raynox 250.
Interesting, this season I’ve moved over to full frame from Canon crop 6D2 from 7D2
The level of detail is amazing and on occasion the dynamic range improvement helps too
Sometimes the insects are perched on white flowers and I just accepted a bit of blown out highlights
With the full frame 6D2 I get the shot without any blown highlights
I’m not sure if it’s due to me having more time now I’m not working so I can spend more time getting out in the right time and place or the new camera having better quality but the lighting and details I’m managing to get now is better than I’ve got before
 
But Laowa have now announced a 50mm 2:1 macro lens for micro four thirds. Without needing the use of a teleconverter, this should be very similar to my A7ii + 2X teleconverter + Laowa 100mm 2:1 macro lens, in terms of the scene sizes it can cover and quite similar in terms of the smallest effective apertures it can use. It is very small and light, compared to the 100mm setup which is big and is heavy enough to give me (so far slightly) sore wrists. However, the working distances may be significantly smaller and one of the reviews I have seen said that some images looked a bit "milky", suggesting reflections inside the lens.

It is not in stock yet, but I am very tempted to try it when it does become available.

It turns out that the working distance with the micro four thirds Laowa 50mm 2X macro lens at its maximum magnification is around 45mm. As the following table shows, this gives a similar scene size to the A7ii with the Laowa 100m 2X macro with 2X teleconverter, but its 45mm working distance compares to around 72mm for the A7ii setup. This would make it more difficult to use a front-mounted flash as I do with the A7ii setup, but given that the Laowa 50mm does not extend a lot like the MPE-65, a hot shoe mounted Venus Optics KX800 should presumably work fine as it does with my close-up lens setups.


1688 1 Scene widths and F90+ eff eq apertures FZ330+R,70D+65,G9+50,A7ii+TC+100 v0.3
by gardenersassistant, on Flickr

There are more issues though if I decide I need more magnification.

I noted previously that the magnification I get with the Laowa 100 2X with 2X teleconverter may not be quite enough for my purposes. With the A7ii setup I can add another teleconverter, and as shown in the table this will increase the magnification for the same 72mm working distance. This raises the obvious question of whether image quality would be acceptable with stacked teleconverters, and how the setup would handle. I did a 55 minutes test session in the garden, with a 1.4X teleconverter in addition to the 2X teleconverter, and captured around 260 images, of which I thought around 67 were worth keeping. That is a keeper rate of around 25%, which is actually rather higher than I usually get, especially with fairly small subjects that are moving around, as was the case for quite a lot of the captures in that session. (The images I kept from that session are in this album at Flickr. Four of the 71 images are additional, tighter crops from four of the 67 keepers.) As to handling, some of the subjects were in somewhat awkward to get at positions, so for example I had to work with the camera pointing almost straight down for some of them. The handling was no worse than with the single telecoverter, which is to say that although I don't like the weight of the setup, it was usable.

The fact that using a 2X and a 1.4X teleconverter produced results that I thought were usable suggests to me that using a pair of 2X teleconverters might work well enough too. This would let me get to the minimum scene size achievable with the Canon MPE-65 on an APS-C Canon 70D. (The minimum scene height of 3mm at 5X magnification on the 70D would need 8X magnification on the A7ii, which the 100mm 2X macro and two 2X teleconverters would be able to achieve, with a 72mm working distance compared to around 40mm with the MPE-65).

I'm not sure how useful I would find 3mm rather than 4mm scene height, as 4mm is quite difficult to handle with small subjects. (Why "with small subjects"? Because when going close in on a part of a large subject, such as a bee's head, it is relatively easy to find the subject and then navigate to the part of it you are interested in. With a small subject, especially one that is in an environment which presents lots of layers at different distances, as was the case in the test session, it can be frustratingly difficult to locate the subject. You may for example be pointing straight at it but if you are not at a suitable distance you simply won't see even a hint of it.)

I'm not even sure how much I would need the extra magnification from adding a 1.4X teleconverter. However, should I need the extra magnification the advantage of using teleconverters is twofold - they don't reduce the working distance, and an extra one can be added and removed without exposing the sensor to the external environment. Dust on the sensor is a huge problem with the very small apertures I am using. After repeatedly cleaning it, I thought I had got the sensor completely clean, but the test session showed that there is still one small dust spot. Once I have the sensor clean I really, really, don't want to open up the camera. I could of course add a close-up lens, but that would reduce the working distance.

Adding a close-up lens would also have a disadvantage in terms of effective apertures. Using a full frame camera may have some benefits such as better dynamic range, but the big difference with the A7ii and teleconverter setup (and it isn't because it is a full frame setup, it is because it is an interchangeable lens setup that allows the use of teleconverters) is that I can get smaller effective apertures, and hence larger depth of field, than I can with my close-up lens setups. With macro lenses etc the effective aperture depends on the magnification, and at present I am targeting the use of effective f/90, which gives around twice the depth of field I can get with my close-up lens setups. The following table shows how the minimum effective aperture varies with magnification for my candidate setups.

[Table corrected 3/9/2020]


1688 2 Some minimum effective apertures by scene height
by gardenersassistant, on Flickr

The green shading shows where I can reach my current target of effective f/90. This shows that the range of scene sizes for which I can achieve effective f/90 would be relatively small for a micro four thirds setup with the Laowa 50mm 2x lens. The option of adding a teleconverter is not available. Adding a close-up lens would reduce the working distance and not improve the range of scene sizes for which I could achieve effective f/90 (because close-up lenses don't cause the effective aperture to change with changes in magnification). I could use extension tubes but this would open up the sensor to dust etc, and would lose infinity focus.

[Corrected 3/9/2020] With a 2X teleconverter on the A7ii and Laowa 100mm 2X I can achieve effective f/90 over a wider range, and wider again with an added 1.4X. With an added 1.4X and an added 2X I would be able to achieve effective f/90 with all the scene sizes I would be likely to use with the setup.

Collectively these considerations incline me towards continuing with the A7ii setup at present. If it turns out that I don't feel the need to add an extra teleconverter then I will probably try the micro four thirds Laowa 50, because I really would like to be using a less heavy rig as I'm a bit concerned about how my wrists are going to hold up with the heavier kit. (I have been doing some wrist-intensive gardening recently and so I don't know if that or the A7ii setup are causing some mild twinges I've been having. If it is the camera then I might move to the lighter setup and live with any disadvantages that gives me.)
 
Last edited:
This thread must have enough in it for a book now Nick! Not sure it would top the bestseller lists but fascinating for macro geeks like us :)

For what it is worth I have now settled on a similar rig to you for bugs - ie Sony A7Rii with Sony 90mm and very often a Raynox 250. Having lost my way for a while with foolish settings I have now settled on - Full Manual with Aperture = f20, Shutter (1/250), iso=400 and flash on TTL mode to make up the balance. Definitely getting a bit more depth of field and the sharpness is holding up ok :)
 
This thread must have enough in it for a book now Nick! Not sure it would top the bestseller lists

I am sure it wouldn't!

but fascinating for macro geeks like us :)

Hopefully so.

For what it is worth I have now settled on a similar rig to you for bugs - ie Sony A7Rii with Sony 90mm and very often a Raynox 250. Having lost my way for a while with foolish settings I have now settled on - Full Manual with Aperture = f20, Shutter (1/250), iso=400 and flash on TTL mode to make up the balance. Definitely getting a bit more depth of field and the sharpness is holding up ok :)

Yes, pretty similar. I'm glad it's going well.

I'm adjusting aperture in line with the magnification so as to keep the effective aperture around f/90 for the moment. I may try making it a bit smaller later on. This compares to the maximum effective f-number of f/40 with your approach, so I'm getting twice or more DOF than you, but with the correspondingly greater loss of fine detail.than you are getting. That makes sense to me given the higher pixel count on your A7rii. In fact I'm thinking about an A7r? not sure which version, but I don't know how much use the extra pixels would be given the extra diffraction softening I'm getting. I could pull back a bit on the aperture and magnification, and then crop more so as to get the same DOF, but I'm not sure there would be any overall advantage in that.

I'm using a manual flash and keeping the power constant at 1/4, and using ISO to control the lightness of the image. Shutter speed 1/200 (I can get issues sometimes at 1/250.)
 
Thanks for the reply Nick - that is interesting (ie varying iso rather then the flash) may give it a bash but for now I am happy that I'm pretty much getting the best out of the single shot approach. Keep up the good work! :)

Since you are using TTL flash it isn't obvious to me what the benefit for you would be of varying the ISO. You would just be giving yourself something extra to handle wouldn't you? The reason I'm doing it this way is because I'm using a manual flash so I either have to adjust the flash level or the ISO, and on the camera I'm using adjusting the ISO is quicker/easier. Keeping the flash level constant also has the advantage of ensuring that I'm not using so much flash power that the recycle time gets too slow. Presumably though because you use larger apertures than I do that is less likely to be an issue for you (although that depends of course on the power of the flash you are using, the light loss from the diffusers and the distance between the flash heads and the subject).
 
Since you are using TTL flash it isn't obvious to me what the benefit for you would be of varying the ISO. You would just be giving yourself something extra to handle wouldn't you? The reason I'm doing it this way is because I'm using a manual flash so I either have to adjust the flash level or the ISO, and on the camera I'm using adjusting the ISO is quicker/easier. Keeping the flash level constant also has the advantage of ensuring that I'm not using so much flash power that the recycle time gets too slow. Presumably though because you use larger apertures than I do that is less likely to be an issue for you (although that depends of course on the power of the flash you are using, the light loss from the diffusers and the distance between the flash heads and the subject).
Well my thinking was that I could potentially use auto iso with the flash fixed at a low-ish level - so for darker scenes the iso would take the strain as it were rather than the flash - so potentially fewer reflections and less glare (with the downside of higher noise - not that there is much of it with the A7Rii - which I would recommend by the way)
 
Well my thinking was that I could potentially use auto iso with the flash fixed at a low-ish level - so for darker scenes the iso would take the strain as it were rather than the flash - so potentially fewer reflections and less glare

I've been wondering about that on and off for quite a while now - throw less light on to the scene and reduce reflections. The thing is, I haven't been able to convince myself that it works. I haven't done any methodical tests, but the odd time I've tried it I wasn't convinced it was helping. Intuitively it seems to me that yes, the bright areas will not be so bright, but everything else will be darker too, so somewhere along the line it will all need to be brought back up again. You can pull up the brightest areas less than the rest to reduce the impact of any reflections, but as long as they weren't blown at capture time you can do that anyway without having upped the ISO and decreased the light on the scene (thus increasing noise, as you point out).

Have you done tests that have shown that it works, or found a credible reference to someone who has? I'd really like to know one way or the other. (I wonder if I could work out a convincing test? Hmmmm....)

(with the downside of higher noise - not that there is much of it with the A7Rii - which I would recommend by the way)

I'm trying to get my head around whether a ii, iii or iv would be better for my purposes. Most (? all) of the more up to date functionality probably wouldn't be much or any help to me, given that I would be using it exclusively for invertebrates, and I'm using my A7ii in a pretty "dumb" mode for that, and I would use an r in the same way. (Why I would be using it exclusively for invertebrates is because of dust on the sensor issues, which I'm fed up with. Once I got the sensor perfectly clean - if I ever did, I haven't managed it yet with my A7ii despite a lot of trying, and a new camera out of the box is no guarantee of a dust free sensor, I know that from experience. - I wouldn't want to change lens. And the macro lens setup wouldn't be suitable for other things.)
 
Have you done tests that have shown that it works, or found a credible reference to someone who has? I'd really like to know one way or the other. (I wonder if I could work out a convincing test? Hmmmm....)

Not tried it Nick - just an intuitive feeling same as you. Will get back to you if I do and get any meaningful results.

I'm trying to get my head around whether a ii, iii or iv would be better for my purposes.

Well IMHO - I would reject the iii as not giving much over the ii (for your purposes(or mine)) so it's a straight choice of money-no-object= iv, otherwise the ii is great.

Cheers
 
Not tried it Nick - just an intuitive feeling same as you. Will get back to you if I do and get any meaningful results.

Same here. I really ought to try and work out a test, given that it matters to me so much!

Well IMHO - I would reject the iii as not giving much over the ii (for your purposes(or mine)) so it's a straight choice of money-no-object= iv, otherwise the ii is great.

Cheers

Now that is really useful about the rii vs riii. Thanks. I wonder how much better the rii would be than the ii that I'm using. More pixels obviously, but anything else I wonder? I need to do some digging. :)
 
I really ought to try and work out a test, given that it matters to me so much!

It looks like it doesn't work.

Below, at the top left, a flash-illuminated raw capture, exposed to the right, using ISO 1250. Top right, capture with the same flash power and aperture, using ISO 320, so it is two stops underexposed.

Bottom left, the ETTR exposed capture, adjusted in Lightroom with Exposure slider at -0.5 stop, Highlights slider pulled right down, Shadows slider pulled right up and a tone curve applied.
Bottom right, the two stop underexposed capture, with Exposure slider at +1.5 stops, with the same Highlights, Shadows and tone curve adjustments.

The bottom two versions look pretty much exactly the same to me, and they were made to look the same with the same adjustments apart from a two stop difference in lightness (what Lightroom calls Exposure, which of course it isn't).

This strongly suggests to me that my guess was right that if the highlights are not blown then underexposing won't help manage the brightest areas. You will get more noise though as the image was made with less light. EDIT. Wrong! These images were made with the same amount of light. I think they will have the same amount of noise (more or less - it depends on whether the sensor is ISO-invariant. But even if it isn't I would guess that the difference in noise would be rather small, and if they were different, I have no idea which of them would be the noisier.)


1690 1 - Two stop underexposure comparison
by gardenersassistant, on Flickr
 
Last edited:
Interesting the underexposure and pulling back brightening with the Sony camera
I would never get away with doing that with canon crop cameras. I could go quite high with ISO for my natural light shots but always had to nail the exposure, I normally went to the right a bit
Haven’t tried that on my full frame canon , underexposing
 
Interesting the underexposure and pulling back brightening with the Sony camera
I would never get away with doing that with canon crop cameras. I could go quite high with ISO for my natural light shots but always had to nail the exposure,
I normally went to the right a bit
Haven’t tried that on my full frame canon , underexposing

Interesting about the Canon crop cameras being poor with lifting from under-exposure compared to using a high ISO. I tried an experiment with my 70D and A7ii.

On the A7ii I wanted to keep the (EF mount) 2X teleconverter attached so as to avoid getting dust on the sensor. I have been having a lot of trouble with that with the very small apertures I am using. I used an EF mount Sigma 70-300 at 70mm on the 2X teleconverter. On the 70D I used an EF-S 55-250 and roughly matched the field of view with what I was getting with the A7ii. I used a tripod, but didn't use a remote release (because I seem to have lost my Canon wired release and I don't have one for the A7ii, and I didn't use a timer because I was in a bit of a hurry. As a result some of the slower shutter speed shots are unsharp, but I don't think that matters much for this particular comparison.

I used f/5 for the 70D and an equivalent f/8 for the A7ii.

Here is the scene. I examined an area near the centre of the scene at 100%. Using 100% for both cameras gives a slightly different field of view for the comparisons as the 70D is 20 mpix and the A7ii is 24 mpix.


1696 1 The test scene
by gardenersassistant, on Flickr

I captured 12 images as raw with each camera: ISO 100, 200, 400, 800, 1600, 3200 all with zero exposure compensation, and ISO 100 with 0, -1, -2, -3, -4, -5 stops of exposure compensation.

The captures looked like this.

1696 2 Unadjusted captures
by gardenersassistant, on Flickr

In Lightroom I then increased the lightness of the 1 stop underexposed image by 1 stop, the 2 stop underexposed image by 2 stops etc. After these adjustments the images looked like this.


1696 3 Adjusted captures
by gardenersassistant, on Flickr

I then compared the (roughly) central area of the frame as shown below for ISO 3200. (Note that Lightroom does not apply luminance noise reduction by default.)
  • Top left, 70D, ISO 100, 5 stops underexposed, raised by 5 stops
  • Top right, 70D, ISO 3200, not underexposed
  • Bottom left, A7ii, ISO 100, 5 stops underexposed, raised by 5 stops
  • Bottom right, A7i, ISO 3200, not underexposed.

1696 8 5 stop under-exposure
by gardenersassistant, on Flickr

Then the same type of comparison for 4, 3, 2 and 1 stops underexposure.

4 stops underexposed

1696 7 4 stop under-exposure
by gardenersassistant, on Flickr

3 stops underexposed

1696 6 3 stop under-exposure
by gardenersassistant, on Flickr

2 stops underexposed

1696 5 2 stop under-exposure
by gardenersassistant, on Flickr

1 stop underexposed

1696 4 1 stop under-exposure
by gardenersassistant, on Flickr

The underexposed Canon images are indeed worse than the related higher ISO shots. The same is true for the A7ii, but to a significantly lesser extent.
 
Really interesting Nick thanks for doing that , it does show the weakness of the Canon sensor to underexposure
I’m routinely at either 400 or 800 ISO for macro natural light shots and normally expose to the right and the noise is minimal certainly don’t need to use noise reduction
For my 7 D mk1 and 550D I use plus 2/3 on my 7D 2 and 6D 2 plus 1/3 exp compensation
I have accidentally over exposed by quite a lot (I use raw ) and it’s absolutely fine but the other way underexposure and brightening in post is noisy ( this was with 7D1) I accidentally knocked the top dial switched to manual mode , can’t do that with the later canons there’s a locking mechanism
 
In this post I had come to the conclusion that I would use my Panasonic G9 with an Olympus 60mm macro for natural light botanical subjects and for invertebrates I would use my Sony A7ii with Laowa 100mm macro and a 2X teleconverter and a Yongnuo twin flash mounted on the front of the Laowa 100. There have been some developments since then.

I am now using a 1.4X teleconverter in addition to the 2X teleconverter on the A7ii. This gives me a setup which goes from infinity focus down to scenes a little over 4mm high (5.6X magnification) by turning the focus/magnification ring on the Laowa 100 by about 120 degrees. It also means that I can use an effective f-number of f/90 for all scenes smaller than around 56mm high (around 0.5X magnification), as illustrated in the following table. f/90 gives me twice the depth of field that I can get with my close-up lens setups.


1706 1 Some minimum effective apertures by scene height
by gardenersassistant, on Flickr

Adding even more glass softens the images even more, but it seems to work ok for my purposes. With the extremely small apertures I am using the images are extremely soft anyway, and adding a bit more softness hasn't made any difference that I have noticed. (I have not done any careful like for like comparisons. I'm in a "good enough is good enough" state of mind about this at the moment. Perhaps I'll test it later. Then again, perhaps I won't. :) )

I have had detailed discussions in this thread at dpreview where I asked for advice as to whether a low pixel count Sony A7siii would help with the issues I am having with flash reflection "hot spots" on insects etc with reflective curved body surfaces. I got lots of helpful information, but the overall it seemed inconclusive to me. However, it also introduced me to the technique of cross polarized illumination, and I spent some time experimenting with that (written up here at dpreview).

Cross polarized illumination deals very effectively with ugly flash reflections. However, it has side-effects which make it unsuitable for my purposes. I have gone back to trying to improve my diffusion. On the Yongnuo twin flash I tried using diffusers which had three layers of silk held in place by step rings, but a trial in the garden showed that they didn't work very well. Today I tried replacing the Yongnuo with the Venus Optics KX800 that I have been using for several years on my bridge and micro four thirds cameras with close-up lenses.


1705 Illustration 1 - KX800 with Plastic paper diffusion
by gardenersassistant, on Flickr


1705 Illustration 2 - KX800 with Plastic paper diffusion
by gardenersassistant, on Flickr


1705 Illustration 3 - KX800 with Plastic paper diffusion
by gardenersassistant, on Flickr

I don't know if it would be practical hand-held out in the field, but by adding a Raynox 250 I can get down to a scene width of around 4mm (9X magnification) and around 2.5mm (14X magnification) with a Raynox 202. It turns out that this flash setup will illuminate scenes from around 0.5X to 14X without needing to make any adjustments.

The focusing light at the top is removable. I had a short session this afternoon to test this setup. It worked ok. As with my close-up lens setups, this diffuser arrangement does tend to restrict what I can get at compared to the much smaller diffusers I have been using on the Yongnuo. There again, if it gives me light I like better then that is a trade-off I'm prepared to live with. It is after all no worse than I have been living with for several years with the KX800 using close-up lenses. And in practice I have found it turns out to be less restrictive than one might imagine from the look of it.

I don't know how today's test shots turned out; they are still making their way through the processing chain (which I'm also experimenting with. More on that another day.)
 
Last edited:
I don't know how today's test shots turned out; they are still making their way through the processing chain (which I'm also experimenting with. More on that another day.)

I have posted eight of the test shots here in the forum.

My post processing experiment with this set of images was to drop Silkypix and use PhotoLab, Lightroom and DeNoise AI. It worked well enough that I think I will stick with that for the moment for invertebrates. For single-image captures of flowers etc I will continue to use Silkypix, Lightroom and DeNoise AI.
 
I spent a lot of the past three weeks or so clearing out my backlog of unprocessed images. I have reclaimed about 2 terabytes of disk space, which I think means I probably went through over 50,000 images and videos. They were mainly from 2017, 2018 and 2019, with a few from earlier. After longlisting, initial processing, shortlisting and finishing I ended up keeping about 1500 of them, a mixture of stills of invertebrates and flowers etc, and focus-stacked images of flowers etc produced from videos. I have posted some of the invertebrate images here in the forum and some of the botanical images here.

It was a good opportunity to test the previously mentioned Silkypix + Lightroom + DeNoise AI workflow for botanical stills and the PhotoLab + Lightroom + DeNoise AI workflow for invertebrates. For the last couple of invertebrate sets I used the just released PhotoLab version 4 with its machine learning Deep Prime noise reduction, which seemed to work well and also faster than the Prime noise reduction of the earlier versions. For all the stills I used Fast RawViewer for initial longlisting. For stacking from video I used PhotoFun Studio for an initial look at the videos and Helicon Focus to do the stacking and retouching, followed by Lightroom and DeNoise AI. In a few cases, for stills and stacks, I used Photoshop for cloning that was too difficult in Lightroom. In all cases I used Faststone Image Viewer to organise files and folders and to review processed images. These arrangements proved to be well up to this demanding task and I now regard them as settled (until the next thing comes along to try!)

Two things struck me as I went through all these images. One was that for invertebrates the large DOF (macro lens and teleconverters) arrangement I have recently been experimenting with really does seem to have some significant advantages. All of the invertebrate images I worked on from the backlog were from earlier, close-up lens setups. I was struck how high the failure rate was for these compared to the large DOF examples I recently worked on, with failures stemming both from complete focus failures and also in-focus but with focus not quite in the right place. The large DOF setup uses manual focus rather than autofocus and does not suffer from the gross focus failures that autofocus gives sometimes. As to focus placement, the close-up lens setups are limited to f/45 full frame equivalent and the focus placement has to be exactly right to get adequate DOF coverage of the subject to suit my preferences; quite often the focus placement is not exactly right. With the large DOF setup I typically have twice or more DOF and this means there is much more latitude as to exactly where the centre of DOF lies, and hence the much lower failure rate. At least, I think that is what is going on. I need more experience with the large DOF setup, including with smaller subjects, to check that the higher hit rate is a reality and not just because of the (limited range of, relatively easy) subjects I have tackled with it so far.

For botanical subjects the exercise reinforced my feeling that I really don't have a systematic preference for stacks as against stills. A lot of the botanical sessions I worked on had both stills and stacking videos for most of the subjects. In some sessions I preferred stills, in other sessions I preferred stacks. And in other sessions it varied from scene to scene. It was very much a combination of what worked in a technical sense for a particular scene, and what "look" appealed to me for a particular scene. As to what worked in a technical sense, for example stacks might not work in bright breezy conditions whereas stills might, and alternatively in poor light stacks might be possible because stacking captures use f/2.8, while stills using a smaller aperture to get the DOF I liked might have too slow an exposure or too high an ISO. As to the "look", stacks with their relatively plain backgrounds and strong separation between subject and background can have a strong visual impact, but the more complex backgrounds of some stills can produce images that look more interesting and varied to me. I did notice though that there is quite a lot of overlap in the look as between stills and stacks; when I was reviewing the images at the end of the exercise, thinking about which ones to post, I discovered that I could quite easily be wrong about whether a particular image was a still or a stack.
 
Last edited:
I need more experience with the large DOF setup, including with smaller subjects, to check that the higher hit rate is a reality and not just because of the (limited range of, relatively easy) subjects I have tackled with it so far.

A short session yesterday was promising. I kept the eight images in this post from the 130 I captured, which is about 6%. That is acceptable for my purposes, especially given that the subject was quite small, was moving around a lot and was on a leaf that was moving in the breeze. Also, if I had not had those eight I would probably have regarded a number more of them as usable. I think the image quality of those eight is probably better than I have previously achieved. (I don't have many to compare them with and I haven't managed to find any of them yet.)
 
Testing continues with the Sony A7ii and Laowa 100mm 2X macro lens with 2X and 1.4X teleconverters. A couple of days ago I had a session testing out effective f-numbers a bit beyond the f/90 I had previously been targeting. The only invertebrates I can find at the moment are rather small, and I spent most of the session using the full 5.6X magnification, or close to it. When it came to processing the test images it struck me that I would really like a bit more magnification, or alternatively more scope for cropping. The image that prompted these thoughts was this one.


1814 06 2020_11_01 DSC06841_PLab4 LR 1300h DNAIcAUTO
by gardenersassistant, on Flickr

This was not the smallest of subjects that I want to tackle (although I think it was a bit small for this type of springtail), but even though I was, I think, using maximum magnification I needed to do this crop. That didn't really get me as close in as I would have liked, but I was wary of cropping even more.


1814 06 crop
by gardenersassistant, on Flickr

Since the A7ii is working well with the odd combination of lenses I am using my first thought was to use a similar camera, an A7rii, which has 42 megapixels compared to the 24 megapixels of the A7ii and should allow more cropping. However, feedback to a question I raised at dpreview suggested that an A7rii might not be feasible, and in any case because of the heavy loss of fine detail from diffraction softening with the very small apertures I am using, I had doubts about whether cropping deeper would actually be useful, even if I had more pixels, because of the underlying lack of detail.

My thoughts then turned to getting greater magnification. One option would be to add a Raynox 250, although this would shorten the working distance significantly. I might try that, but I decided to try using a pair of 2X teleconverters instead of one 2X and one 1.4X. I took delivery of a second 2X teleconverter this morning.

The Laowa 100mm macro has a maximum of 2X magnification, with a working distance of around 72mm and apertures from f/2.8 to f/22, covering a minimum scene size of 18mm x 12mm. With a pair of 2X teleconverters it becomes 400mm with a maximum of 8X magnification, with a working distance of around 72mm and apertures from f/11 to f/90, covering a minimum scene size of 4.5mm x 3mm. When I fitted the pair of 2X teleconverters everything seemed to work as expected. Later on I took it out for a short session in the garden, using the KX800 twin flash.

I only found two subjects. I think they may have been a springtail nymph and possibly an aphid nymph. Both were small, somewhere around 1mm in length I think. Both were moving around and, especially the aphid, stayed in view long enough for me to have a number of attempts. I was working at or near to maximum magnification.

The setup handled fine. For quite a lot of the shots the focus was not placed where I would have liked, but given that the subjects were quite small and were moving around I was not surprised about that. Of more concern was that adding even more by way of teleconverter power would soften the images even more than before, perhaps to the point of not being usable. With the aphid I had time to vary the f-number a lot to get some idea of what sort of aperture to best use at the higher magnifications. I have posted eight of the shots here in the forum and several of them are soft not so much from a misplaced focus plane but, I think, from too small an aperture. This one, which used not quite as small an aperture, is probably the best of them in terms of detail captured, and it seems promising enough for me to continue experimenting with this setup.


1816 01 2020_11_03 DSC07676_PLab4 LR 1300h DNAIcAUTO
by gardenersassistant, on Flickr

One thing I will do is to try cleaning the sensor, again. The small apertures, coupled with the aggressive processing I am using, really makes dust spots show up strongly. And there were also some softer but much larger circular problem areas. Given how soft they are I doubt they are dust on the sensor. It may be something on the front or back of one (or both) of the teleconverters. Or perhaps dust inside one of the teleconverters. I bought the second one second hand (at a fairly modest price) from Wex Photographic, and they mentioned that it had dust inside, saying that it would not affect photos. However, given the way I am using it perhaps that is what I am seeing. I need to do more testing.

In the meantime though, I think this avenue is looking quite promising.
 
Last edited:
Finally, approaching seven years after starting this thread, I am putting away my close-up lenses. I have given my blog its first update in the more than 18 months since I created it to reflect this change (this post and the previous two). I think it may also be time to update my portfolio, which I have also not updated since I created it along with my blog in early 2019.

In the past week I have had another half a dozen sessions with the twin 2X teleconverter setup on the A7ii, with the Laowa 100mm 2X macro and Venus Optics KX800 twin flash. This has convinced me that I can get results that are more to my liking with this setup than by using close-up lenses (on bridge, micro four thirds or APS-C cameras). Also, to my surprise, although having usable autofocus at all magnifications was one of the attractions of using close-up lenses, it turns out that I am getting at least as good a focus hit/fail ratio using manual focus with the A7ii as I get using autfocus with close-up lenses. I think part of the reason for this is that because I have a lot greater depth of field with the A7ii setup (two to four or more times greater than with my close-up lens setups) it is much easier to get the subject into the in-focus area.

This also means that my bridge cameras (FZ330, FZ200) are going out of use. My equipment will be:
  • Botanical subjects. Panasnonic G9 and Olympus 60mm macro
  • Invertebrates. Sony A7ii, 2 x 2X Kenko teleconverters, Laowa 100mm 2x macro, Venus Optics KX800 twin flash
  • Common birds in flight (if/when I get to do any more of this). Canon 70D, 55-250 EF-S STM
  • Other. Panasonic TZ90 travel camera.
I am continuing to work on my post processing.

I have dropped Silkypix. For botanical stills I am now using DXO PhotoLab, Adobe Lightroom and Topaz DeNoise AI, and using Helicon Focus, Lightroom and DeNoise AI for botanical stacks, which I will also use for invertebrate stacks if I get round to doing any. For invertebrate stills I am using PhotoLab, Lightroom and DeNoise AI, and will also use this for any common birds in flight and other images, if and when I do any of this.

For invertebrates I am (work in progress) changing my processing workflow to reduce the time I need to spend on it. Rather than preparing an initial longlist using FastRawViewer, I am putting all the raw files from a session (typically 500-600 at the moment) into PhotoLab and using a single preset on all of them (rather than using ISO-specific presets). This produces TIFF files which I then load into Lightroom, apply AutoTone and export JPEGs at my target output size of 1300 pixels high. I then load the JPEGs into DeNoise AI and produce noise reduced, sharpened 1300 pixel high JPEGs using the same settings for all of them. Getting to this stage takes a fair amount of computer time, although it isn't too bad as I have a rather powerful PC with a powerful graphics card the PhotoLab, Lightroom and DeNoise AI all make use of. However, it takes very little of my time; two or three minutes perhaps.

The reason for doing this is that because of the extremely small apertures I am now using it is very difficult to "read" the images when looking at them for longlisting in FastRawViewer or any other image viewer. I find it impossible to tell how much detail will be revealed by post processing, which makes it very difficult to be confident about which images to include and which to reject. With this new workflow I don't take any decisions about the images until I have versions that give me a good idea of their potential. I then go back and forth between Lightroom and DeNoise AI in several cycles refining the selection and processing.

I have had to add an extra step at the end of the selection/processing, prior to the final sharpening/noise reduction, which is to go through the list of selected images and examine them closely to try to remove all dust spots. The very small apertures I am using, in combination with some very aggressive post processing, make any dust spots stand out. This is not just dust on the sensor. There are another eight surfaces on which dust can settle (both ends of each teleconverter, both ends of the macro lens and both sides of a UV filter on the macro lens). Dust on any of these surfaces can produce visible results, ranging from small, dark spots for dust on the sensor and progressively larger and more diffuse circular areas on the surfaces further away from the sensor. I have thus far found it impossible to clean the equipment to the point where no blemishes are visible in plain areas on the final images. I have decided to accept that rather than spend additional hours trying to get the equipment perfectly dust free, which might not last long anyway, I will to spend some time for each series of images dealing with dust spots, developing an approach to that which is as efficient and effective as I can make it.

I am feeling reinvigorated as far as invertebrates are concerned. There are two aspects to this. One is that I am getting results that I like better, including with small subjects with which I have previously had limited success. The other is that I suddenly have a much better range of subject matter. I don't know why I never did it before, but I contacted the church warden of the church opposite our house and got his permission to roam around the church grounds taking photos. The grounds are a rough square of around 100 metres on the side, of which the church and small car park take up well less than a quarter. The rest is a relatively unmanaged wood (difficult to get at quite a lot of it because the brambles are so extensive - I explored the whole site, and have extensive scratches to prove it despite wearing my heavy workman's trousers). I have put some images of the grounds in the next post. These are taken from this album at Flickr, which gives a more accurate indication of the balance between the various types of area than the images in the next post, which are more like one image from each type of area.

I have felt for some time that our garden, to which I have now been restricted for many months, has not provided enough interest in terms of numbers and variety of invertebrates. There seems to be plenty more going on in the church grounds. At the moment this is almost all rather small subjects, which is fine as I am now better able to tackle these. I have no idea whether there will be larger subjects come the spring and summer. I hope so, but I know from the nature reserves that some areas are radically better than others for the sort of subjects I have typically tackled over the past few years.

My current activity is to work out how best to use the Sony setup. I don't yet have a good feel for what apertures to use for different types of subject for the wider range of magnifications I now have available, and that is what I am currently exploring, mainly at the moment over at the church grounds.
 
Last edited:
I think you are very lucky to have that church opposite you,

Absolutely.

it must be amazing in the summer for insects.

That would be good. I explored more of it today and a lot of it seems to be quite barren at the moment as far as findable invertebrates go. It seems that I tried the (currently) best little area first.
 
Last edited:
Absolutely.



That would be good. I explored more of it today and a lot of it seems to be quite barren at the moment as far as findable invertebrates go. It seems that I tried the (currently) best little area first.
It is winter, I think the biggest things that I see at the moment are woodlice. If I lived in a rural environment I might get moths at night but sadly I don't.
 
Copied from a post at dpreview, here is a short overview of my photographic 2020. I have had to split it over two posts because there are 10 images. There are 1300 pixel high versions of the images in this album at Flickr.

A year of three techniques

Most years almost all of my photos are close-up/macros, mainly in our garden with occasional trips to local nature reserves. This year all of them were close-up/macros. I didn't get to any of the nature reserves, so it was mainly in our garden, or at the end of the year in a very small patch of woodland next to the church opposite our house, which I had not previously sought permission to go into.

This was the year when focus-stacking from 6K video fully settled into my "toolkit" as a routine method to use with flowers, buds, berries, seed pods etc using natural light, hand-held. I used it most of the time apart from when it was particularly breezy. Here are three of them.

#1 A stack of 78 6K video frames

1863 01 006 1732 1 2020_05_15 P1184931_0007 78 C4-Edit LR 1300h DNAI1,13
by gardenersassistant, on Flickr

#2 A stack of 189 6K video frames

1863 02 1859 06 2020_12_31 P1246871_0199 189 C4 LR 1300h DNAIauto
by gardenersassistant, on Flickr

#3 A stack of 39 6K video frames

1863 03 1853 10 2020_12_19 P1246473_0027 39 B28,4+innerC4 LR 1300h DNAI0,32
by gardenersassistant, on Flickr


I used a second hand-held, natural light technique for flowers etc - aperture bracketing. This gives me 7 shots from f/2.8 to f/22 for a single press of the shutter button. Having a set of images at different apertures lets me choose the one (or occasionally more than one) where I best like the balance between the amount of the subject that is in focus and the way the background looks. In very breezy conditions this would be the only technique I would use, with focus stacking being impractical. Otherwise I would use both video (for focus stacking) and aperture bracketing. Using both techniques several times for each subject meant that if none of the stacks worked I might still have something to work with from the aperture bracket sets. And sometimes I just prefer the look of single stills, or sometimes I only feel like shooting stills (and sometimes only videos), and for some subjects I have learnt that stacking is highly unlikely to work and so I often don't bother with video for those subjects any more.

Here are three stills that were taken from aperture bracket sets.

#4 I chose the maximum aperture f/2.8 shot for this one.

1863 04 1661 4 2020_06_25 P1209687_PLab3 SP9-Edit LR 1300h-DNAI
by gardenersassistant, on Flickr


#5 I chose f/11 for this one. That was probably the first one in the sequence where I felt enough of the subject was sufficiently in focus for my purposes, probably referring to the yellow "fur" on the right hand petal, and that left the background better defined (more intrusive) than in the previous one. I know that many people much prefer plain backgrounds, and will for example use use plain cards behind the subject or use background replacement in post processing to produce plain backgrounds, but I'm happy with quite busy backgrounds if to my eye they complement the subject or at least don't fight too hard against it (a matter of personal visual preference of course). In fact I like the variety that less defocused backgrounds can produce. I find the relatively plain backgrounds I get from stacks a bit monotonous sometimes.

1863 05 1733 1 2020_05_24 P1185491 SP10 LR 1300h DNAI2,18
by gardenersassistant, on Flickr


#6 And sometimes a smaller aperture is needed to get enough of the subject into focus to suit my visual preferences, which lean towards getting the whole subject in focus where practical. As well as involving a trade-off as to the look of the background, using smaller apertures can involve trade-offs involving shutter speed and ISO. I have that side of it automated now, so the camera uses base ISO as long as the shutter speed is 1/80 sec or faster (as for #4 above), and then hold the shutter speed at 1/80 sec as it increases ISO up to a maximum of ISO 3200 (as for #5 and #6), at which point (which I don't reach very often) it holds the ISO at 3200 and slows the shutter speed.

1863 06 1659 27 2020_06_21 P1198149_PLab3 SP9 LR 1300h DNAIc
by gardenersassistant, on Flickr



Continued in next post
 
The third technique I used this year was entirely new (for me) and used with a combination of kit I had not used before, for photographing invertebrates down to 1mm or so long, hand-held, often as they move around, using single stills with flash. This involves the use of very small apertures to produce a lot greater depth of field than you normally get with single images, especially for small subjects (up to 8 times bigger than the largest depth of field I had previously been able to get). Such small apertures produce extremely soft images which need strong post processing to produce results that are usable for my purposes (which is another matter of personal preference and judgement of course).

I started experimenting with this technique in the summer and by the end of the year I had found a combination of kit I was comfortable using, which involved a 2X magnification macro lens and two 2X teleconverters to give a total maximum magnification of 8X and which let me set f-numbers up to f/90. By the end of the year I had settled on setting the f-number to around f/45 most of the time, which taking account of the magnifications I was using gave me effective f-numbers up to around f/400. And by the end of the year I had sorted out post processing to use on these extremely soft images.

#7 One of the things the extra depth of field helped me do was to get shots of highly active insects and spiders which didn't stay in one place for long and sometimes didn't stay still when they did stay in more or less the same place for a while. This first example is of modestly sized wasp which was in one place for a relatively long time as it tried to pull flesh off of a pigeon corpse, but by the end of the year I was photographing much smaller animals in the 1 to 2mm size range as they moved around.

1863 07 1712 021 2020_08_30 DSC00211_PLab3 LR 1300h DNAIcAuto
by gardenersassistant, on Flickr

#8 I also found that I could get much closer in on some of my favourite subjects than I had been able to previously, like this snail.

1863 08 1850 042 2020_12_17 DSC08301_PLab4 LR 1300h DNAIcMedHi
by gardenersassistant, on Flickr

#9 And I had much more success with small subjects in the 1-2mm size range, like this springtail.

1863 09 1824 21 2020_11_08 DSC09794_PLab4 LR 1300h DNAIcAutoMed
by gardenersassistant, on Flickr

#10 And finally, something entirely unexpected turned up at the end of the year. It turned out that I could use the hand-held, flash-based, small aperture, large depth of field approach for water droplets like these.

1863 10 1846 07 2020_12_12 DSC07463_PLab4-Edit LR 1300h DNAIcMedHi
by gardenersassistant, on Flickr


It has been a horrible year in so many ways for so many people. Isolated in our own little world with my wife it turned out to have some significantly positive aspects as far as my photography goes, for which I am extremely grateful.
 
Last edited:
I’ve started trying out LED lights instead of flash to add a bit of fill light
As soon as I start finding more insects I’ll post up pictures
It worked well today on a green hairstreak butterfly today but I missed focus so will try again
The idea is that I can see see the light effect before I take the shot and can vary the light intensity and angle easily
 
Back
Top