First examples using lasagne tin diffuser

GardenersHelper

In Memoriam
Messages
6,344
Name
Nick
Edit My Images
Yes
I described the construction of a "Lasagne tin diffuser" in this thread. (My wife suggested calling it a "Quaker bonnet diffuser" because of the shape of it, for example see these, especially the ones from half way down to the bottom of the page.) Since building the diffuser I have used it for several sessions but not processed any of them yet. I had another session with it in the garden yesterday and have gone through the images, selected which ones to process and processed them. They are in this album at Flickr, from which the images in this post are taken.

All of these images were captured hand-held with minimum aperture (equivalent to about f/29 in Canon APS-C terms) and base ISO, using a Raynox 150 on my FZ200, with a shutter speed of 1/2000 sec apart from the last one which was (don't know why I changed it) 1/1000 sec. The reason I used such a fast shutter speed (1/2000 sec is the fastest for which flash works effectively on the FZ200) was because the conditions were bright and I quickly discovered that in a few cases which had a very light coloured background that was highly reflective (none of these I think) the ambient light was strong enough to cause over-exposure of highlights at 1/1000 sec. (The second image in the Flickr album has the bright background that was giving me problems.) I wanted the flash to be the dominant light source, because it was a test session for the diffuser, and so I upped the shutter speed as far as I could.

1.

0744 015 2015_06_27 P1930623_DxO LR 1300h
by gardenersassistant, on Flickr

2.

0744 027 2015_06_27 P1930713_DxO LR 1300h
by gardenersassistant, on Flickr

3.

0744 030 2015_06_27 P1930780_DxO LR 1300h
by gardenersassistant, on Flickr

4.

0744 035 2015_06_27 P1930801_DxO LR 1300h
by gardenersassistant, on Flickr

5.

0744 044 2015_06_27 P1930846_DxO LR 1300h
by gardenersassistant, on Flickr

6.

0744 051 2015_06_27 P1930929_DxO LR 1300h
by gardenersassistant, on Flickr

7.

0744 053 2015_06_27 P1930934_DxO LR 1300h
by gardenersassistant, on Flickr

8.

0744 093 2015_06_27 P1940249_DxO LR 1300h
by gardenersassistant, on Flickr
 
Lighting looks very Nice Nick...

Thanks Bryn.

now what is your opinion??

Ha ha, you noticed. :D

Well. I'm not sure. That's why I said nothing.

I tried to pick good ones for this post, so as to try to make for an attractive post to look at. In retrospect I feel a bit uneasy about that though, given that this is as much as anything a post about the diffuser, and I really ought to give a fair, balanced view of it. So I'm afraid my choice for the OP may have been misleading.

But even with these I'm not sure.

#1. I've been trying to get nice pictures of these for years. It's been getting better recently, but I think the results I got this time were better than I've achieved before.

#2. I find these pretty difficult. The shiny curved surface on the thorax always suffers from colour change in the area behind the head oriented towards the flash and often gets bleached or blown completely. This example came out better than I've seen before I think. It varied. Here is one that was badly bleached.


0744 022 2015_06_27 P1930671_DxO LR 1300h
by gardenersassistant, on Flickr

Here is one that I think is a bit better on the thorax, but the narrow line along the abdomen is ugly, and badly bleached/borderline blown at the end nearest the thorax.


0744 020 2015_06_27 P1930667_DxO LR 1300h
by gardenersassistant, on Flickr

#3. Not really happy with the line of white/highlight areas along the body.

#4. OK I think, but an easy subject.

#5. The ridged area at the back end of these guys has been giving me severe problems with all the various diffusers I've been trying recently. It is highly dependent on the angle. It can look like a featureless bright nothingness, even when it isn't strictly speaking blown. And sometimes it's ok. So I don't really know what to make of this one. It looks ok. But from a few degrees different angle? Don't know.

#6. Not perfect, but not too bad.

#7. Bleached/blown area on the thorax. Difficult subject, but all the same, not good. Not sure whether there is a bit of over-exposure playing into that. And here is a worse version. Definitely blown.


0744 052 2015_06_27 P1930933_DxO LR 1300h
by gardenersassistant, on Flickr

#8. I think this shot is ok, but here it is from another angle. Not so good. See the blown line on the abdomen, and the wing is a bit iffy.


0744 091 2015_06_27 P1940242_DxO LR 1300h
by gardenersassistant, on Flickr

And then there are ones I didn't post. For example, I don't like the blown lines along the legs in this one.


0744 013 2015_06_27 P1930611_DxO LR 1300h
by gardenersassistant, on Flickr

Or the smaller blown areas on the legs in this one.


0744 029 2015_06_27 P1930722_DxO LR 1300h
by gardenersassistant, on Flickr

Nor the blown area on this one. (Mind you, this shot did show that the diffuser throws light ok for small subjects. This one used the 150 and 250 stacked. Uncropped it was about 3.5:1.)


0744 010 2015_06_27 P1930565_DxO LR 1300h
by gardenersassistant, on Flickr

There are a couple of small areas blowing out on the thorax here.


0744 011 2015_06_27 P1930574_DxO LR 1300h
by gardenersassistant, on Flickr

So, mixed results. Not a cure-all. But quite possibly the best arrangement I've used thus far in terms of achievable (given appropriate operator input) image quality. In terms of handling it has been ok by and large. It's light and robust, which is good. It does reach forward more than my pie tin diffuser though. I've noticed that a bit when working in amongst foliage. It gets in the way a bit more and makes some angles unachievable that I might be able to get at with some other arrangement, but I haven't felt that has been significant problem.
 
I'd not thought of moving the shutter speed upwards to make it all flash lighting. Did that though cause the flash to work harder - put more light down ?

I generally think the results are fairly good. I would not want my images to be totally flat - so I think you need some highlights put down, without blowing them out.

My question would be is there detail in the whites ? Can these be recovered if you were to pull the highlights down in a RAW converter, if you shot them in RAW or dogde/burn whatever the right way is in PP Software. If so then I think the diffuser is working pretty well - would like to see if off a really reflective subject, like a beetle for instance.

I am assuming the images are unaltered from camera for a diffuser test ?
 
#2 I don't see what you are... not seeing the bleached on the abdomen in fact it looks well exposed maybe you are referring that it looks different to the underside but that part is in shadow from the flash so would expect colour is disappear.
 
#2 I don't see what you are... not seeing the bleached on the abdomen in fact it looks well exposed maybe you are referring that it looks different to the underside but that part is in shadow from the flash so would expect colour is disappear.

The bleaching referred to the first of the additional images. The area outlined in red here on the thorax.


0744 022a 2015_06_27 P1930671_DxO LR 1300h annotated
by gardenersassistant, on Flickr

The abdomen comment related to the second additional image. Although the bright area on the Thorax (outlined in green) looks a bit better to me on this one, the blue outlines shown the line I mentioned and the red outline shows the bleached/borderline blown area at the left hand end of that line.


0744 020a 2015_06_27 P1930667_DxO LR 1300h annotated
by gardenersassistant, on Flickr
 
The bleaching referred to the first of the additional images. The area outlined in red here on the thorax.


0744 022a 2015_06_27 P1930671_DxO LR 1300h annotated
by gardenersassistant, on Flickr

The abdomen comment related to the second additional image. Although the bright area on the Thorax (outlined in green) looks a bit better to me on this one, the blue outlines shown the line I mentioned and the red outline shows the bleached/borderline blown area at the left hand end of that line.


0744 020a 2015_06_27 P1930667_DxO LR 1300h annotated
by gardenersassistant, on Flickr

Still not seeing the bleaching Nick, Looking at the largest I can on flickr... the colour is all there and no detail has been lost.

#1 is a little worse on the thorax but still detail is there.
 
I'd not thought of moving the shutter speed upwards to make it all flash lighting. Did that though cause the flash to work harder - put more light down ?

Don't think so. If the shutter is open at least as long as the flash pulse (and the synchronisation is good) there is no need for the flash to work harder. It would have to work harder if I was using HSS/FP flash, but the FZ200 has a leaf shutter (I believe, although I've never actually found a solid reference to that) and it doesn't have the flash syc issues that a focal plane shutter has, so it can use ordinary, non-HSS/FP flash at much higher shutter speeds than focal plane shutter cameras.

I generally think the results are fairly good. I would not want my images to be totally flat - so I think you need some highlights put down, without blowing them out.

Agreed. But the problem I see (although I see that Bryn doesn't), for example in both the annotated images in the previous post, is that the areas on the thorax are not just a bit brighter, they are different in colour (different in a way I don't find appealing). I assume that one (or possibly two) of the channels is getting clipped, so although it isn't blown to featureless white the colours are wrong. Sometimes the colours are wrong (for example I've seen a red poppy petal turn bright yellow when overexposed). Sometimes the images get what I think of as "bleached", where the colours move towards white. Although the hue may still be more or less ok, the saturation gets lower and the luminance higher, and it looks very different from the as seen scene. I get this a lot with petals catching the light at a certain angle.

My question would be is there detail in the whites ? Can these be recovered if you were to pull the highlights down in a RAW converter, if you shot them in RAW or dogde/burn whatever the right way is in PP Software. If so then I think the diffuser is working pretty well - would like to see if off a really reflective subject, like a beetle for instance.

I am assuming the images are unaltered from camera for a diffuser test ?

No, they are all altered. I always post process, so a test for me is a question of what state I can get an image to with post processing. The out of the camera version (and I'm not really sure what that means for raw) doesn't matter to me. A lot of my stuff looks more or less horrible (for example because of under-exposure) before it is processed.

I always shoot raw. I very frequently (much more often than not) pull the highlights down. I suspect this is partly because of the characteristics of the small sensor on the FZ200 than I'm using. I have to do it even though I often underexpose.

I'd like to see it with a more highly reflective subject too. I haven't seen one yet to try it on. Unfortunately, I'm pretty sure I'm not going to like it when I do.
 
Still not seeing the bleaching Nick, Looking at the largest I can on flickr... the colour is all there and no detail has been lost.

#1 is a little worse on the thorax but still detail is there.

We are seeing this very differently. Here is a 100% crop of the processed image (this is the first of the two additional images). This was cropped in Lightroom from the as posted version. I've just checked - it has had highlights pulled down to the maximum extent that Lightroom allows. (As always, this is done from a raw original. In fact the raw was sent to DXO for its PRIME noise reduction then back to Lightroom as DNG. So no loss in that process.)


P1930671_DxO Crop
by gardenersassistant, on Flickr

The bright area has a quite well-defined edge, and looking at the area just above the middle of the top edge I can see green and gold spots. These suddenly disappear at the edge of the bright area. To my eye, that is lost detail.

I then passed it across from Lightroom to CS2 as tif and applied really (ridiculously) aggressive highlights. Still that detail looks lost to me.


P1930671_DxO Crop Extreme CS2 Highlights from tif
by gardenersassistant, on Flickr
 
At this level of cropping it is obvious what you are seeing but that is not there on the 2048px image on Flickr or not ad extreme.

The solution to this maybe you need to lower flash power and then lift the shadows in post.
 
Don't think so. If the shutter is open at least as long as the flash pulse (and the synchronisation is good) there is no need for the flash to work harder. It would have to work harder if I was using HSS/FP flash, but the FZ200 has a leaf shutter (I believe, although I've never actually found a solid reference to that) and it doesn't have the flash syc issues that a focal plane shutter has, so it can use ordinary, non-HSS/FP flash at much higher shutter speeds than focal plane shutter cameras.

Guess that may be possible. When I use High Speed on my Canon more power is used, till it eventually fails to expose.


Agreed. But the problem I see (although I see that Bryn doesn't), for example in both the annotated images in the previous post, is that the areas on the thorax are not just a bit brighter, they are different in colour (different in a way I don't find appealing). I assume that one (or possibly two) of the channels is getting clipped, so although it isn't blown to featureless white the colours are wrong. Sometimes the colours are wrong (for example I've seen a red poppy petal turn bright yellow when overexposed). Sometimes the images get what I think of as "bleached", where the colours move towards white. Although the hue may still be more or less ok, the saturation gets lower and the luminance higher, and it looks very different from the as seen scene. I get this a lot with petals catching the light at a certain angle.

Yes - I see what you mean on the real pixel peeping versions - certainly losing some colour detail. Though the worst highlight is in the water looking closely.

No, they are all altered. I always post process, so a test for me is a question of what state I can get an image to with post processing. The out of the camera version (and I'm not really sure what that means for raw) doesn't matter to me. A lot of my stuff looks more or less horrible (for example because of under-exposure) before it is processed.

I always shoot raw. I very frequently (much more often than not) pull the highlights down. I suspect this is partly because of the characteristics of the small sensor on the FZ200 than I'm using. I have to do it even though I often underexpose.

I'd like to see it with a more highly reflective subject too. I haven't seen one yet to try it on. Unfortunately, I'm pretty sure I'm not going to like it when I do.

Fine - I find RAW always needs adjustments to bring the best. Though I tend to compare flash diffusion trials on RAW to RAW output. Just me I expect.

At this level of cropping it is obvious what you are seeing but that is not there on the 2048px image on Flickr or not ad extreme.

The solution to this maybe you need to lower flash power and then lift the shadows in post.

Certainly worth experimenting with this to see if a -1 flex comp helps.
 
I think they look good Nick, I like the highlights and think they look well controlled and definitely not bleached, without them the images would look flat.
 
At this level of cropping it is obvious what you are seeing but that is not there on the 2048px image on Flickr or not ad extreme.
The solution to this maybe you need to lower flash power and then lift the shadows in post.
Certainly worth experimenting with this to see if a -1 flex comp helps.

Here we go. A well underexposed (raw) capture - notice the big gap at the top of the histogram.


0745 1 2015_06_29 Example 1 OOC, well underexposed
by gardenersassistant, on Flickr

Raising shadows in Lightroom had virtually no effect, so I sent it across to CS2 as tif and raised the shadows.


0745 2 2015_06_29 Example 1 Shadows lifted
by gardenersassistant, on Flickr

Even at this size the bright area on the thorax looks bad to me. (And pulling down the highlights to try and improve matters has almost zero effect.)

Just to be sure though, here are 100% crops of the area.


0745 3 2015_06_29 Example 1 OOC, well underexposed, 100pc crop
by gardenersassistant, on Flickr


0745 4 2015_06_29 Example 1 Shadows lifted, 100pc crop
by gardenersassistant, on Flickr



I'd like to see it with a more highly reflective subject too. I haven't seen one yet to try it on. Unfortunately, I'm pretty sure I'm not going to like it when I do.

I then found a dark subject with a shiny thorax. Well that's how it looked where I found it, in a pretty shady area. Looking at the image I now see it looks like a fairly ordinary and not particularly dark and not particularly shiny fly. Even so the results were not at all to my liking.

This one is underexposed too - see the gap at the top of the histogram.


0745 5 2015_06_29 Example 2 OOC
by gardenersassistant, on Flickr

Here it is with shadows raised in CS2.


0745 6 2015_06_29 Example 2 Shadows lifted
by gardenersassistant, on Flickr

And the 100% crops. Again, I don't like what I'm seeing here.


0745 7 2015_06_29 Example 2 OOC, 100% crop
by gardenersassistant, on Flickr


0745 8 2015_06_29 Example 2 Shadows lifted, 100% crop
by gardenersassistant, on Flickr

Here too, pulling down the highlights in CS2 had virtually no effect.

This was not as difficult an example as I thought when I was photographing it. I am sure that really shiny beetles will be much worse. Basically, this diffuser, like the others I have tried, doesn't produce results I like with curved reflective surfaces.

Even on the camera LCD I could see the problem remained. I went and photographed some flowers instead in the nice late afternoon light, using natural light with the 70D. That made me feel better.

I have no idea where to go now with diffusion. I think I'll wait for Bryn to explain what he is using and perhaps try that. I do wonder if I'm fighting physics though.
 
Last edited:
I think they look good Nick, I like the highlights and think they look well controlled and definitely not bleached, without them the images would look flat.

Thanks Neil. With the exception of one small area in #7, I don't think the images in the OP are bleached or blown. That's why I chose them.:) But as I've tried to illustrate in the subsequent posts, I have plenty of examples where the highlights are not well controlled and are bleached or blown.
 
Back
Top