A couple of leading questions, focus and zoom with image quality

Messages
2,196
Edit My Images
Yes
I have a Canon 400D (which is EF-S mount, and 1.6x crop), and the following lenses (all have a UV filter for 'protection':
Canon EF 50mm 1.8
Canon EF-S 18-55 /Kit lens/ 3.5-5.6
Canon EF 28-135 3.5-5.6 IS
Those aren't too important for the first of the questions

Canon 70-300 4-5.6 IS, this filter is a HOYA UV(0)

I would like to take scenery shots, and general wildlife. I generally don't take people, but, got the 50 to see whether it was something I would enjoy.

I posted an earlier post wondering whether the focus was out in the camera/lens in general, and was advised that I should try taking photos of paper straight on, at distances and several apertures to determine what the best mix was for the lens. However, another post states that there will always be problems with lenses, so don't be over-critical with test-cards etc.

When I did this, it became clear that beyond around 220mm, the lens became soft. The following set were taken with the aperture set at 10, which seemed to be good for the lens. Bottom left is 170mm, top left 220mm, top right 300mm.
All images were taken under the same circumstances, and as quickly as possible, IS was on, longest shutter speed was 1/500, (shortest 1/640, similar results with IS off).

This bird wasn't the greatest of subjects, but to me, the image is much blurier than I have seen other people's examples, supposedly with the same lens. Image here is 225x225 pixels taken from [basically] the same area of the photo (total sensor 3882x2592, so it is quite a big crop, but with a bird 10" high, it is quite small on the sensor, even at 300mm). The 300mm image is one of the better ones that have been taken, but I am not sure how.


So, question number 1

1a Is this too much blur/soft focus overall, or am I being too critical
1b Is is possible to arrange it so that the bad focus is at the other end of the range? (i.e. by sending to Canon, way out warranty so would need to be paid for). I understand that the lens cannot be perfect all the way through the range.
1c Is this just too much for the lens

I think there is going to be a related question further on, as I think I know what the answer is going to be,
Thanks in advance, Gareth
 
Can't see any pictures?

Did you use a tripod? For testing I would want tripod mounted, IS off and use the self timer to try and utterly eliminate shake.
 
When I confirmed using a piece of printed paper, I did use a tripod, and I had IS off. There was a grew fuzz around the printing, and the 'grain' of the paper could not be seen at the higher focal lengths.
I should also state (which was the reason I was intending on editing the post) was the photos above were shot using style 'neutral', RAW images, +3 applied to sharpness under raw, then when I looked at the RGB settings, I felt they looked a bit better with +45 sharpness.

The photos show here, I used the link generated by the gallery upload thing.
The URL should be: http://www.talkphotography.co.uk/gallery/showphoto.php?photo=57011 (I think)
 
I dont know the lens but it seems to me that you are pixel peeping too much ;-)

I also wonder what you get if you focus on something rather nearer? you can still compare focal length effects looking at something 20 feet away instead of however far away this is (looks like waaay in the distance).

I'd suspect the answer is that you are expecting too much.

I would try again but check what iso, use a tripod, do not use the filter and try looking at something nearer.

I would also say that you should review your use of in camera sharpening - again I'm not familiar with the canon, but I have seen a few tests on the olympus kit that I use which shows that image quality is actually degraded by increasing the sharpness. Maybe someone else can point you to similar info on your camera. or read this http://www.wrotniak.net/photo/43/e510-fs.html for with a view on general applicability. Increasing sharpness can introduce artifacts that you dont want.
 
Thanks for the suggestions. I am using Natural/Neutral in the camera, there is no in-camera adjustment being made on this setting, as far as I know (when I open the file in DPP, it shows 0 against the sharpening figure, the sharpening I added appeared to improve, not degrade the image)
I shall try again without the filter, but previously when I had removed it to test, there was little difference.
Measuring it out, the bird was around 7-8m away, was as close as I could get really (was peering around the corner of the house, using it for a bit of stability). Minimum focusing distance for this particular lens is 1.5m

It is possible that I am just pixel peeping, but I have seen pictures that people of taken, of birds half the size, where you can see the individual parts of the feathers on the birds (half the size!). Admitedly, this bit doesn't show that, but my belief is that none of these shots would be clear enough to allow this.
 
Could we see the full images they're from? There does look to be a fair amount of chromatic aberation going and a lack of colour/contrast.

Things can be clear at the pixel level but you'll need really good lenses.
 
Any chance of getting the original image up for us to look at?
 
Can't open those files. I suspect they're HUGE!

But anyway, it's not going to be a focus issue at f/10. You've got plenty of depth of field. Camera shake is unlikely also, even if subject blur cannot be absolutely ruled out - birds move very fast, but there's no evidence of that.

And while I'm not a UV filter fan, they very rarely affect sharpness; it's flare they induce under rare circumstances (eg bright lights at night).

No, the problem is, I'm sure, obsessive pixel peeping. You're looking for a problem and, unsurprisingly, you've found what you're looking for. But you're being completely unrealistic expecting to get super-fine detail on mega enlargements like that.

If you think about it, if all that was needed for birding was a 300mm consumer zoom and a very tight crop on the PC, then nobody would bother with £6k lenses and sit in a hide for hours. Birds are extremely testing; you need very long, very sharp lenses, a close viewpoint, and at least as much fieldcraft as photographic expertise.

There's nothing wrong with your kit :)
 
... Canon 70-300 4-5.6 IS, this filter is a HOYA UV(0)
I assume it's not one of the early ones affected by the recall? I would lose the filter before drawing any conclusions myself.

I think my copy of this lens is really quite sharp, but this shot was taken at around the 165mm - 200mm mark.

Stroller.
 
@Strolls: That squirrel is lovely. I do not believe it was a recalled lens, I think I have checked before, but will check again. I have had the lens from new, and it was registered, but maybe Canon don't contact people for consumer lenses.

@Hoppy: Initially I wasn't really looking for a problem, I just noticed that none of the images looked great, and wondered why, then started looking closer.

>>There is nothing wrong with your kit
I assume what you mean is, that the lens is working as designed, this lens is not out of spec. Which leads me to what I thought the second set of questions were going to be:

2a If above 200mm, the focus/sharpness is going to drop off, is there a usual way people use to stop themselves going too far into the non-sharp area (rather than continously looking at the distance markings on the outside of the lens)?

2b I have been toying recently about an upgrade. I can only really afford to get either a lens or camera (monetary wise, and wife-wise). I have been thinking that the reach of the -300 is not quite enough, and that I am pushing the lens. This lens cannot use an extender.
I was leaning more towards something like the 50d, as I had found before that the grain of the 400d was a lot for some of the shots I had taken at high(er) ISOs (however, I have since updated the firmware, and the grain has died back a lot, previously a shot like the top left one, would have had so much sky noise it would annoy me). I like landscapes/sunsets which seems to make it worse. However, if I get more serious about landscapes, and move full-frame for the extra non-zoom (50d is still 1.6 crop I believe), I don't want to loose too many lenses. On the other hand, I was thinking about the canon 100-400 IS. I don't feel confident enough at the moment, to make do with primes (although the only one I have tried is the 50). The advantage of the 100-400 is that it could take an extender at a later date if required, however it isn't much longer.
Unfortunately prices are going the wrong way at the moment.
Is there an alternative to the lens? I did think about the Sigma 150-500mm OS, but then worried that as a non-canon make, the chances are that I will have similar issues at the longer focal lengths again. I am also concerned that if I use an extender, then I will loose the ability to auto-focus, or focus confirm from the camera.
 
Get rid of the filter before you do anything!

I almost got rid of a Canon 70-200 f4 because it was as soft as butter. Turned out it was the filter and it was not a cheap one either. Eliminate that before you spend any money.

I found that my lens was actually tack sharp, so check it first, it won't cost anything to do.
 
@Strolls: That squirrel is lovely. I do not believe it was a recalled lens, I think I have checked before, but will check again. I have had the lens from new, and it was registered, but maybe Canon don't contact people for consumer lenses.

@Hoppy: Initially I wasn't really looking for a problem, I just noticed that none of the images looked great, and wondered why, then started looking closer.

>>There is nothing wrong with your kit
I assume what you mean is, that the lens is working as designed, this lens is not out of spec. Which leads me to what I thought the second set of questions were going to be:

2a If above 200mm, the focus/sharpness is going to drop off, is there a usual way people use to stop themselves going too far into the non-sharp area (rather than continously looking at the distance markings on the outside of the lens)?

2b I have been toying recently about an upgrade. I can only really afford to get either a lens or camera (monetary wise, and wife-wise). I have been thinking that the reach of the -300 is not quite enough, and that I am pushing the lens. This lens cannot use an extender.
I was leaning more towards something like the 50d, as I had found before that the grain of the 400d was a lot for some of the shots I had taken at high(er) ISOs (however, I have since updated the firmware, and the grain has died back a lot, previously a shot like the top left one, would have had so much sky noise it would annoy me). I like landscapes/sunsets which seems to make it worse. However, if I get more serious about landscapes, and move full-frame for the extra non-zoom (50d is still 1.6 crop I believe), I don't want to loose too many lenses. On the other hand, I was thinking about the canon 100-400 IS. I don't feel confident enough at the moment, to make do with primes (although the only one I have tried is the 50). The advantage of the 100-400 is that it could take an extender at a later date if required, however it isn't much longer.
Unfortunately prices are going the wrong way at the moment.
Is there an alternative to the lens? I did think about the Sigma 150-500mm OS, but then worried that as a non-canon make, the chances are that I will have similar issues at the longer focal lengths again. I am also concerned that if I use an extender, then I will loose the ability to auto-focus, or focus confirm from the camera.

Interesting post CP :) I think that you are wrestling with the conflicting demands of birding. It is an extremely testing discipline, both on the kit you need, the photographic skill, and the field craft. Everything goes against you, and it is no wonder that the folks that deliver the best results tend to use something like a 1D3, maybe a 600mm f/4, and even an extender on top. That is, a camera with amazing AF even at f/8, fast drive, 1.3x crop for even more reach, and the longest lens possible. I find that combination extremely hard to use, even if I could afford it.

If I was you, the first thing I'd try is getting closer. A few people have a garden hide, but I have found common garden birds take quite well to having a tripod parked near the feeding station, pre-focused and fired by an extended remote cable from a safe distance. You do run the risk of starlings crapping on your camera though ;) So you put it in a plastic bag, which catches the light and makes a noise in the wind which they don't like. So you need to have it set up as an almost permanent feature for familiarity. Then you find that the birds won't sit in your pre-focused spot. The problems don't go away. Edit: set everything on manual - focus, exposure - with mirror lock. This will give you minimal shutter lag - birds react so quickly, they are already moving at the sound of the mirror going up, believe it or not!

I think a Kenko Pro 1.4x extender will work on your 70-300mm IS - it doesn't have the long protrusion of the Canon extenders. But you'll lose AF. More problems LOL. There's not much wrong with the sharpness of your lens at 300mm, and it is certainly a better way forward than shooting at a shorter focal length and enlarging on the PC.

I don't understand some of your comments in 2b re firmware and full frame etc. The sharpness problem you have is because the image you're getting isn't big enough. Moving closer is the easiest way around that. A 400mm lens will help, and 500mm will help a bit more, and so on, but when talking about extenders remember that you lose AF beyond f/5.6 on consumer cameras.

My advice would be try the tripod/remote route, and maybe a Kenko extender. Cheap, relatively easy, and it will give you a feel for what is possible and what further might need to be done.
 
@Strolls: That squirrel is lovely. I do not believe it was a recalled lens, I think I have checked before, but will check again. I have had the lens from new, and it was registered, but maybe Canon don't contact people for consumer lenses.
No idea if they contact you or not. This lens was released in late late 2005 / early 2006 I think, and the problem only affected copies from the initial batch(es). If you bought yours since 2007 then it certainly won't be affected.

http://www.usa.canon.com/consumer/c...ct&keycode=2112&fcategoryid=216&modelid=11922
If you find yours is in the affected range of lenses then you can find the same not on the Canon UK site, phone them up & they'll give you details for returning it for repair.

But sling the filter. I kinda think there's a bit of a swiz being perpetuated on newbies with the "filter for protection" thing. My cousin visited recently & she's bought a new dSLR with an 18-135mm (or so) lens to take photos of her kids; all she wants is 1/2 decent "precious memories" before they grow up, and in her case I think the filter (I caught her she cleaning it by rubbing it with her t-shirt) is justified. But I can't help thinking that camera-shop salesmen made up this "wisdom" just to sell more cameras - perhaps people tend to follow it longer than is justified because it's amongst the first pieces of advice they receive when they first buy an SLR.

Stroller.
 
I think that the filter may indeed by the culprit darn it.
I don't know which hoya filter this is, as I ended up with too many when I bought them, thinking that Hoya was a good brand... but it seems that there are some to steer clear of (I have worked out so far the purple and green are not so great [pity as one of the polarizers I have in 77mm is purple]. What about the grey coloured packs?)

I have taken a series of photos, 16 in total, all at F/10, ISO 200, AV mode. Focal lengths, 70 mm, ~120mm, ~200mm, and 300mm, with continuous shooting set (most of these are the second shot), IS OFF, on a tripod, measured at 3m, which is twice the minumum focal length. They are in the image here (warning ~700k, saved at quality 12, but there was a noticable difference on quality 10). Each one is a 360 px crop of roughly the area of the photo which is in focus. One side is with the filter, one side is without.
Guess which side is which. I think the second set ~120mm is an anomaly.
http://86.153.170.59/filtercompare.jpg
 
I think that the filter may indeed by the culprit darn it.
I don't know which hoya filter this is, as I ended up with too many when I bought them, thinking that Hoya was a good brand... but it seems that there are some to steer clear of (I have worked out so far the purple and green are not so great [pity as one of the polarizers I have in 77mm is purple]. What about the grey coloured packs?)

I have taken a series of photos, 16 in total, all at F/10, ISO 200, AV mode. Focal lengths, 70 mm, ~120mm, ~200mm, and 300mm, with continuous shooting set (most of these are the second shot), IS OFF, on a tripod, measured at 3m, which is twice the minumum focal length. They are in the image here (warning ~700k, saved at quality 12, but there was a noticable difference on quality 10). Each one is a 360 px crop of roughly the area of the photo which is in focus. One side is with the filter, one side is without.
Guess which side is which. I think the second set ~120mm is an anomaly.
http://86.153.170.59/filtercompare.jpg

CP, I'm not a filter fan but unless you have a really nasty one fitted it is unlikely to be causing serious damage to the sharpness. Even cheap filters don't really do that; they cause flare. But chuck them away anyway :D

Honestly mate, I wish you'd stop testing things. I don't buy the new shots you've posted. The right hand side looks like it's got a touch of camera shake, specifically mirror vibration. What's you're shutter speed? The fact that you've not mentioned it suggests that you don't think it's very relevant. A low ISO setting suggests the same. A tripod is never a guarantee of sharp images.

I'm as sure as I can be that once you get a decent frame-filling image and stop blowing up tiny sections you will be very pleasantly surprised and satisfied. I think your starling pics were taken from an area one two-hundredth the size of the frame!!! That is ridiculous :eek: ;)
 
Yes, what I have posted are small snippets, but the size of the photos that are produced are too big for most people to download.
You are right, the shutter speed was lower than I would have wanted had I properly stopped to do the math (1/40th), however it is identical each side, and I did take care when pressing the release. I honestly did just miss it off the post above (it was on Av mode, so the camera took care of it, the other settings I selected). I also forgot mirror lockup.
I will take your advice and start just taking photos again, rather than looking for problems. I think I will keep the filter on, or maybe try and get a better one (have worked out that it must be the purple hoya, as the grey one is still sealed), as most of the time when I use my camera, it will be in a back-pack, where possibly the wife can get to it. When I shoot at home I will remove the filter. I have to admit that I have had the lenses two and a half years, and (touch wood) not scratched them, but I have worn glasses for the last 20-odd years and know how irritating scratches can be. [hmm, what a rip-off, the glass in my glasses, 2 small pieces, cost more than the >10 elements in the lens!]
I'll try and entice the birds to eat where the camera is set up as you suggested. Unfortunately, my wife doesn't agree too much with feeding them, so I think I'll have to go dig up some worms. (or maybe get a non-leathal snail trap!)
Thanks for the help and suggesions

Edit:
My comments about the firmware were that my initial thoughts on the 50d as a next stage on, were that some of the shots I have taken, which I have had to do at 800 or 1600ISO are useless because of the noise (one in particular was of a fish in an aquarium, had large red pixels all over, was F/3.5, 1/6th second, ISO1600, so wasn't much more I could do). With the updated firmware (from 1.0.4 to 1.1.0, I hadn't realised the firmware wasn't actually on the main canon website) the noise handling, admittedly I haven't tried it so extreme as the shot I am thinking of, seems to be a lot better, which reduces my desire for the 50d somewhat. I don't know why the noise should be better, in theory it should be a function of the electronics, not the processor, so it could be just in the mind until I reproduce the exact same conditions (the camera was warm).
 
Back
Top