Is there a problem here?

Justiceiro

OK- I keep getting the feeling, while using my 20d, that I am just not getting the sharpness that I used to. I was going over some photos from November, when I first bought it, and they look OK. Now it seems that everything is taking on a "soft" appearance. I have noticed this on all of my lenses; The 18-55mm kit, the 50mm f1.8, and the 35-135 f4 USM. I thought that perhaps the DOF was too tight for things on the horizon, so I tried two shots at different apertures, but they look more or less the same. Here are two photos, shot with the 35-135mm. Photo 7106 has the following exif:
Shutter speed 1/125 f5.6 ISO 100 Focal lnght: 135mm
Photo 7107 has the following exif:
shutter speed: 0"5 f32 ISO 100 Focal length 135mm
Both of these were shot, on a tripod, with mirror lockup, originally RAW. so what is the deal here? Do I have aproblem with the lens, the camera, or is it just not possible for the autofocus to lock onto something that far away?

Justiceiro

OK, here is the one at f32 0"5. They don't look that different to me, and they don't look very sharp. Is it me? Do I just suck at this, or what?

jfriend

They don't look very sharp to me

Justiceiro wrote:

OK, here is the one at f32 0"5. They don't look that different to me, and they don't look very sharp. Is it me? Do I just suck at this, or what?

It's a little hard to compare these two because the 2nd is brighter and has more contrast than the first. True lens sharpness and contrast are hard to distinguish with a low res view like this. In fact, the software sharpening algorithms don't actually sharpen the image - they boost local contrast on edges to make things appear sharper.

Posting links to high res versions would make the comparison easier and perhaps more meaningful. Also, I would do a high DOF test shot at f/16, not f/32. At f/32, you can start to see meaningful sharpness degradation from diffraction. While it gives you more DOF, the objects at the actual focal plane might not be quite as sharp at f/16.

Also, a couple questions:

• What did you actually focus on in this shot?
• Have you tried taking some test shots in manual focus. You can take a sequence of focus-bracketed shots in manual focus mode, turning the focus dial a small amount between each shot. This will help you determine if your lens is not capable of producing a sharp image at all or if some of the shots are sharp, then it's more likely an auto-focus issue.

But ... based on the sample I can see here (I'd rather see equal exposure , original, high-res versions to be more confident), neither of these looks particularly sharp to me if this is the actual non-cropped image.

Justiceiro

jfriend wrote:

It's a little hard to compare these two because the 2nd is brighter and has more contrast than the first. True lens sharpness and contrast are hard to distinguish with a low res view like this. In fact, the software sharpening algorithms don't actually sharpen the image - they boost local contrast on edges to make things appear sharper.

Posting links to high res versions would make the comparison easier and perhaps more meaningful. Also, I would do a high DOF test shot at f/16, not f/32. At f/32, you can start to see meaningful sharpness degradation from diffraction. While it gives you more DOF, the objects at the actual focal plane might not be quite as sharp at f/16.

Also, a couple questions:

• What did you actually focus on in this shot?
• Have you tried taking some test shots in manual focus. You can take a sequence of focus-bracketed shots in manual focus mode, turning the focus dial a small amount between each shot. This will help you determine if your lens is not capable of producing a sharp image at all or if some of the shots are sharp, then it's more likely an auto-focus issue.

But ... based on the sample I can see here (I'd rather see equal exposure , original, high-res versions to be more confident), neither of these looks particularly sharp to me if this is the actual non-cropped image.

So if I wanted to run a meaningful test, should I do bracketed focusing at f16? Is that the best aperture for that kind of test? And would it be better to test with my 50mm prime, or set the 35-135 at a medium zoom? Would it be better to run the sharpness test at f8?

Also, for these shots I set the autofocus for the center point only, instead of full field (centering the AF sensor on the boat). Shoudl I run the AF test with the full field, or center point only?

Some of the "close up" shots that I have recently taken are OK, or at least better than this. What I am wondering is, do I need a longer telephoto to get sharp pictures of distant objects, or should I be able to do this with the 35-135? Assuming, of course, that there isn't a camera issue.

jfriend

Some ideas

Justiceiro wrote:

So if I wanted to run a meaningful test, should I do bracketed focusing at f16? Is that the best aperture for that kind of test? And would it be better to test with my 50mm prime, or set the 35-135 at a medium zoom? Would it be better to run the sharpness test at f8?

Also, for these shots I set the autofocus for the center point only, instead of full field (centering the AF sensor on the boat). Shoudl I run the AF test with the full field, or center point only?

Some of the "close up" shots that I have recently taken are OK, or at least better than this. What I am wondering is, do I need a longer telephoto to get sharp pictures of distant objects, or should I be able to do this with the 35-135? Assuming, of course, that there isn't a camera issue.

If you want to just test out your lens to see if it's capable of producing sharp images at all, then you should run a manual focus test with no auto-focus and using small focus steps to see what the best result you can get is. First run it at f/16. If you get good results there, then run it at your widest aperture. This should show you the best your lens can do. If neither of those is sharp, then you probably have a lens problem (assuming the tests were run appropriately). If both of these are acceptably sharp, then you move on to testing the auto-focus system.

You should not need a telephoto to get sharp-looking landscape photos. You will need a telephoto if you want distant objects to look sharp under high magnification. There just isn't enough resolution (e.g. pixels) to capture a distance object as razor sharp when looked at under high magnification. A distance object should look razor sharp when viewed at a normal magnification. For example, if you took a picture of a sailing ship in bright sunlight and printed it at 8x10, it should look very sharp. If you cropped the image and looked at a small piece of the rigging under magnification on the same photo, it would not look all that sharp because there's just not enough pixels to express fine detail in it because it was such a small part of the overall image.

You should also know that image contrast contributes greatly to our perception of sharpness. A dimly lit image or an image without much lighting variation will often not look very sharp even though it's not really a sharpness quality issue. So, you'd probably be best using medium to high contrast subjects under good lighting conditions for your test photos. That will also let you use a reasonably good shutter speed, even at the small aperture to help rule out camera motion/shake (even if you're on a tripod).

erich6

f/32 is too small an aperture to run your test at. Generally you'll get soft performance from your lenses at this small of an aperture.

Since you're doing outdoor shooting try f/8. This is generally a good enough aperture to get sufficient depth-of-field, particularly if you use your 50 mm prime.

Also, can't tell for sure from these pics but it seems like it may have been windy when you took your shots. Take care of blocking the camera from the wind. Buffetting can cause "camera shake" even when on a tripod.

Try auto focus (just center focus point on a well-defined object should do it) and compare it to manual focus and see what you get.

Erich

ShannonW

Hey,

I too have had issues with focus. I posted this thread a while back and it fine tuned my camera. Things are great now. Everyone was helpful in getting the problem fixed. Hope this helps.

The name of the topic was. ..Help with 20D????

jfriend

An example of how contrast matters to perceived sharpness

jfriend wrote:

You should also know that image contrast contributes greatly to our perception of sharpness. A dimly lit image or an image without much lighting variation will often not look very sharp even though it's not really a sharpness quality issue.

I took the image that you attached to your previous message and enhanced the contrast in CS2. Note how much sharper it appears now. If I had been able to do this on your original instead of this small web size, it would look even better than this.

So, right now, I think I need to see your original to offer any more of an opinion on this one. It may actually be plenty sharp and just need more contrast (either with better lighting or contrast-enhanced in post processing).

Here's the contrast-enhanced image. You may not actually want it this contrasty, but I did this to illustrate the perception impact on sharpness. This looks quite reasonably sharp to me, particularly since I'm only working with a low-res web size.

Justiceiro

Thanks for the help, everybody. I am going to take your suggestions and try another outdoor test, using the focus bracketing on manual.

When I shot this it was not that light (just after sunrise) so perhaps I need more light to bring out contrast (your rework does look better). It was also windy as hell and I have a rather light tripod (that's what happens when you blow all your money on a body, there's nothing left over for accessories.)

Sorry I can't post the original- I need to get set up on smugmug or something to hotlink photos. I can send you the original via e-mail (it's in RAW format, so its rather big). But let me try your suggestions first before you waste a lot of your time tinkering, unless you like tinkering. Let me know.

Thanks for the link shannon, I'll give that a read.


I actually have some tests I ran a week ago under more controlled conditions at closer range- they appeared hsarp enogh, but I am having issues in the field. Check out this example with the same lens:

Justiceiro

and here's a 100% crop. These look OK, so perhaps its the other issues. Then again... Can photos ever be sharp enough? Anyway, I will do some mroe screwing around and get back to you guys. Once again, thanks for the assistance.

jfriend

Try again in better conditions

Justiceiro wrote:

Thanks for the help, everybody. I am going to take your suggestions and try another outdoor test, using the focus bracketing on manual.

Low light, windy and lightweight tripod are, in deed, not the best recipe for sharpness. My guess is that those all contributed here and your camera/lens is probably fine. The pencil box picture looks pretty good. We'll await the results of further tests in better conditions.

Justiceiro

jfriend wrote:

Low light, windy and lightweight tripod are, in deed, not the best recipe for sharpness. My guess is that those all contributed here and your camera/lens is probably fine. The pencil box picture looks pretty good. We'll await the results of further tests in better conditions.

I need a heavier tripod as well. It can be quite difficult to do good photography on the cheap, no?

wxwax

jfriend wrote:

Low light, windy and lightweight tripod are, in deed, not the best recipe for sharpness. My guess is that those all contributed here and your camera/lens is probably fine. The pencil box picture looks pretty good. We'll await the results of further tests in better conditions.

I agree.

I'm surprised your autofocus didn't go hunting at f32 in low light.

wxwax

Justiceiro wrote:

I need a heavier tripod as well. It can be quite difficult to do good photography on the cheap, no?

No doubt about it. Unfortunately. :cry

jfriend

Good trip definitely required

Justiceiro wrote:

I need a heavier tripod as well. It can be quite difficult to do good photography on the cheap, no?

Yep. If you want to read about why it's worth investing in a good tripod, read this article. Thom Hogan basically concludes that sooner or later you're going to end up buying a good, high quality tripod so you might as well just get one from the beginning.

Andy

jfriend wrote:

Yep. If you want to read about why it's worth investing in a good tripod, read this article. Thom Hogan basically concludes that sooner or later you're going to end up buying a good, high quality tripod so you might as well just get one from the beginning.

Before you go to bythom, start right here, on Dgrin:

http://www.dgrin.com/showthread.php?p=65553#post65553

ian408

Just a quick question. When you first bought the camera, you mentioned your
shots looked sharp. With your most recent shots, you mention that you shot
RAW.

Just a reminder that an image shot in jpeg will be processed (and may include
sharpening too) in-camera. Whereas all processing is left to you when you
shoot RAW. I mention this only because the 20d's default is jpeg.

Ian

Sigal

That's amazing
I just had focus problems last weekend while working on the challange photo.

My husband gave me the idea to test while taking a photo of a newspaper.

I'm still not sure what the problems is, I tried 3 different lenses.

I use EOS 300D, I'm just going to take it to Canon, and ask them to clean the camera and check it's focus.

Sigal