BetterPhoto Q&A
Category: New Answers

Amazing digital sharpness?

The level of detail and sharpness in his pictures is AMAZING! While I don't expect to even approach Ansel's level in my life, I wonder if I even have a chance at getting some of the amazing detail he got with a digital camera today? Specifically my Digital Rebel XT.

Many of Ansel's photos are made with very small apertures (f/45) as with 'Rose and Driftwood' and the exposures are several seconds long.

Now I know that it's film first of all, so there's that, but is part of the key to getting amazingly sharp pictures to use small apertures and long exposures? The technical side of me would think that it would be possible to get greater quality with a slower soaking of light... but maybe that's just not the case.

I know that film was slow in the past and sometimes needed long exposures, but by 1932 (the date of the rose picture) it wasn't needed so much so I was assuming that the f/45 and 5 second exposure enhanced the detail and was indeed why Ansel did that.

Is there any merit to this with today's digital cameras? Might it still apply to film but not digital? Or might it be good either way just with slight differences because of the capturing medium or device?

I'd like any suggestions anyone has on getting the sharpest images one can get on a digital camera these days.

Side note... I did just get a Canon 50mm f/1.8 lens today. While I'm pleased at the speed increase from my default 18-55mm f/3.5-5.6 lens, I've got to say that it's pretty loud and rather slow at focusing. It seems like it's fighting to focus sometimes for a couple seconds before I'll get a confirmation beep. Is that normal for this lens? It was purchased new and can be exchanged if needed.

Thanks again in advance, I know there's some great advice out there.

To love this question, log in above
May 26, 2005

The shots above were taken one after another with a remote from a tripod. The left shot was in P mode and the camera chose 1/6 f/1.8. The shot on the right was taken in Av mode and I chose f/22 (maximum of this lens) and the camera decided a 25sec exposure.

All other settings were the same. I used the incandesent WB setting, no flash, ISO-100. Both images were taken with a Canon 50mm f/1.8 lens. I forced the focus to the center point and after focus was confirmed I switched the lens to manual focus. Both images were focused on the center red dot of the record button.

One thing that I was dissapointed with was the lack of a crisp focus at f/1.8 even though the camera confirmed focus. Is this just the way it is with such a large aperture or is there a quality issue with this lens in particular?

As a side note, I think this finally answers my light question from before about whether a picture would be better/worse depending on the aperture. It seems it definately IS much better with a smaller aperture and slower shutter.

To love this comment, log in above
May 27, 2005

Smaller apertures do increase your sharpness, as well as your DOF, of course.

I don't think the length of the exposure really affects the sharpness. It just goes along with using the smaller aperture. Longer exposures actually offer more chance for vibration that could affect your sharpness.

I've also read that some digital cameras have problems with noise on very long exposures.

To love this comment, log in above
May 27, 2005

Anyway, I've been taking pictures non-stop and I did start leaving it on AF but it keeps moving focus slightly and it throws off my tests when I'm trying to compare the same crop of pixels.

I've found though, that I just can't get an impressive picture at 1.8 so far. 2.0 is ok, but it's not until 2.8 that I start to see things pop out.

I've switched to taking pictures of my carpet, so there's a ton of detail to be seen.

I started taking pictures at one step incriments and from 2.8 to 9 really isn't much different. So now I'm with you where it doesn't make much difference, but in the 1.8 to 2.8 range there is still a big difference in detail 'poping'. Even with an unsharp mask on a 1.8, it's not as good as the 2.8 shot.

is there any concern about the quality of the lens? I don't see many other options for a 50mm f/1.8 so I didn't think that would be an issue. There is the 1.4 for $300 vs mine at $75 though...

To love this comment, log in above
May 27, 2005

I found on dpreview.com several examples of the fact that most lenses go soft at their maximum and minimum apertures. I'm not so worried about it anymore I guess if it's common enough for them to say it's common. I guess the f/1.4 lens would have an advantage in that respect in that it may be soft at f/1.4, but it might clear up at f/1.8.

For my final test I lined up a UPS package label and locked to the middle focus point and focused on the bull's-eye. I switched to using the remote capture software so I could look at the previews full screen as I went. Everything was taken in Av mode and with AF turned on. I mis-spoke before about the WB, it's on tungsten, incandesant isn't an option. Not that it's that big of a deal.

I took 5 shots at each f-stop and used the best of each for this lineup. The box was at a slight angle to the camera, as the camera was on a tripod looking down at the box on the ground so it couldn't be perfectly perpendicular. It turned out good to have that though, because I can see now that what's happening is the focus is moving forward from where it should be.

I found the results very interesting in that at f/1.8 the lens starts soft, then gets a little better at f/2.0, but then gets a lot worse at f/2.2. f/2.5 gets a little better than f/2.2, but it's not good until f/2.8 and not great until f/3.2. f/3.5 is great as well, but then f/4.0 gets a little soft again. Up from there is too near identical to matter, although the further you go the more paper texture detail you start to pick up which might be a good thing or maybe not depending on the situation.

So I found out that it's not a total softness except for f/1.8 and to a lesser degree f/2.2. The other settings are all just slightly front focused it seems. Maybe there's a firmware update that addresses that if it's the camera's fault?

In a real photo shoot, if I were to shoot at one of the f-stops that is off, I might focus on a person's eyes and their nose would be clear and their eyes would be soft. That's not a great thing I don't think. I wonder if there is anyone out there that does this thorough of reviews of all lenses... I'm curious to see how the Sigma 50mm f/2.8 stacks up with this Canon 50mm f/1.8 and the Canon 50mm f/1.4.

Well, I learned a lot with this test, maybe somebody else will learn something too.

Still open to any suggestions anyone might have. Or just some real world input maybe. But at this point, I'm going to avoid shooting at less than f/2.8 unless I absolutely need it.

I'm still curious to know if I can get close to the level of detail I see from Ansel with any digital out there today.

To love this comment, log in above
May 27, 2005

To love this comment, log in above
May 27, 2005

To love this comment, log in above
May 27, 2005

To love this comment, log in above
May 27, 2005

Not only did nothing he said contribute to the subject of this post, but the Ansel downplay was just rude and arrogant. Typical of a person with a lack of there own talent.

Not to mention that I didn't say anything about wanting to shoot color or black & white, yet his comment made it sound like he was saying I should give up trying to make good B&W prints because it's too easy.

As a point of fact, digitally, it's basically the same. With film, sure, there are a ton of differences, but not digitally. Especially since most photographers tell me they shoot color and then desaturate it with one method or another later.

To love this comment, log in above
May 27, 2005

To love this comment, log in above
May 27, 2005

B&W photography is most deffinately NOT easy by any means. In many ways it's quite a bit harder than color because you have to focus on how the contrast and tonal differences will end up without the benifit of color. A tanned face against a white house is fine in color, but it's very hard to set them apart in B&W.

I'm still searching for techniques to edge towards the sharpest and most full of detail I can get with digital. Anyone have a picture to share that was taken with a Digital Rebel or 20D maybe?

To love this comment, log in above
May 30, 2005

I skimmed over some of this post, but read most of your comments etc. What I did notice is that you were doing a DOF test (ie on the carpet) on an object, that really didn't have any DOF. It was flat carpet. Try your DOF test on say a row of trees and see what you get. I was looking at the carpet shots, and some were clearer that others, but I don't think it was the appeture, I think it may have just been slight camera movement etc. Being so close to the carpet and taking the shots, all you have to do is be slightly out at such a large appeture and your pics will be not as sharp as you expected.

To love this comment, log in above
May 30, 2005

Front-to-Back Sharpness:
Not only did Adams use a large format 8x10 field camera for nearly all of his work, the field camera also allows some things that cannot be done with medium format, small format or digital cameras (with some special exceptions). They are called "adjustments" which allow shifting, tilting and "swinging" both the front lens board (the board in front of the bellows that holds the lens) and the back (shallow box behind the bellows that holds the film). It is the tilting and swinging in particular that allows the front-to-back sharpness that he achieved from extremely close to the camera to "infinity." I'm not going to try to describe how this works here, it requires some diagrams. If you want to know the details of how it works, find a library book (Ansel Adams "The Camera" is one of them) that covers large format photography and how to used a field camera's "adjustments."

Color Separation in B/W:
This is done using filters designed for B/W photography. The most common are yellow, orange, red and green. Less common and a bit more specialized is a blue filter. The goal in using them is shifting how specific colors render in gray . . . making them render lighter or darker gray than they would without the filter. A red filter will increase contrast between white cloud and blue sky by making the blue sky very dark while leaving the white clouds relatively unchanged. Yellow and orange have similar effect, just not as dramatic . . . orange being between the effects of yellow and red (use the orange and red with people with caution; they affect skin tones some). Green is used quite often with green foliage . . . to make it lighter. Sometimes also used with photographing people against blue sky background . . . provides a bit more separation between flesh tones and sky tones.

Very Small Lens Apertures:
Ansel used extremely narrow lens apertures. Sharpness overall is affected as aperture is made narrower. It's called "diffraction" limiting as it limits the sharpness possible. It's the effect of light diffracting and spreading some as it passes through a hole. The smaller the diameter of the hole, the more it diffracts (spreads). Because Adams used 8x10 large format, he could get away with f/45 and f/64 apertures. Making prints from 8x10 inch negatives requires very little enlargement (zero for an 8x10 print). The effect of aperture diffraction increases as the enlargement increases. It is a problem with medium format and smaller. You will likely find that your lenses hit diffraction limiting and start to show its effects on sharpness at about f/16. The "sweet spot" for most lenses is about f/8 to f/11 . . . in that magical area of minimal softness from lens aberrations, but before diffraction effects become detectable. Stopping down narrower than about f/11 with small or medium format is a tradeoff that gains greater depth of field while giving up a little overall sharpness.

-- John Lind

To love this comment, log in above
May 31, 2005

Natalie H. - From the beginning I was using a tripod, although I was pressing the shutter at first. Then I started using my remote. Then I quickly switched to using my camera's remote capture software via a USB connection when others started suggesting camera movement. Unless the mirror movement caused a problem, there was no movement.

The carpet shots I did were not supposed to have any DOF, I didn't want them to have any. I wanted a perfectly in focus shot of a flat plane of carpet, but realized that part of what was working against me was the carpet's height. If the autofocus locked on a space between two strands of carpet for example, then the top of the strands would be out of focus as such a large aperture. That's why I moved on to the UPS box.

The UPS box did have DOF, as it was at an angle, and being a flat object I could be sure the focus was on the correct plane for proper focus. I did that on purpose hoping to show a front or back focus if that was what was happening. Indeed it was. As the shots show, I just couldn't get it in focus properly at the smaller apertures. Until f/2.8 I'm just not happy with how my camera/lens is able to focus.

What I found though, is what I've seen on photographic tests done by others as well. That's the tendency of this camera and/or lens to front focus just slightly. By slightly, I'd say it's off by less than an 1/8 inch, but it is off. At such a large aperture as f/1.8 this is noticable.

dpreview.com is one of the places that I saw tests done that match mine in their results. They have a focus chart that they took pictures of and show the same front focus happening with a 300D and a Canon f/1.4 lens. That leads me to believe it's the camera's fault.

The review I'm refering to, in case you'd like to see it... is their review of the 300D. Specifically page 17.

http://www.dpreview.com/reviews/canoneos300d/page17.php

The bottom line is that I just proved for myself what others have said, and that's the fact that lenses are soft at their largest aperture and their smallest. That's part of the reason why the 'sweet spot' mentioned is in the f/8-11 range. By then, you're far enough away from the extremes for the best possible light behaviour.

Thank you all for your input. I know I'm better off for it.

To love this comment, log in above
May 31, 2005