I've been reading for too long (I think so) about everyone saying "look at 100% crops of the picture" to judge sharpness and blah blah blah.
If I get a bad or negative idea in my hdea, I cannot get it out, without a jackhammer.
Looking over my shots of my 24-70 that I KNOW is sharp, I keep looking thinking...these 100% crops look like ass.
Same with my 70-200, I did some tests and I'm like, well the lens isn't front focusing (by no means precise tests, but little pop quizzes if you will), but it sure doesn't seem right.
So in reality, what can you expect from 100% crops?
I'm probably thinking too much about it, but I've had some recent missed focuses (don't see it being motion blur at 1/2000 of a second) and some other issues, that make me wonder, have I just been lucky with the photos I get, that are sharp? Hopefully it's just thinking too much, but like I said, once a bad idea is in my stupid head, it doesn't want to come out.
Later I think I'll do some shots from a tripod, and and multiple focal lengths, blah blah.
ChrisDM wrote:
Forget 100% crops, look at images at your largest regular output size. That's the only measure that's relevant to you.
Chris M
www.imagineimagery.com
Strangely enough, i've told myself that about 73 times since I've made the post, and at least 276000 times before that today.
I agree 100%, who cares what it looks like at 20X30, if I'll never print it.
I'm almost 100% positive, I don't have any issues what-so-ever in regards to equipment, however (this has to have come from playing music for too many years and tweaking every little thing), I have to wonder...if there is an issue?
AGeoJO - maybe I'm missing something here, but I can't see how the top image can be a 100% crop, if the bottom image is full frame from a 1Ds MkII - with a 4,992 pixel wide frame.
Robert,
I applied a fair amount of sharpening, not overly but still not a fair comparison. If straight from the camera then we are comparing the strength of the AA filter and of course, the larger sensor size of the FF camera plays a significant role in the apparent difference.
Mike,
I used LR 2 to give me a 1:1 ratio and that's what I cropped the image onto. I exported that file as jpg medium and sharpened for web presentation. I set the quality to about 75% to minimze the jpg artefact but still within the allowable size limit for FM. I took it as a compliment that you didn't buy that it was a 100% crop .
RobertLynn wrote:
Strangely enough, i've told myself that about 73 times since I've made the post, and at least 276000 times before that today.
I agree 100%, who cares what it looks like at 20X30, if I'll never print it.
I'm almost 100% positive, I don't have any issues what-so-ever in regards to equipment, however (this has to have come from playing music for too many years and tweaking every little thing), I have to wonder...if there is an issue?
Hanging out in equipment-based forums, you'll likely become biased by "equipment based" photographers, gearheads etc... For many there's really no practical application to the quest for microscope level sharpness (they don't print 24x36 etc), rather the quest for this ultimate sharpness is simply a driven hobby. There's nothing wrong with that, except these types of discussions deviate too far from the practical. For me, I shoot lots of different stuff. I shoot weddings professionally and primarily use a 24-70 and a 70-200 2.8. I've never examined microscope sharpness, corner sharpness etc on these lenses, because for what I use them for it simply doesn't make a difference. But I also sell large format landscape prints in a local gallery. So for my landscape lenses, I have no use for fast, heavy glass. But I do seek corner sharp, lightweight glass to shoot this work with. So I simply like to point out the utility of considering the application when considering your lenses and their relative performance. And the next time you hear someone complaining about a portrait lens's lack of corner shaprness wide open, think to yourself, "So what?"
Chris, I agree with and appreciate everything you're saying.
However, when viewing photos that other people take, even with the EXACT same gear, I always am like...wtf is the problem here? I've been told, and I believe that my lenses are sharp, however at the same time, the more I look at the results others are getting, and the results I'm getting, I think that I've got an issue.
Tomorrow, when the light is good, I think I'm going to go outside with the tripod, and both lenses. Shoot them both at 2.8, 4, 5.6, 8, 11, and 16, and at 24 28 35 50 70mm on the 24-70, and the same marked ranges on the 70-200 to see what happens.
If I don't, I'm going to drive myself nuts...and if I'm already nuts, I'm going to drive myself the rest of the way to crazy.
100% crops as an evaluation is an absurdity, and here is why:
You are mapping 1 pixel on image to 1 pixel on computer monitor, and that is a HUGE image to evaluate from a mere 18-24" viewing distance!
I just did a test on my home system to illustrate. I drew a line across a feature in the photo, it was physically 27cm long on my monitor at 100%. I stored the photo, then I looked at the full width of the photo vs. the length of the line...it was 23.6% of the total frame width. In other words, the 27cm line at 100% meant that I was viewing effectively a 45" wide photo from 18-24" away, and no one would be doing that! It was the same as viewing a 50x enlargement from my APS-C frame, from 18-24" away.
Gary, what do you think, do my 100% crops seem on par with what they should be? I mean, you've seen the pictures my 24-70 and 70-200 have produced. I'm figuring your 100% crop was your new 1dmk3, so I'm thinking yours would be much better right off of the bat, but I believe you owned a crop first?
Gary, thanks. I understand what you mean about the full file. However, I don't know any site that I can load my full file size on. Photobucket is pretty much a waste.
AGeoJO wrote:
Mike,
I used LR 2 to give me a 1:1 ratio and that's what I cropped the image onto. I exported that file as jpg medium and sharpened for web presentation. I set the quality to about 75% to minimze the jpg artefact but still within the allowable size limit for FM. I took it as a compliment that you didn't buy that it was a 100% crop .
Sorry, I still confused!
The vertical distance from the eye to the centre of the lips is 113 pixels in the "whole image" but 264 pixels in the "100% crop" That's only a 2.3 fold enlargement. If the whole image resized as posted is 530 pixels high, then the crop at 530 pixels high should be a 6.3x enlargement if the "true" size is 3328.
What am I missing - a 100% crop is pixel for pixel the same as the original size, without any resizing applied, just a portion of the original, right?
.