100% crop rant...

Ok, I just deleted half a response in another thread because I thought better of it. And I know we have all different levels of experience here.. but there is one practice that is just driving me nuts.... okay two. The one that bugs me most is the use a the 100% crop. It really shouldn't used a reference or any indication how a final printed image will look. Let me explain why. When you blow something up to 100% on screen, you are seeing depending on what monitor you are using basically somewhere between a 72 and 82 dpi image (resolution of monitor/max res of screen). The minimum output of most comercial and home printers is blended 240 dpi - and more commonly 300 dpi. So, what that means is the dots on the final image are anywhere between 3 to 4 times smaller then the pixels on screen. Or conversly, the screen pixels are HUGE compared to the size of dots on a printed image. Secondly, the pixels on screen glow - and light bleeds because of it. A printed image is reflective - and does not have any light bleed. So looking at a screen image at 100% magnification is the just about the same looking at a printed image at 400% plus magnification. That is why somethings like CA that shows up at 100% magnification doesn't show up at all in the printed image.

There are good uses for being able to zoom in to images. But the common practice here of posting 100% crops is to see if the lens is sharp enough is silly. Print something. It's like saying my car doesn't accelerate between 120 mph and 140 mph well... there must be something horribly wrong... when you in reality you normally don't or can't drive over 85mph anyways. Worse yet, there are many here who try to make others feel that unless you buy a lens that is pin sharp on 100% crops, you will be suffering is just about as logical as saying that unless your car can sustain 180 mph, well then you won't be able to get to work on time, or that they will be able to get to work earlier then you because they have a car that can go that fast.

I am not against anyone getting "L" lenses, or any other lens what so ever. What people buy really isn't that big of a deal to me. If they want to buy a car that goes 180... I don't care either. But telling people that unless a lens looks sharp at a 100% crop - it must be bad is just silly. I see this a lot when preople review f2.8 lenses. What looks soft at 100% crop may look wonderfully sharp at 11x17 double page spread. There are a lot of other factors like focus speed and propor use of DOF that will impact real world sharpness.

There are times a 100% crop is usefull. I used it to figure out I have a decentered element in my lens. But I first saw the softness in one corner in a print first.

If you really want to see how sharp a lens is... read it's MTF chart. If it doesn't seem to be performing to that standard, send it in to be calibrated.

If you a lens collector and having the sharpest lens in town is your thing... very cool. Nothing wrong with it. But to the average Joe/Jane who is shooting for fun... it really is overkill some of the standards being pushed here. Collecting and shooting are two different things.

Anyway... I am done. Just had to get that off my chest. I hate visiting the lens forum because it just gets my hair standing on end. It is such an emotional place for some. Cheers.

I think you are missing the point. Could you compare two race cars by evaluating their performance running at 65MPH on a highway? They may perform equally under those conditions, and a Honda Accord might also be pretty close. But crank it up to 180MPH, and you separate the men from the boys. Even though people aren't usually going to print at the dpi you are seeing with 100% crops, it does let you see what a lens is really producing... and if you are comparing 100% crop images on the same monitor, it is a good test of what the lens/body is capable of. The sharper things are at 100%, the better they will be blown up or even printed later. Reviewing 100% crop images is just another datapoint in evaluation, not THE datapoint.

Just to add a little to the argument, Canon's 1 series bodies produce the least sharp images out of the camera. The x0D series are sharper. The xx0D bodies are the most sharp of DSLRs. Canon P&S cameras will generally be sharper out of the camera than any of the DSLRs. This is from Canon's own literature. So, if you compared a 100% crop from a 1D2 and a 350D, you might very well see a sharper image at that point from the 350D. After processing, however, the 1D2 image should come out on top.

100% crops used with the same body on the same screen, do give some information about lens comparisons... but you have to use them intelligently and not out of context.

It isn't so much the 100% crop comparisons that bother me as much as the downright s***ty pictures that are used. "Here is a shot of the grass in my front yard at dusk wide open ISO 3200. Did I get a good copy?" Yeah bub. I like to use 100% crops to show something, like the effects of NR or sharpening. The other thing that bugs me is the way the crops are made. There is a upsize comparison in another thread that has a bunch of different upres methods, but the crops aren't the same! Sloppy work! Here is the way they should be done:

Source:

I agree. The pixel peepers here don't print enough. I print about %40 of the pictures I take for my business. They look absolutely insane at 300dpi on a 8x10 or 5x7. Tack, tack sharp. Do they look ok on my 1920x1200 monitor? They look ok.

I think more people here need to print. Printing is the end result of your photography. Looking at them on the monitor is just proofing them or previewing.

There are so many factors as he mentions regarding how a %100 crop will look on everyone else's screen. Print your shots and you'll be a lot happier.

The only 100% crops I like are the ones involving pretty girls and burning gunpowder.

The true test of a good lens and a good shot is a 100% crop. Yes many shots when viewed at 100% could be off focus a bit yet still yield a decent 4xt6 or 5x7 print. The better your image, the bigger you can print and/or crop to print. If all your shots are consistently off at 100% - you need to know that in case your lens has an issue.

curtis... ah... old lever action winchester style rifle... when I was a kid I would have died for one... alas my parents prevailed.

Hellahot - I guess my issues is, and yes I know this a "canon" forum, but so much attention is paid to it now that we can use computers to look at the images under rediculas magnification. For a hundred years photographers produced wonderfull images blown up way more then 5x7 that there was never a need to be able to pixel peep to validate their geer. The proof was in the end product. So what is different now? 10 years ago, did you get a magnifying glass out to examine your 8x10s to make sure your equipement was performing? It just seems the pendulim has swong way to far in the direction of technology, and less on common sense and good photographic skills. I know when my shots aren't as good as they need to be... I don't need to peep at anything. Anyway...

Croasdail wrote in post #2081005
curtis... ah... old lever action winchester style rifle... when I was a kid I would have died for one... alas my parents prevailed.

Hellahot - I guess my issues is, and yes I know this a "canon" forum, but so much attention is paid to it now that we can use computers to look at the images under rediculas magnification. For a hundred years photographers produced wonderfull images blown up way more then 5x7 that there was never a need to be able to pixel peep to validate their geer. The proof was in the end product. So what is different now? 10 years ago, did you get a magnifying glass out to examine your 8x10s to make sure your equipement was performing? It just seems the pendulim has swong way to far in the direction of technology, and less on common sense and good photographic skills. I know when my shots aren't as good as they need to be... I don't need to peep at anything. Anyway...

Did people use a loupe to view their negatives?

For what it's worth, part of my editing process is using the magnifier tool with RSP to check the sharpness of each shot at 100%. That's because minor focusing errors and motion blur can go undetected when viewing the downsized image on screen but still show up in a print - even a 4x6. So this is part of the process of selecting the best image from a group. Prints cost money, so I would rather know that it's sharp before I print.

And through experience, I pretty much know how much softness will be noticeable at certain print sizes. This process also helps me determine how much cropping I can get away with.

If I take two identical shots, one with the 50mm F1.4 and one with the 18-55 kit lens, you won't see a difference (usually) when you look at the entire images downsized on the screen. But at 100% the sharpness of the 50mm will really jump out at you. Whether or not this difference translates into a sharper print depends on the print size and the amount of cropping.

Viewing at 100% has definite value, once you have the experience to realistically interpret what you see.

Croasdail wrote in post #2081005
It just seems the pendulim has swong way to far in the direction of technology, and less on common sense and good photographic skills. I know when my shots aren't as good as they need to be... I don't need to peep at anything. Anyway...

Photo equipment has become a consumer commodity, blame Canon, Nikon et al for it. It sounds silly but photo gear isn't sold to someone looking into getting into photography under the heading "Imagine how good the pictures you take will turn out when you learn how to use this camera and produce great shots!" it is sold under "Amazing low noise levels, unparalleled sharpness, crisp vivid colours and a depth of field as smooth as butter". Well when someone is sold on those characteristics, it only makes sense that they're going to focus on them when they get home and say "let me see at 100% if that noise is really that low, and if the lense is really that sharp like I was told it was".

I couldn't care less personally, my gear is some of the cheaper gear on the forum but it is what it is. I do agree with your point anyway

Pixel peeping has it's limited use and I do mean limited.
Focusing on "sharpness" as somehow the holy grail of end of be all of image is a silly notion and if that's the first thing they look at, then I guess that's their thing.

I say composition and use of light is the most important thing. Distortion and sharpness are minimums that have to be met not to distract from the composition and the tones.

I think folks starting out early get too technical and becomes measurbators. I was a bit like that at the beginning too. Hopefully, like childhood, all will eventually grow up and get over it or just move onto another past time.

Every one knows the best way to evaluate a lens in by how much it costs.

Dan-o wrote in post #2081892
Every one knows the best way to evaluate a lens in by how much it costs.

Precisely. That is why I have the best, because mine were the cheapest.

We have always had 100% crop. Even when I got my first FSLR in 1973. Back then we called it a loupe. Nobody complained about their use at that time - I don't see any reason for someone to complain now.

Rad

Radtech1 wrote in post #2082322
We have always had 100% crop. Even when I got my first FSLR in 1973. Back then we called it a loupe. Nobody complained about their use at that time - I don't see any reason for someone to complain now.

Rad

granted - we did have loops, but unless you were looking at medium sized negatives or greater, you were no where near the same magnification that a 100% is now.... at least as I remember it. You pretty much knew the grain size based on the film you were using... and if you pushed the film, you pretty much knew what it would do. But, yes, in general we did use loops. But on 35 mm negatives, it wasn't really all that usefull. Now if you had quoted me back what I used to do all the time - embarassingly I would project them using my handy kodak projector to rediculas sizes - you would have nailed me. That is a lot closer to this 100 % is now. A 30Ds image at 100% is the equivalent to a 47" print. Not sure how many of you all are printing to that size - I don't very often.

But thanks for the replies and feedback - it is always intersting to learn from you all.