I've seen a lot of posts around here suggesting people using crop sensor bodies to go to Full Frame for the IQ. Is this a "leap of faith" kind of thing, or are there any examples out there of the same subject photographed with with the same lens and different body?
I was kicking around the idea of going from a 50D to a 5D Mark II, since everyone raves about the increase in image quality. I wouldn't want this to be a case of "the emperor's new clothes" for me though. "Look at these images, crafted from the finest pixels!" If I can't tell the difference, the upgrade isn't worth it.
So if any of you could show 100% crops of the same subject with crop vs full frame bodies, or could point me to a site that does that, I'd really appreciate it.
I dont know why everyone gets so hung up about 100% crops. Surely the best test of IQ is to see the result printed out for full effect (unless you want to make a poster of the smallest part of your image)
I'd guess it boils down to your creativity and if you're the kind of shooter who likes that ultra wafer thin DOF with say an 85L well no cropped camera on earth is gonna give you that ultimate "shallowness" (is that a real word) that only full frame will.
It's probably a hair splitting difference but it's there for sure.
Ian.Dobinson wrote:
I dont know why everyone gets so hung up about 100% crops. Surely the best test of IQ is to see the result printed out for full effect (unless you want to make a poster of the smallest part of your image)
Exactly, the sharpness differences between my 5D and 40D don't matter to me, a little USM evens that up.
The difference is in the DOF at the same aperature and camera to subject distance.
You'll find that the increase of quality has to do, mostly, with pixel pitch and resolving power. However, you have to bear in mind that if you were to take a picture with a crop body at a certain focal length, then take a picture with a full frame body at the same focal length WITHOUT recomposing, then crop the full frame picture to match the crop picture, the quality will be about the same.
However, the quality difference is not a leap faith. Whenever you increase the size of your format (in this case, CMOS sensor) quality is always increased. The reasons are two fold:
A) Pixel pitch - Larger pixels = less noise. Less noise = greater sharpness. This point becomes moot when the full frame camera's pixel pitch approaches that of a crop body.
B) Glass resolving power. This is the biggest issue. The glass we're using now in lenses is pretty darn good, but the resolving power is still not high enough for the crop sensors to get lots of minute detail at their very small pixel pitch, which is why pixel peeping on crop bodies will yield a bit of smearing, regardless of lens used. As you increase the format, the resolving power of the lens because less important as you don't need the resolution be nearly as fine for the image to be clear. The difference in format, of course, requires you to compose differently in order to achieve the same framing. But if you're projecting a picture of a house on to a small format and then the same frame house on to a larger format, the resolving power of the lens attached to the smaller format will need to be much higher than that of the larger format to achieve the same sharpness.
As an aside, the DoF issue is not due to the larger sensor, but the requirement to recompose to achieve the same framing.
NumberFive wrote:
So if any of you could show 100% crops of the same subject with crop vs full frame bodies, or could point me to a site that does that, I'd really appreciate it.
Thanks!
Here but it doesn't really prove anything though...
This proves nothing though. There is more to it than sharpness. Nobody is forcing you to go to FF, but I don't think you'll find any photographer who believes Crop IQ is better than FF. The images take on a different look imo. It's not necessarily due to the Sensor size, but maybe due to the better sensor technology.
I think you're going to find that crop bodies are perfectly capable of producing very good images.
For anyone who wants the ultimate in sharpness, detail and quality, you're better off getting yourself a medium or large format camera. These cameras will blow away any 35mm size format, digital or otherwise.
The problem with arguing about sharpness on 35mm or smaller formats is that it just feels like splitting hairs after a while. Pixel peeping hurts the eyes (and brain) after a while. Even worse is that it seems that after you get to a specific size for a poster, the sharpness/resolution of an image doesn't even matter. Eyes have very little resolving power of objects that are far away. Thus, if you print out a poster sized image from a crop sensor body, you're not going to stand 4 inches away from it like you would a 6x7. Rather, you'll stand a couple feet away and your eyes will resolve the poster just fine, without noticing the diminished quality. Chances are, a poor quality enlargement will be more due to a poor printer or poor post production.
To get an idea of what I'm referring, take a close look at any billboard out there and note that it's a series of VERY LARGE dots of CMYK.
This issue can't be resolved in a web forum. 100% crops don't do the full image justice, and web-sized images look largely alike between crop and FF. Add to that the photog's. skill - someone who knows what they are doing with a crop can blow away a poorly executed shot on a FF. I wouldn't let a bunch of crops or test images guide your decision.
For me, it came down to this - FF or not, almost every picture I saw here, flickr, wherever, that I really liked, was taken on a FF camera. Add in the technical details back in the day - file size, larger, brighter viewfinder, etc. - and it made sense, for me. But again, very likely that may have been due to the skill of the person taking those pictures.
So I didn't have irrefutable proof in front of me, but it was more than a leap of faith. Having said that, many of the images I took with my crop sensors and EF-S lenses are still amongst my favorites.
I wouldn't sweat it from a purely IQ perspective. Like you said, trust your eyes. Also keep in mind that on this forum, many people pixel-peep as an endeavour unto itself. Not that there is anything wrong with that necessarily, but if you only go my what you see on this forum, you can drive yourself crazy second-guessing your equipment. Just my $.02
The differences between images taken with an APS-C and a full frame sensor are many and mostly subtle. Combined, they may make a compelling difference or be completely irrelevant depending on your style of shooting.
As many have already pointed out, FF offers reduced DoF for the same subject distance, aperture and angle of view. Or same aperture and focal length and framing. Or which ever way you look at it.
Another contributor to the overall look of the image is the way that a FF sensor will render the outer areas of a lens' image circle. This will mostly manifest itself with reduced corner sharpness and often a noticeable amount of vignetting if you are shooting at or close to wide open. Granted, you can simulate this to some extent in post, but different lenses have their own characteristic signatures in this regard. While a crop sensor benefits from the usually better center portion of the image circle, for my taste I find this tends to render images a bit clinical/sterile. Of course that may be just what you need.
This also brings me to another point. While the image corners may be a bit more vulnerable with a FF sensor, comparing a 15MP crop sensor to a 15MP FF sensor, the FF sensor does not require glass with the same resolving power as mentioned in previous posts. Where my 24L was a tad soft on my 40D wide open, it seems tack sharp on my 5D. IMHO a high resolution FF sensor like the one in the 5D2 does not make this a moot point, because you still have a significant advantage over a 21MP crop sensor.
Another obvious difference is angle of view at given focal length. With a FF sensor you start using many of the classic lenses the way they were designed to be used. For me, a 24-105 only really makes sense as a general purpose walk-around lens on full frame. You suddenly have much more compelling choices of lenses in the (ultra)wide to mid tele range.
Then you have the fact that for the same resolution, a FF sensor will have larger photo sites/pixels. This usually transfers to more light gathering capabilities and less noise. Going from a 40D to a 5D, I was quite taken with the IQ improvement. With the 40D I rarely went beyond ISO400 where with the 5D I don't hesitate to use ISO800 in every day shooting situations. Also, the 5D files hold up to post processing much better than anything else I have ever worked with. May not be FF specific, but I'm pretty sure the sensor size gives the necessary advantages.
All in all, depending on what you shoot, a FF sensor may be an advantage and it may be the opposite. But just comparing crops/static shots will not give you the full picture. I guess on top of all the theoretical aspects and the test images, a leap of faith is still required :-)
Wow, thank you guys very much for the responses and picture examples. Those have been most helpful. The DOF in the bike shot especially shows a striking difference.
Ian.Dobinson wrote:
I dont know why everyone gets so hung up about 100% crops. Surely the best test of IQ is to see the result printed out for full effect (unless you want to make a poster of the smallest part of your image)
I think it depends on the photography style. In macro photography, every pixel counts. In other types of photography, those same pixels may or may not count as much as the whole.
joekraft wrote:
This issue can't be resolved in a web forum. 100% crops don't do the full image justice, and web-sized images look largely alike between crop and FF. Add to that the photog's. skill - someone who knows what they are doing with a crop can blow away a poorly executed shot on a FF. I wouldn't let a bunch of crops or test images guide your decision.
For me, it came down to this - FF or not, almost every picture I saw here, flickr, wherever, that I really liked, was taken on a FF camera. Add in the technical details back in the day - file size, larger, brighter viewfinder, etc. - and it made sense, for me. But again, very likely that may have been due to the skill of the person taking those pictures....Show more →
If one were so inclined to cycle through pbase photos using a very expensive body vs. a cheap body, you'll more than likely find a higher percentage of good photos from the expensive body.
However, it would be a case of correlation inferring causation to assume that the expensive body is the cause of better photos. You'll find that frequently the people buying the expensive bodies (and lenses) are the people that are investing their time and effort into the hobby. These types of people are also the type of people that are likely to take it more seriously and consciously try to make better photographs vs. the person who buys an entry-level DSLR and uses it as a glorified point and shoot camera.
In fact, you can take a look at photos created with the EF-S 18-55 3.5-5.6 and while a majority of the photos may look somewhat crummy, many will look very nice and well composed. It doesn't have many of the characteristic traits of nicer zoom lenses (such as a lack of DoF or heightened chromatic aberrations), but you'll find many nice photos nonetheless.
Zara wrote:
As many have already pointed out, FF offers reduced DoF for the same subject distance, aperture and angle of view. Or same aperture and focal length and framing. Or which ever way you look at it.
michael49 wrote:
The difference is in the DOF at the same aperature and camera to subject distance.
That's what I've been saying all along! Didn't I get raked over the coals in another thread because I said that a FF camera will give you less DOF at the same subject magnification?
Seth Tower wrote:
That's what I've been saying all along! Didn't I get raked over the coals in another thread because I said that a FF camera will give you less DOF at the same subject magnification?
Keeping the same aperture and subject distance requires you to change focal length between a full frame and crop frame camera to maintain the same composition. The greater focal length used on a full frame camera gives you a thinner DoF than the smaller focal length used on a crop camera. If you were to maintain the same focal length, aperture and subject distance, the DoF will be exactly the same. The composition would change.
It's why a point and shoot, which uses a very small focal length between 4mm and 20mm will have a very large DoF and very little background blur unless you focus on objects VERY Close to the camera.
But, yes, same focal length and framing will affect DoF due to a necessity to change your distance from the subject.
I'm not sure where you find disagreement. I said DoF is reduced for FF at:
- same subject distance, aperture and angle of view: FF requires a greater focal length at same subject distance to achieve the same AoV - thereby resulting in reduced DoF at corresponding aperture values.
- same aperture and focal length and framing: FF requires reduced subject distance for comparable framing at equivalent focal lengths - thereby resulting in reduced DoF at corresponding aperture values.
Zara wrote:
I'm not sure where you find disagreement. I said DoF is reduced for FF at:
- same subject distance, aperture and angle of view:
FF requires a greater focal length at same subject distance to achieve the same AoV - thereby resulting in reduced DoF at corresponding aperture values.
- same aperture and focal length and framing: FF requires reduced subject distance for comparable framing at equivalent focal lengths - thereby resulting in reduced DoF at corresponding aperture values.
I thought by "angle of view" you were referring to focal length (considering the two are intimately bound).
There is nothing subtle about the difference if you enlarge the images with the same framing both to 30x40...which is the size of the portraits I sell. The difference is clearly obvious--the larger format wins.
Comparison at 100% means nothing. That's like enlarging a 1-in square ceraminc tile and enlarging a 2-in square ceramic tile and comparing them. They're both just square ceramic tiles.
But if you take 100,000 of the 1-in tiles and 200,000 of the 2-inch tiles and build the same mosaic from them, the 200,000 2-inch tiles will give you a mosaic that is both larger and more detailed than you get with 100,000 1-inch tiles.
That's what happens with this comparison. A 5D2 has both more "tiles" and larger tiles, and provides a larger and more detailed image. But you would never see that by looking at the individual tiles.
