Sensor size vs MP res

<p>In terms of print quality alone in a 16x24 print how would a Canon 5DII (21.1MP, full-frame 24x36mm) compare to a Sony nex-7 (24.7MP APS-C 23.5 x 15.6 mm) or a Canon EOS M (18.5MP APS-C)?</p>

<p>Thanks.</p>

<p>Assuming equal quality lenses (quite a stretch, for a bunch of reasons) the system with the larger number of photosites and a larger sensor will always have the potential to be enlarged more with equivalent print quality. But...</p>

 

<ul>

<li>Differences in the camera pixel dimensions must be quite large before the resulting image difference is significant. For example, looking at, say, 21MP and 24MP full frame sources images you would almost certainly be unable to see a difference. (Do the math to calculate the difference in resolution at a given print size to see what I mean. Or calculate the relative print dimensions at a constant resolution.)</li>

<li>The larger sensor - with equivalent lenses - can produce greater system resolution with lenses that have the same lp/mm resolution. This is one of the incontrovertible facts about sensor/film formats. </li>

<li>However, these differences are largely academic unless you are regularly producing very large and very high quality fine art prints. If you don't regularly print larger than, say, 13" x 19" or if you almost always share your photographs electronic in email or on the web... most of this doesn't make a bit of visible difference when it comes to image quality.</li>

<li>There are potentially some other functional differences that may or may not make a difference to you. These include a wide range of factors such as camera/lens size and bulk, cost, number of useful apertures, issues related to DOF, and more. Do not ignore these important things and fixate only on something called image quality.</li>

</ul>

<p>Dan</p>

<p>G. Dan +1</p>

<p>I'm not sure what you are saying about lens quality. One of the really cool things about the NEX (and the other mirrorless cameras) is their ability to use just about any kind of lenses. I'll put my Leica 35/2 ASPH or my Contax 21/2.8 up against any lens Canon wants to submit. In fact the lenses are likely to be BETTER on the NEX. There might be fringing issues with some lenses (though I don't seem to see it) but the quality is spectacular. Still it would be nice if the NEX-7 had a full frame sensor and I think that matters.</p>

<p>I once tested the 5D against the Sony A350. Theoretically the APS-C A350 had a higher resolution sensor, but in practice with equivalent lenses the 5D was better. The Canon 50/1.4 is probably not as good as the Minolta 50/2.8 macro but it still seemed to have better resolution even with the lower resolution sensor, though the A350 with the 50mm lens beat the heck out of the 5D with it's 24-105/4L lens. So you're right, resolution isn't everything.</p>

<p>A smaller format lens with a shorter register (distance from rear element to focal plane) and a smaller image circle is going to deliver more lines/mm than a larger format lens. So if you're comparing a good EF consumer lens on a FF camera with a good EF-S (consumer) lens on a crop camera, with both cameras being of similar generation and MP count, then both should yield similar resolutions of photos.</p>

<p>As Dan points out in his last bullet point, though, there are other differences -- DoF, noise, dynamic range, high-ISO capabilities -- that may be relevant.</p>

<p>I usually find that image quality is mostly determined by how good a photographer you are.</p>

<p>I agree with Bob, and I think technical issues are often not relevant when the images are strong.</p>

<p>My current show is going up for the fourth time in 18 months in October, it's 20x30 prints from 11.3MP. Nobody has yet complained about the quality of the prints, in fact, the only problem was someone who was scared by one of them. It's all drawn from <a href="http://spirer.com/npg2/">this online gallery.</a> </p>

<p>A friend who used to shoot with at 5Dii now shoots with an iPhone. His last show, which was reviewed very positively in a big city newspaper, was 20x20 prints from his phone. People respond to images, not test charts.</p>

<p>A friend lent me his NEX-7 for a weekend. I loved it - not enough to run out and buy one, but it's a very nice camera.</p>

<p>To me, the most noticeable difference was this: the NEX files contained considerably more noise at ISO values over 200 than the 5D2. That makes sense given the pixel size. I'm sure that software noise reduction would have worked well on the NEX files. I had some success taming the noise using Lightroom's native noise reduction.</p>

<p>The second thing that I noticed on the NEX is that the auto focus is slower and far less reliable. For the people who complain about the 5D2's AF, I would hate to see them with a NEX in their hands. I would guess that about 20 percent of my AF shots on the NEX were back-focused, and I was avoiding challenging focusing situations.</p>

<p>I would consider a NEX-7 for serious projects if I needed a camera with its feature set and its limitations wouldn't cost me shots. (I would NOT do serious work with an iPhone no matter who has had success with one.) But let's be realistic. A DSLR it's not. It's not going to be as fast or as controllable. Yes, you can attach a lot of lenses, but it's not the same as a DSLR with fast glass and integrated autofocus and flash control. Different tools for different uses.</p>

<p>The NEX-7 has an excellent sensor and a lot of nice features. It's a brilliant tool for travel photography or other applications where you need to keep your gear light (climbing, perhaps?). But a DSLR, regardless of sensor size, is going to provide options that the NEX can't handle.</p>

<p>Generally NEX-7 should not be compared to a DSLR.<br /> However, how else would you expose some of the deficiencies, or shortcomings of the NEX-7?</p>

<p>Very good opinion, Dan!</p>

<p>There are other things that NEX-7 misses or has them wrong, but... let's do not spoil the moment.</p>

<p>In addition to pixel count, there is somethimg like pixel quality,... that remains largely unqualified.</p>