Old 12MP 5D or 60D for 16X24 prints

[lad_lueck]

<p>I've been printing full-bleed 13" x 19" landscapes from my 11 MP 1Ds and a 24-70, usually at f5.6 to f11. I shoot RAW frames, and process in PS ACR.<br>

'Grain-sniffing' the prints from my Epson R1800, they look better than what I got shooting Fuji color film, and optical printing to 8" x 10".<br>

The number of megapixels isn't as important as the amount of magnification you need to get from the on-sensor image to the print.<br>

The *lens* is the limiting factor, and the higher mag needed for the crop camera outweighs the MP's.</p>

[bob_estremera]

<p>Thanks everybody,<br>

John Crowe - yes, it's the traditional enlargement quality from a MF negative from an enlarger that I want to come close to. And I've considered setting up that kind of gear but the cost and workflow would kill me. I convert to black and white and can't develop at home, nor do I want to.<br>

If Mauro is correct, even though I'm not a crazy MP hound, I think there might be something to the 18MP advantage.<br>

And also, I need to look at glass as the other factor. Maybe glass first as someone (John?) suggested. I'm going to go check those recommendations out.<br>

Good shooting to all. I'll post with my final decision and photographic outcomes when the day comes.<br>

Thanks again and good shooting to all,<br>

Bob</p>

[danielleetaylor]

<p><em>A common misconception is that more megapixels equals better resolution regardless of sensor format. </em></p>

<p>That the 60D out resolves the 5D is not a matter of theory or misconception, but of observable fact. That said, you are correct that it's a misconception that more MP are always better regardless of format. I would rather shoot with a 6 MP 10D than a tiny sensor, super noisy 14 MP P&S.</p>

<p><em>If we consider lens resolution in terms of line pairs per millimeter, the larger format can potentially resolve more line pairs per frame width.</em></p>

<p>That would only be an issue if sensors were close to out resolving lenses. They're no where near that point. We're still well within the envelope where you can record a smaller portion of the image circle at higher resolution. Though, as I mentioned earlier, this does result in softer (i.e. lower contrast) details.</p>

<p><em>3. It is common to overestimate the effect of a particular increase in photo sites. For example, some used to imagine that going from, say 8 MP to 10 MP would make a significant difference. It didn't.</em></p>

<p>12 MP to 18 MP is significant. It's 50% more pixels. (Some will cite lpmm or linear resolution along one axis as evidence that it's not significant, but that misses the point that we look at 2D images.)</p>

<p>That said, I'm not sure how important the difference is for architecture. Subject matter is a huge variable. If you ask me "will I see the difference in a wide angle landscape?" I would say yes, absolutely. In a portrait? It won't be as significant and perhaps won't matter at all. In architecture? Depends on what architecture you're shooting I suppose. The side of a glass skyscraper doesn't present a lot of fine detail, for example.</p>

<p><em>With the right lenses, optimal shooting technique, and skillful post-processunfpg and printing, excellent 16 x 24 prints are possible with the current camera.</em></p>

<p>Very good point. I'm thinking lenses might be his weak spot right now.</p>

[mauro_franic]

<p>OP, what focal lens are you planning to use more often and how many prints a year (unique - not copies of the same) you plan to produce?</p>

[david_littleboy__tokyo__ja]

<p>FWIW, when the 5D came out, there were some good tests showing that it was very similar to 645 for larger prints. Here, the 5D2 (with a good prime) gets very close to (but maybe not quite as good as) the Mamiya 7 + Provia 100F scanned on a Nikon 9000.</p>

<p>As has been pointed out way above, the enlargement from APS-C to 16x24 is a very large enlargement factor, and one's lens infelicities are going to be noticeably. To reiterate my math, to get those extra pixels worth of quality from the 60D requires the lenses cough up the same contrast at 100 lp/mm as one's 5D lenses do at 50 lp/mm. Of course, on a 5D2, that goes up to 70 lp/mm or so. The experience here was that after the 5D2 arrived here, I replaced all my lenses. The 17-40 turned into a Zeiss 21/2.8, the 24TSE was replaced with the 24TSE II, and the ultra-cheap 55-200, which had coughed up some decent images on the 5D, was replaced with the 70-200/4.0 IS. I seriously doubt, for example, that the 10-22 on the 60D would be anywhere close to the 17-40 on the 5D2. (It's a flipping good thing that the economy didn't trash my business till after I had upgraded.)</p>

<p>So my _opinion_ here is that 18MP on an APS-C camera is crazy _IF_ one thinks one is going to get 18MP of FF pixels worth of image quality out of it. For example, my cheapest lens here, the 80/2.8 prime for the Mamiya 645Pro, which I bought used for US$150, used with an adapter on the 5D2 outperforms (is noticeably sharper than) the 70-200/4.0 IS on the 5D2. And the 60D would show an even larger difference with those to lenses with its 40% higher pixel density; and the 70-200/4.0 is the best lens you can get. Oops. (That said, with a good lens at f/8, the 60D will produce better images than an APS-C camera with fewer pixels.)</p>

 

[mauro_franic]

<p>It's a Catch 22. There is no replacement for sensor/film size but to the OP's question......</p>

<p>...The 5D's pixel limitation will be a show stopper for 16x20 and larger prints. Regarding tests showing the 5D similar to 645 MF, or the 5DII similar to the 6x7 MF; they are wrong and I would either ignore them or spend one afternoon testing that yourself. </p>

[g dan mitchell]

Having made a number of prints from a 5D, it is simply untrue that " pixel limitation will be a show-stopper" at 16 x 20 (or

24) with this camera.

 

The "similar to" comparisons to FILM medium format don't map perfectly, but in many ways digital format performance can

be roughly equivalent to the next larger film format in film. This is, for example, why many former 4 x 5 f LF film

photographers are now have moved to MF digital, and why we recognize that cropped sensor DSLR images can, in most

ways, compete with those from 35mm film.

 

And just because the previous poster says you should ignore me, you don't necessarily have to. :-)

 

Dan

[mauro_franic]

<p>Dan, a 16x20 print from a 12MP camera is like an 8x10 print from a 3MP camera. While it may indeed look good enough, it is (to me) not display quality.</p>

<p>Would you advice the OP to get a 12MP camera for his purpose to produce 16x20 prints? I do not.</p>

[scott_ferris]

<p>Did nobody look at the link provided by Miles? It answers the specific question perfectly. The 5D MkI prints out to 20x30 demonstrably better than 50D ones, the 50D and 60D are effectively identical. That is the answer to the OP.</p>

<p>Now my first reply was to ask what FOV was needed, and it turns out that a lens upgrade is in order too, but, that does not detract from the fact that in that very relevant and specific criteria of that one test, the 5D prints big better. Any lens upgrades should be done with an eventual move to FF in mind.</p>

<p>Given Bob's specific interests and lens use the 90 TS-E would be my first purchase, a used 5D my second, a 70-200 f4 my third, but I would make a plan for all three. For out and out architectural detail shots the 90mm TS-E is far and away the most useful single upgrade.</p>

[thierry nguyen cuu - nomad]

Hey Bob,

 

You can't go wrong with the 5D.

12MP would be more than enough for most applications.

[bob_estremera]

<p>I should address some of the questions from some of the posts here.<br />My style of architectural shooting, while it often involves the 'glass and steel monoliths of NY, mostly centers around the classic style. Much is in masonry, stone and brick. Lots of little details and texture. I would say the demands are not dissimilar to the rendering of leaves, foliage, rock formations or similar landscape criteria.<br>

I know that when I can afford a used 5D MKII all of this will be mute.<br>

I was out shooting just today and found myself most frequently at the long end of my EF-S 55-250.<br />Imagine gargoyles on the 10th floor and wanting to capture the snarling mouth framed by brick and mortar.<br>

Mauro, I 'imagine' I would want to print a variety of individual and unique versions, perhaps a few hundred a year. Much of that might be several prints of the same image. I would be using the services of some high end inkjet printers, specialty black and white inks on quality rag stock.<br>

I also did take a photozone.de look at the recommended 70-200 lenses. I can see the need for something of that sort because I loath the idea of too much sharpening. In the digital world, I seem to notice that sharpening often just creates 'edge contrasts' without actual detail as in enlarger prints from MF or the true resolving of details with fine glass.<br>

I will take a look at the links as mentioned and see what additional information I can glean from them.<br>

As always, I am very appreciative of all your feedback and help with this.<br>

Thanks, Bob</p>

 

[danielleetaylor]

<p><em>As has been pointed out way above, the enlargement from APS-C to 16x24 is a very large enlargement factor, and one's lens infelicities are going to be noticeably. To reiterate my math, to get those extra pixels worth of quality from the 60D requires the lenses cough up the same contrast at 100 lp/mm as one's 5D lenses do at 50 lp/mm.</em></p>

<p>You are confusing contrast and resolution. You are also using a contrast of 50% MTF as if it were some form of hard limit on resolution. The ability to resolve detail doesn't start to disappear until around 10% MTF.</p>

<p>Further, we are not stuck with the MTF we are given from any particular lens/sensor/film combination. The whole point of sharpening tools is to alter detail contrast. If the gap were very wide between the detail contrast of a FF sensor and of a crop sensor, then the FF sensor would have a distinct advantage. In practice the gap is narrow enough that it can be closed just by using different USM settings.</p>

<p><em>I seriously doubt, for example, that the 10-22 on the 60D would be anywhere close to the 17-40 on the 5D2. (It's a flipping good thing that the economy didn't trash my business till after I had upgraded.)</em></p>

<p>From what I've seen those combinations are pretty much equal. I have more experience comparing the 17-40L and Tokina 11-16 f/2.8, and the crop combination beats the FF combination in that case.</p>

<p><em>So my _opinion_ here is that 18MP on an APS-C camera is crazy _IF_ one thinks one is going to get 18MP of FF pixels worth of image quality out of it.</em></p>

<p>There are numerous tests, image samples, and reviews which contradict your opinion. Have you truly seen none of them? And why would you form an opinion before seeking them out and reviewing them?</p>

[danielleetaylor]

 

<p><em>Did nobody look at the link provided by Miles? It answers the specific question perfectly. The 5D MkI prints out to 20x30 demonstrably better than 50D ones, the 50D and 60D are effectively identical. That is the answer to the OP.</em></p>

<p>I played with those samples when they first came out. In the Wyofoto test the same sharpening settings were used for all samples, and they were fairly weak settings. You can easily make the 50D sample look better in print than the 5D or 5D2 samples with optimum processing (LCE and stronger USM). If you apply optimum processing to all samples there's not a lot to differentiate between them in this case. The test target is more demanding than some subject matter, but not as demanding as a typical landscape, and it fails to differentiate well between the three bodies. (Again, when proper processing is applied.)</p>

<p>Also, the 50D is not equal to the 60D. Differences at base ISO may not be large, but they're there.</p>

<p><em>Now my first reply was to ask what FOV was needed, and it turns out that a lens upgrade is in order too, but, that does not detract from the fact that in that very relevant and specific criteria of that one test, the 5D prints big better.</em></p>

<p>Sorry Scott, the 5D is not going to make a better large print than a 60D where the lenses are good, the processing is optimal, and the subject is challenging. Resolution is resolution and that 50% jump in pixels makes a difference when you're trying to reach 24" and 30" from small format.</p>

<p>Mauro can testify that I've pushed digital files about as hard as they can be pushed when he and I have debated how small format digital compares to 35mm film and MF film at various print sizes. While we disagree slightly in that I think a 5D can make a good 16x20 print, I think he would agree with me when I say there's no getting around having more pixels when you print larger. The 18 MP sensor will simply present more fine detail when the subject matter offers it, and will enlarge better as less interpolation will be needed to reach the target size. So if the criteria is the best possible 24" print, and Bob won't be buying T/S lenses, then the answer is pretty clear. (If he will be buying T/S lenses then that throws a wrench into things because they are very useful for architecture, depending on his style.)</p>

 

[g dan mitchell]

Mauro, do you do a lot of printing from the cameras you speak about? Or is this mostly theoretical on your part? I suspect the latter, given your odd notions about the potential qulality of a print from a 12 MP full frame body. I suppose that you are welcome to your opinion, but it is most certainly a minority opinion.

 

To those who like to think that cropped sensor bodies are equal to or better than full frame bodies in terms of resolution, you have to ignore a lot of basic facts about photography to come to that conclusion. If you assumptions about a larger format having tpno resolution advantages were true, we would not see any advantage in moving from, say, 35mm film to medium format film.

 

It is fair to point out that many who are not careful and skillful shooters and processors and who do not make large prints may find no real advantages in the larger formats for their work. It is also useful to point out that for certain types of shooting that among cameras at a given price point the features of certain cropped sensor bodies may br more compelling to some photographers than those of certain full frame bodies. But,in fact, better resolution capabilities are one of the most obvious and significant advantages that come from larger sensor/film formats.

 

Dan

[scott_ferris]

<p>Yes Daniel,</p>

<p>And I believe I have surprised both <a href="../film-and-processing-forum/00XJiR">you and Mauro </a>on the capabilities of FF digital capture. The real problem with both your opinions is that neither of you own a FF digital of any generation, many of the posters in this thread do. Endlessly regurgitating tests is one thing, owning and using the cameras in question, and making prints from them, is something else.</p>

<p>In this thread, specifically, I'd trust the opinions of G Dan Mitchell, Dick Arnold and Puppy Face. I am just too argumentative to take seriously most of the time..........</p>

[danielleetaylor]

<p><em>To those who like to think that cropped sensor bodies are equal to or better than full frame bodies in terms of resolution, you have to ignore a lot of basic facts about photography to come to that conclusion. </em></p>

<p>To claim that the 5D out resolves the 60D one would have to ignore both basic mathematics and direct observation. No photographic fact contradicts the statement that the 60D out resolves the 5D.<br>

<strong></strong></p>

<p>The DP Review measured resolution of the 5D sensor in JPEG (h/v) is 2300/2000 absolute, 2500/2500 extinction. While they have not tested the 60D yet, the 7D uses the same sensor. 7D JPEG is 2500/2450 absolute, 3100/3050 extinction. 7D RAW is 2500/2500 absolute, 3800/3900 extinction.</p>

<p><strong>Observation always trumps theory.</strong> It cannot be the other way. But basic mathematics could tell you all you need to know before observation as 18 MP > 12 MP. Since both sensors operate in a density range that is well within the resolution of good lenses, and since the 60D has noise and tonal characteristics equal to or better than the 5D, there's no reason to question that 18 MP > 12 MP.</p>

<p><em>If you assumptions about a larger format having tpno resolution advantages were true, we would not see any advantage in moving from, say, 35mm film to medium format film.</em></p>

<p>False analogy. If you use the same film in both bodies then the film resolution per unit of area remains the same, but the MF body uses a larger area. The resolution per unit of area is not the same between the 5D and the 60D. The 60D is much higher in terms of resolution per unit of area, enough so that it offsets the larger area used by the 5D and results in greater total resolution for the image.</p>

[danielleetaylor]

<p><em>And I believe I have surprised both you and Mauro on the capabilities of FF digital capture.</em></p>

<p>I was not surprised by any of your samples, nor are they relevant to the discussion at hand.<em> </em></p>

<p><em>The real problem with both your opinions is that neither of you own a FF digital of any generation,</em></p>

<p>Please refrain from stating things which a) you do not know, and/or b) would be irrelevant to the discussion at hand.</p>

<p>FYI, the statement that the 60D out resolves the 5D is not an opinion, it is an observable, verifiable fact. I can get an ISO 12233 chart and reproduce the test and results published at DP Review. So can Mauro. So could you. Neither opinion nor theory can trump these simple, reproducible observations. We can debate their relevance in the context of Bob's needs...the lenses he owns and subjects he shoots...but the 60D out resolves the 5D. This is established fact. <em> </em></p>

<p><em>Endlessly regurgitating tests is one thing, owning and using the cameras in question, and making prints from them, is something else.</em></p>

<p>I grow tired of telling you that I have tested FF and crop bodies and that my casual results match the reproducible, verifiable, scientific tests published online. <em> </em></p>

<p><em>In this thread, specifically, I'd trust the opinions of G Dan Mitchell, Dick Arnold and Puppy Face. I am just too argumentative to take seriously most of the time..........</em></p>

<p>If they claim the sun orbits the Earth shall we believe them, or believe scientific observation and tests?</p>

[markonestudios]

I think in this case the greatest mileage would be gained by upgrading the glass rather than the body. As has been mentioned earlier, and from my own 400D experience, with careful technique, excellent 16x24s can be produced using that camera.

<p>Since you are finding yourself on the long end of your 55-250, a 100-400L would get you up close and personal with your intended detail shots, comfortably and with <a href='http://www.photozone.de/canon-eos/204-canon-ef-100-400mm-f45-56-usm-l-is-test-report--review'>very good image quality</a>

[StuartMoxham]

<p>Wow the 5D crop really does look better than the 50D in the wyofoto link Miles provided us with. I would guess that the 50D could be sharpened up a bit to make it look as good but I think I would prefer the 5D over the 50D. Its just got to be easier to work with the 5D files.</p>

[bob_estremera]

<p>Here's a question about lens resolution:<br>

I use the tests at photozone.de as a starting point for many of my lens decisions.<br>

As Mark points out, a 100-400 would be a good range for me. But when I compare the sharpness at comparable focal lengths, by the numbers, that lens and my lowly EF-S 55-250 are equivalent. Same goes for my second generation kit 18-55 IS. It also tests very high on photozone.de. On the other hand, on photozone, the 70-200 lens recommended earlier in this thread, is an obvious improvement and worthy of consideration. Am I missing something with the 100-400? Is this an example of the test data not reflecting the real world resolution? For $1600, I would want something that I'm not searching, with squinted eyes, for subtle differences in resolution. And in a real world test, if the charts resolutions are equivalent, would I, or most of you, be able to tell, with certainty, which 16X24 print was made from which lens?<br>

Thanks still, Bob</p>

[scott_murphy_photography]

<p>I have a Nikon D700 and have no problem at all producing non-pixellated, sharp as a tack, 16 x 20 prints from the 12.1 MP sensor. They key is keeping the ISO low, proper exposure and sharp focus. Are you not experiencing this with your current camera?</p>

[mauro_franic]

<p>Dan,</p>

<p>"Mauro, do you do a lot of printing from the cameras you speak about? Or is this mostly theoretical on your part? I suspect the latter"</p>

<p>As you may know I routinely print from MF 6x7 at 24x30 and 16x20 on my Epson 7880 on either Epson Ultra Smooth paper or Ilford Fibre Gold. I have stopped printing large from my own work using DSLRs and most of the 40D, 7D, 5D and 5D2 prints I do now are commissioned. Based on my experience, I strongly discourage people to print larger than 11x14 from a 5D or lower MP camera (and I have turn down several print requests).</p>

[g dan mitchell]

<blockquote>

<p><em>"As you may know I routinely print from MF 6x7 at 24x30 and 16x20 on my Epson 7880 on either Epson Ultra Smooth paper or Ilford Fibre Gold. I have stopped printing large from my own work using DSLRs and most of the 40D, 7D, 5D and 5D2 prints I do now are commissioned. Based on my experience, I strongly discourage people to print larger than 11x14 from a 5D or lower MP camera (and I have turn down several print requests)."</em></p>

</blockquote>

<p>Mauro, no, I don't know what you print or how - that is why I asked the question. That and the fact that your points remind of those made by others I know whose experience and skill with digital processes does not match their experience and skill with older film processes - and who blame the processes rather than their abilities. </p>

<p>I do my printing from mostly 5D and 5D2 images on my 7900, generally to Ilford Galerie GFS. Most who like my work and see my prints remark on the high quality of the prints at the sizes we are considering here.</p>

<p>You are doing a tremendous disservice to your customers if you "strongly discourage" them from printing larger than 11 x 14 from high quality 5D originals. If you cannot print larger than that from excellent 5D captures, there is likely something quite wrong with your post processing workflow - perhaps it relates to your output sharpening methods? Perhaps you are a film person who is expert at film techniques but has not quite mastered the processes of photographing and printing using digital technologies? Such people often make somewhat uninformed claims about the limitations of digital photography and printing that are based more on the fact that they have not brought their skills with digital to the same levels as their skills with optical/film media and techniques.</p>

<p>Some of your clients, if they are reading this, may want to look into printers who understand how to produce the highest quality prints from the 5D and, likely, other DSLRs as well that they use. There are many expert and highly-regarded photographers/printers who would disagree emphatically with your point of view - and they can back up their positions with their work.</p>

<p>A preference for film is one thing. I know other excellent photographers who still prefer to work with film for a wide range of reasons, and who do outstanding work that way. A few of them even keep the entire workflow in the optical/chemical mode, though many others scan before post-processing and printing digitally. But a few let their personal preference for the older media overcome common sense, and they make blanket and wrong statements about the medium with which they are less familiar.</p>

<p>Dan</p>

<p> </p>

[clintdunn]

<blockquote>

<p>Based on my experience, I strongly discourage people to print larger than 11x14 from a 5D or lower MP camera (and I have turn down several print requests).</p>

 

</blockquote>

<p>...biggest joke of the year. If you can't get world class prints bigger than 11x17 with a 5D classic than you are either A) incompetent, or B)working with lousy files produced from poor technique. I have a 20x30 print from my old 5D that looks unbelievable but I used remote shutter release, tripod, mirror lock-up, low ISO, stopped the lens down and used a professional printing service.</p>

[StuartMoxham]

<p>G Dan when you make a 24x30 inch print from a 12 mp file how close can you usually view the print. If view the print from about reading distance to look at the tiny details how well do they hold up and is it possible to see tiny details?</p>