7D vs. 5D Mark II...I know it's been done to death...but...

[picturesque]

<p>Zvia--you will find that modern conversion programs such as Lightroom are much improved re handling overexposure. So when I am minding my exposure, to avoid underexposure, I don't mind overexposing by up to a stop. The conversion programs can recover very well these days, and what really makes noise look bad is having to pull the exposure up on an underexposed image. So shoot to the right, and maybe then some. Try it.</p>

[philip_wilson]

<p>Daniel - the only reason I ask is that a comparison means testing both items - not just one. By the way I find that with the 7D I try and shoot at ISO 800 or below. ISO 1600 is quite usable with care and with very careful post processing you can get acceptable shots (indoor sports) at ISO 3200. On the 5DII I find that the images are good up to ISO 1600 and very usable at ISO 3200. In my experience at higher ISO the 5DII is 1-2 stops better than the 7D.<br>

On the subject of objective tests Popular Photography measured the 5DII resolution at 2820 lines vs 2610 for the 7D (both at 100 ISO). The 5DII had better colour accuracy in their tests (6.3 vs 7 delta E - lower being better). As for noise at ISO 50 the 5DII recorded 0.7% and 0.9% at ISO 100 the 7D was 1.3% at 100 ISO. In these tests the 5DII did not reach 1.3% until ISO 1600 and ISO 3200 (both 1.3%). At ISO 800 the 7D recorded 2.1% about the same as the 5DII at ISO 6400 (which recorded 2%). While Popular Photography does not have a 125 year history like Amateur photographer it has been around since 1937 and is the top selling photography magazine in the world. Interestingly their test equipment (like almost everyone else's) appears top confirm my own findings from owning and using booth bodies.</p>

[philip_wilson]

<p>Zvia - on a serious note I find DPP does the best job with 7D images at high ISO - especially if the exposure is slightly off (1/3 or 2/3 of a stop). I find Adobe CS5 works fine with lower ISO images from the 7D. What I really notice is that the 5DII can tolerate more exposure errors at high ISO than the 7D. I am not sure why this is but you can generally get away with a RAW file that is 1 stop underexposed on the 5DII but with the 7D this does not work so well.</p>

[dcstep]

<p>Don't worry Scott, I've found some of your "work."</p>

[scott_ferris]

David,

 

I doubt it, there are two Scott Ferris' here on photo.net, I am the one with capitals and no images. There are four

photographer Scott Ferris' that I know of. I know this because another two of them wanted some of my domain names.

Tomorrow I will post a couple of images though. One will be a surfer, for Daniels amusement, the other will be an image I

know the 7D could not print out as well as the 1Ds MkIII/5D MkII.

[keith reeder]

<blockquote>

<p>Keith put up one image that I embarrassed by offering a fairly similar image of much higher quality</p>

 

</blockquote>

<p>Heh! You did <em>what</em>?</p>

<p>Please! It's getting silly now.</p>

[dcstep]

<p ><a name="00Y2QO"></a><a href="../photodb/user?user_id=657840">Scott Ferris</a> <a href="../member-status-icons"><img title="Frequent poster" src="../v3graphics/member-status-icons/1roll.gif" alt="" /></a>, Jan 16, 2011; 12:47 a.m. said:</p>

 

<blockquote>

<p>David,<br>

I doubt it, there are two Scott Ferris' here on photo.net, I am the one with capitals and no images</p>

</blockquote>

<p>What makes you think that I wouldn't know that?</p>

 

[scott_ferris]

<p>Sorry Keith, I did not embarrass you, this is not a personal attack, I apologise if you felt it was. But, if you don't see that my horse image is substantially better image quality than your lion then we are talking about different things. I really would like to illuminate this discussion.</p>

<p>To reiterate my opinion, I believe the 7D and 5D MkII cameras are both superb imaging tools, both are far more capable than most of us. There are very good reasons for getting either one over the other. One of the reasons for getting the 5D MkII over the 7D is because the image quality is better. I think we are all agreed on those points. Where we start to disagree is on where the image quality differences start to show. Although I do not own both, I have used the 7D a good bit and will be using one again next month, I believe (along with 75% of the owners of both that expressed their opinions in this thread and Zvia after using both), that the image quality differences are noticeable sooner than you, Dave and Daniel do.</p>

<p>Dave and Daniels only support for their positions seem to be downloaded bench test images, I have pointed out why that is a weak basis for a real world comparison, test images do no reflect real world ones very well. As an example of this difference look at the IQ from your lion image and David's very nice real world raptor image that is totally destroyed by noise. I would say his image is closer to a real world image, but I can understand if people who predominantly shoot soft toys in controlled (all be it dim) lighting conditions disagree. Both make effortless studio cameras, the studio gives you the most test like situation, you are controlling the light, contrast etc 100%. A dark bird against a lighter sky in miserable, low light and low contrast lighting conditions really does change things up, the 7D is supposed to be a better birding camera, but in David's image there is no doubt the 5D MKII would have made a better image.</p>

<p>Processing is very important, you can't process both cameras files the same, but in a comparison you can't just change one to try to level them, you must process both to get as good as you can. That is what we do to all our files all the time, some need more work than others. 7D files generally need more work than 5D MkII files, for some, that alone, is a decision maker.</p>

<p>Both on the screen, and in prints, I can see differences between the different cameras files in lots of situations. Not every time for every subject though, I fully agree with the LL article about the G10, a camera I also own. BUT, I know when I can take an image that will be noticeably different between the 5D MkII and the 7D or the G10 and when the scene presents an opportunity to make a very very similar file with any of them, it is not rocket science, just basic imaging knowledge.</p>

<p>In the same way that I have a G10 and use it instead of the 1Ds MkIII on occasions, I tried the 7D to see if it would compliment my bag. After using it for a month I concluded that it didn't offer enough to me to purchase one. My biggest issue was that the crop factor "advantage", in real world use, did not transfer from the small advantage the 7D has in test images, and I was disappointed in the number of times, on screen and in print, that the image quality was not up to what I was used to with the 1Ds MkIII.</p>

<p>Here is an image that I know from experience would not print out as well had I taken it with a 7D, that is just a conclusion I came to after using both enough to make this kind of comment. As a complete aside, it was taken with a 17mm TS-E, a lens for which there is no 7D focal length equivalent.</p><div></div>

[mattb1]

<p>Wow, sorry but I could not get myself to read every single post. I think I have seen most of it before. </p>

<p>It is some funny, in a sad way, how people with camera A feel the need to convince everyone that they made the best purchase choice and want to make everyone see the world their way. Shouldn't it be enough that they relate their opinion and how happy they are with their camera? But, no for some reason they feel the need to "prove" they are right, is it a superiority thing? Why is it threatening to them that so many others have a different opinion? Please, it is getting very old. </p>

<p>It reminds me of the 35mm film people who would state that the 35mm format was the best and other formats were just fooling themselves and others. Sure35mm had advantages, but it was not the best at all things.</p>

<p>If someone says they prefer camera B to your camera A, that does not mean camera A is bad for everyone. Nor does it mean that you are not smart because you choose camera A. It means that for some set of preferences, camera B is a better tool for them.</p>

<p>The 7D is a great camera, and so is the 5DII. Slightly different tools. Lately I have recommended the 7D more because those people would do better with that too. In this case though I would recommend the 5D or 5DII or wait for the 5D III. </p>

[dcstep]

<p>Scott, I do concur that your sample image with a 7D would be problematic with noise, either on the computer screen at 100% or printed large. However, if you only uploaded it in the size you're showing in this thread, few would see the difference, so long as you didn't use high-ISO. To shoot it with the 7D it would need to be on the tripod at low ISO.</p>

<p>OTOH, if you shot it at ISO 1600 with the 7D you'd likely end up with cartoon like distortion, even at this relatively small size.</p>

[danielleetaylor]

 

<p>Scott - <em>But, Daniel, they are both correctly exposed. </em></p>

<p>They are both exposed within the dynamic range of the sensors. We cannot know that both sensors received identical exposure without having a calibrated test target in the scene. Given that these two test scenes both have a relatively narrow scene brightness range, and that you could be several stops off and still have a perfectly usable image with modern RAW converters, that's a huge variable for a noise test.</p>

<p><em>You said "regardless of EV" then you argue that "You cannot assume your sensor received the same exposure, the same amount of light, as Keith's did.", you can't have it both ways. </em></p>

<p>Regardless of <strong>scene EV. </strong>In other words, regardless of scene brightness. Please do not trim the quote as it changes the meaning.</p>

<p>Look at the Imaging Resource Low Light Test series for any body. They list scene brightness in lux across the column headers, and ISO down the rows. In each cell is the test image and the exposure information. Note that they adjust exposure time in response to changes in scene brightness to keep the exposure at the sensor, the amount of light hitting the sensor, identical. If lux decreases by half (1 stop), then the time the shutter is open increases by 2x (1 stop) to compensate. So regardless of scene brightness the actual exposure at the sensor remains the same within a row. And if you study the samples you find that noise does not change for a given ISO based on scene brightness, but only changes when exceptionally long shutter times come into play.</p>

<p>Therefore <em>a properly exposed ISO 3200 test shot is a properly exposed ISO 3200 test shot regardless of scene brightness.</em> (I've changed one word to attempt to avoid confusion.) The whole point of having variable apertures and shutter speeds is to compensate for scene brightness to insure that the right amount of light hits the sensor or film to form a usable image. In the case of a noise test it's critical to insure that the same exact amount of light hits the sensor for each camera.</p>

<p>So what's the problem with your image versus Keith's? You used matrix metering and program AE to take your shots. You know the amount of light in your scenes was different because neither of you controlled it. You know from EXIF data the shutter and aperture that AE chose. But the critical thing you don't know is scene brightness (in lux or stated as EV). So you have no idea how much exposure AE chose to give each scene. You do not know if you have identical exposures, the same exact amount of light hitting the sensor in each case. You only know that you have two exposures that fit within the dynamic range of the sensors.</p>

<p><em>My image is correctly exposed, </em></p>

<p>It is more correct to say that your image did not receive the same amount of exposure as Keith's. And judging from the noise in the gray values in your scene as compared to a calibrated reference (the IR ISO test series for your model), your exposure most likely received 1.5 to 2 stops more light than a calibrated exposure at ISO 6400 should have received. Put another way, your exposure looks as clean or almost as clean as their ISO 1600 test, and is certainly cleaner than their ISO 3200 test. Now either your sensor is magic and nearly two stops better at high ISO than their test camera was, or AE gave the scene more exposure in your case. I'm pretty confident it's the latter.</p>

<p>If anyone wants to judge noise levels of a 1Ds mkIII vs a 7D, they should turn to the calibrated tests at either Imaging Resource or DPReview. There are no conclusions to be drawn from your attempt.</p>

<p>If you do not understand what I am saying here after two long explanations, or why it's impossible to assume that your AE program and Keith's AE program chose values which resulted in the same precise exposure in two very different scenes, then you do not understand exposure at all.</p>

 

[danielleetaylor]

<p><em>The reason you are failing here, the reason you get into these arguments and end up contradicting yourself, is because the only experience you seem to have with the 5D MkII is to download a few bench test images then print 8x10 crops from 16x24 image files.</em></p>

<p>I have not contradicted myself. But I do not think you have a technical understanding of exposure sufficient to understand your own mistakes. Like a snap shooter you assume that setting ISO and AE means you get the same "correct" exposure every time. This is surprising to me for someone who claims to be so technically demanding, but I don't see another conclusion at this point.</p>

<p>BTW, I have stated in these types of threads, directly to you in fact, that I borrowed, shot, and tested a 5D mkII and a 7D to my satisfaction before choosing to purchase the 7D. Is it so hard to be honest in these debates Scott?<em> </em></p>

<p><em>The problem with that, and with your entire position, is that those image files are not good representations of real world shooting image files,</em></p>

<p>I think it is absolutely hilarious that when debating FF vs. crop sensors in focal length limited scenarios you state that the real world is less demanding than the lab. That is to so say that any differences seen in tests are less likely to be seen under real world conditions. You even give a whole list of reasons why. When debating FF vs. crop in non focal length limited scenarios your position changes 180 degrees, and it's suddenly the real world that's more demanding.</p>

<p>Let's get down to your overall problem in these threads: you claim that there is a large difference in IQ between the 5D mkII and 7D, yet you are completely incapable of posting evidence to support your position. In addition you willfully ignore and wave aside any and all evidence to the contrary, no matter how precisely executed or convincing. As Keith pointed out with his Family Guy link, you simply sit and repeat your position over and over again, as if saying the same thing one more time will disprove all other evidence and commentary.</p>

<p>My real problem with these threads is that I get suckered in and then, enjoying a good debate, I don't know when to recognize that the debate is not good and it's time to just stop. I should have long ago started replying to your posts with just one sentence: "Do you have any evidence for our consideration yet?" Because until you do, you really have nothing to add to the conversation, and zero right to lecture those of us who have attempted to measure, understand, and judge these sensors.</p>

<p>Both Dave and I have made and downloaded test shots between these cameras; processed them to the best of our ability; made 24" test prints (or 10" crop prints that represent the resolution at 24"); and reviewed them with peers. Our conclusions are the same as those in the Amateur Photographer print test: there's no significant difference in IQ at low to mid ISO.</p>

<p>I don't think you've ever done this. I think you've looked at a few RAW files out of camera and concluded that the 5D mkII was better. It is if you're pixel peeping out of camera. But those differences do not survive post processing and printing. You would know that if you ever performed an honest print evaluation.<em> </em></p>

<p><em>The surfer is pretty cool, shame it is over exposed :-)</em></p>

<p>No, it's not. Either calibrate your monitor or spend some time studying professional surf photography.</p>

[danielleetaylor]

<p>Scott - <em>You and Dave have both put up one image each that neither of you took and it has been suggested why they are a meaningless comparison anyway,</em></p>

<p>Suggesting something does not make it so, and your only valid question was addressed. <strong>Do you have any evidence for our consideration yet?</strong></p>

<p>I'll note that there is an incredible number of test images available online, yet you cannot find even one set to prove your point. Hmmm...</p>

<p><em>Keith put up one image that I embarrassed by offering a fairly similar image of much higher quality,</em></p>

<p>Your image was not similar at all. It received 1.5-2 stops more exposure than Keith's.<em> </em></p>

<p><em>rather than accept that you chose to question the methodology getting yourself in terribly contradictory knots.</em></p>

<p>I have not contradicted myself at all. But your assertion that I have has revealed that you do not properly understand exposure, a strange contradiction for someone who is so <em>technically demanding</em> that they must shoot FF.<em> </em></p>

<p><em>I have still seen no "hard evidence" from anybody to support your assertions,</em></p>

<p>How could you with your hands covering your eyes?<em> </em></p>

<p><em>Just on the numbers game, of the owners of both, 4 times more people say the 5D MkII is noticeably better than the 7D. Are they all wrong? Really?</em></p>

<p>http://www.nizkor.org/features/fallacies/appeal-to-popularity.html</p>

<p>http://www.nizkor.org/features/fallacies/biased-sample.html</p>

<p>In addition to the two fallacies your numbers represent, you need to define "noticeably". If you mean pixel peeping out of camera with no post work, then they are correct. I have said I see as much. If you mean high ISO, then they are correct. I have said I see as much. If you mean 24" prints of properly post processed, low to mid ISO images, then yes they are all wrong. They could not reliably choose between such prints in a double blind study.<br /><br /></p>

[danielleetaylor]

<p>Philip - <em>Daniel - the only reason I ask is that a comparison means testing both items - not just one. </em></p>

<p>I don't mean to be short with you, but I have to keep answering that question for Scott and it gets on my nerves. I do not have extensive hands on experience with the 5D mkII, but I did borrow and test both bodies to my satisfaction before buying.</p>

<p><em>By the way I find that with the 7D I try and shoot at ISO 800 or below. ISO 1600 is quite usable with care and with very careful post processing you can get acceptable shots (indoor sports) at ISO 3200. On the 5DII I find that the images are good up to ISO 1600 and very usable at ISO 3200. In my experience at higher ISO the 5DII is 1-2 stops better than the 7D.</em></p>

<p>I agree that the 5D mkII is better at high ISO with the caveat that the latest RAW converters and post processing techniques may reduce that gap considerably as Keith asserts. My evaluation is a little more positive than yours. I think 3200 is pretty good, and 6400 acceptable with very careful post processing.</p>

<p>I appreciate your notes from Popular Photography and do not dismiss them. Again I will say that out of camera the 5D mkII is a bit better. The two questions are:</p>

<p>* To what degree can post processing eliminate the difference?</p>

<p>* After post processing, how visible is any remaining difference in print?</p>

<p>My assertion is that at low to mid ISO with proper post processing the differences are essentially eliminated, that it's impossible to distinguish between prints at that point, and very nearly impossible to distinguish between post processed images even while pixel peeping.</p>

[danielleetaylor]

 

<p>Scott - <em>But, if you don't see that my horse image is substantially better image quality than your lion then we are talking about different things.</em></p>

<p>Of course you're talking about different things. Different exposures to be exact.</p>

<p><em>I believe (along with 75% of the owners of both that expressed their opinions in this thread and Zvia after using both), that the image quality differences are noticeable sooner than you, Dave and Daniel do.</em></p>

<p>Aside from the already discussed fallacies, I don't believe all these people stated when the differences are noticeable, only that they saw differences out of camera while pixel peeping. Did Zvia make test prints and evaluate them? How many people that you count have? Let's see a table of who, what size test prints they made at what ISOs, and when they felt differences became apparent by size and ISO. Seriously. If you're going to make an appeal to popularity, you better at least document the popularity correctly.</p>

<p><em>As an example of this difference look at the IQ from your lion image and David's very nice real world raptor image that is totally destroyed by noise. </em></p>

<p>David's raptor image is a terrible example in that the scene brightness range exceeded the sensor DR and the raptor was severely underexposed by AE due to the white sky background. The 7D will not tolerate underexposure at ISO 6400. A 5D mkII is cleaner at ISO 6400, and I believe has a bit more DR. But a 5D mkII in the same situation wouldn't have done much better because AE still would have severely underexposed the target due to the sky. If David was focal length limited in this situation (most likely) the 5D mkII would have made a worse image.</p>

<p><em>Here is an image that I know from experience would not print out as well had I taken it with a 7D, that is just a conclusion I came to after using both enough to make this kind of comment. As a complete aside, it was taken with a 17mm TS-E, a lens for which there is no 7D focal length equivalent.</em></p>

<p>Unless you were using movements or a really high ISO you're not telling us about, I see no reason why that shot couldn't have been made with a 7D, or why the prints wouldn't have been every bit as good so long as you nailed the exposure. That is a situation where underexposure would not be tolerated well by the 7D, but probably would be tolerated or at least tolerated better by a 5D mkII.</p>

 

[Dave Luttmann]

<p>Wow, 27 pages and still no sample from anyone to show a "striking" difference in detail. Guess that's about it then. Samples have been provided under lab conditions showing there is no difference, and many paragraphs have been written (without a single sample) saying the difference is striking. Pretty easy to see what's going.</p>

<p>I agree Daniel, I enjoy debate. But one cannot debate with someone that passes over a lab test claiming it's invalid.....and instead expects you to simply believe his opinion.....without any evidence.</p>

<p>Funny, if the difference is so striking.....why can't they post a single image to show that? Kind of obvious why! ;-)</p>

[philip_wilson]

 

<p>Daniel I use my 5DII and 7D extensively and indeed probably take as many shots with each one - although I carry my 5DII more often. The reason for this is that I generally use the 7D for sports (Ski racing and Ice Hockey) where I take many more shots as I often need good shots of many athletes. In general here are my findings (some of which I stated on about page 5 or 6 of this post)<br>

1 I find the 7D performs about as well as the 1DIIN it replaced - AF almost as good high ISO performance very similar (slightly more noise but the extra pixels allow you to get similar results in post processing) and low ISO performance is better due to the extra resolution.<br>

2 Both the 5DII and 7D make very high quality images at low ISO and I have used the 7D up to 13" x 19" without any issues. That said the 5DII for me produces higher image quality. One factor we have not discussed is lens performance and without wishing to start a new debate I find that the higher pixel density of the 7D reveals lens short coming much more than the 5DII sensor does. The main lens I see this with is my 16-35 F2.8 II which is a good lens but not a great one. When shot wide open and at longer focal lengths the image edges from the 7D are not great.<br>

3 I find that at higher ISOs the 7D is more sensitive to slight exposure issues (especially underexposure) than the 5DII. I suspect that this is due to the fact that the 7D sensor has a lower dynamic range than the 5DII sensor at ISO 800 and above (from test charts I have seen). I also find that at higher ISOs the 7D images benefit from the DPP RAW convertor whereas with the 5DII it is difficult to see any difference between DPP and ACR. If you can suggest any better RAW convertors let me know (my ACR is the latest version 6.6.0.205 that used with CS5 on a Mac).</p>

 

 

[philip_wilson]

<p>Daniel I forgot to add - look at the test images at the bottom of page 47 in the AP you reference (despite being in Canada I have subscribed for many years as it is probably the best photography publication). These are the pictures of the lines from the resolution chart. While AP show that they can resolve 30 lines at 100 ISO from both the 7D and the 5DII look at the images honestly. The 5DII images are much sharper than the 7D images and I can count the 5DII lines easily - the 7D lines are much harder to count. You have to admit that the 5DII crops do look much crisper. Whether this bothers you or not depends on how big you print and what you are looking for.</p>

[dcstep]

<p>Daniel said:<br>

"David's raptor image is a terrible example in that the scene brightness range exceeded the sensor DR and the raptor was severely underexposed by AE due to the white sky background. The 7D will not tolerate underexposure at ISO 6400. A 5D mkII is cleaner at ISO 6400, and I believe has a bit more DR. But a 5D mkII in the same situation wouldn't have done much better because AE still would have severely underexposed the target due to the sky. If David was focal length limited in this situation (most likely) the 5D mkII would have made a worse image."</p>

<p>With 20-20 hind-sight, I'm wishing that I'd thrown the 1.4x TC on the 500mm and shot it with the 5D2 with the same settings. I was NOT focal length limited. The crop was only 10% or so to align and balance things. It would have been a very good comparison if I'd thought to try the 5D2 for the same shot. I too think that the 5D2 would have done better in these conditions in this setting.</p>

 

[ken_max__parks]

<p>If money is the issue, then get the 7D</p>

<p>If money is NOT an issue, then get both the 7D and the 5Dmk2.</p>

<p>Enough said!</p>

[scott_ferris]

<p>Daniel,</p>

<p>Many thanks for all the advice on exposure and such, must be honest, I really don't know how I got on before your comments, I suppose I just muddled through blindly relying on my cameras green square mode! Anyway thought you might like to see how I have come on after following your advice, being as how you are such a knowledgeable and gracious professional surf photographer an' all, here is my best selling print this past Christmas season.</p>

<p>Take care, Scott.</p>

<div></div>

[adam_mueller]

<p>5D Mk2. I take pics of my little girl with it no probs.<br>

I have the 7D too which is great also but the build feels better on the 5D2 & aftre going FF you wont go back. :-)</p>

[scott_ferris]

<p>The point of the last post was to show another picture the 7D could not have taken and printed anywhere near as well as a "FF" camera can. Here is a 100% crop, from the second image in the sequence, with similar horrific back light to David's raptor image, it retains detail and smooth gradations in colour long after the 7D would have blown out and blocked up.</p>

<p>To me, like David's image and my dinner image, it is just another real world image the 7D could not shoot anywhere near as well as a FF camera can. I know for a fact the real world is much harder on sensor performance than bench test sampling. Anybody that doesn't know that, after using a camera and looking at the results, isn't using a camera like I do.</p><div></div>

[Dave Luttmann]

<p>Scott, being that you know this so well, it should be easy for you to produce a side by side comparison showing the striking difference as has been suggested. That said, I see nothing in that image that couldn't have been captured with a 7D, 1D2, 1D3, Nikon D2X, etc, etc. Seen hundreds like it.</p>

<p>Still waiting for a comparison set.</p>

[scott_ferris]

<p>Of course you don't Dave, you shoot weddings and portraits where you can control the light, contrast and dynamic range to a large degree, well I do when I shoot weddings and portraits. The three images I have used as examples of problem areas are beyond the control of the photographer and would print differently with a 7D or a FF camera file.</p>

<p>Yet again, I don't have any 7D files with me, I didn't buy the camera after using it for a month because it didn't do what I needed, but I do get to use one again for ten days from the 9th of Feb, when I get back home, and I have said I am happy to post images here or start another thread, and if people have specific comparison images they'd like me to shoot I am happy to do it. Now in the mean time how about you post me a comparison file, you shot, in very contrasty lighting, very strong back light or dull fine tones of grey (like the dinner image) and tell me you can't see a difference in medium sized prints.</p>