5DII vs 7D test, part 2
jmphotocraft
Registered Users Posts: 2,987 Major grins
Here is a landscape test of the two cameras, with more realistic sharpening applied to the 7D image. I'd say the 7D is holding its own.
http://www.jmphotocraft.com/5DII_v_7D/shore.htm
http://www.jmphotocraft.com/5DII_v_7D/shore.htm
-Jack
An "accurate" reproduction of a scene and a good photograph are often two different things.
An "accurate" reproduction of a scene and a good photograph are often two different things.
0
Comments
Got bored with digital and went back to film.
An "accurate" reproduction of a scene and a good photograph are often two different things.
Also, the 24-70 is hardly the lens I would use for tests with such images (depending on the particular copy its focus can be questionable). I'd use a prime, possibly the 135, stepping to match the frame content.
Neil
http://www.behance.net/brosepix
He did focus manually, and I guess he used a zoom so that he could get a similar field of view with each lens. The exposure seems slightly different (7d more exposed). There is some colour fringing on the 7d images which may be due to the higher pixel density exposing issues with the lens?
PS, I think the 5d would really show it's superiority at a higher iso such as 3200 or 6400.
The 100% crops are different in magnification because I did 1200x800 crops of each camera. This results in a wider FOV on 18mp than 21mp.
The color and exposure differences you see are characteristics of the sensors. Both were exposed the same (manually) and both were developed the same in DPP.
Your monitor is probably closer to 100ppi. If not, it's time to upgrade.
As it says in the test, I used manual focus with live view at 10x.
An "accurate" reproduction of a scene and a good photograph are often two different things.
I have to agree here also.
I have both camera's, and as your "test" show, even at ISO 100, the 5DMkII clearly out performs the 7D in the IQ department.
Bump those puppies up to ISO 800 and above. One leaves the other behind fast!
IMHO, that's why I love the 5DMkII so much. I shoot a lot at higher ISO, and for the most part, get great images without much processing work later. Unless I've shot purposely way under-exposed, even up to ISO 6400, I rarely use noise-reduction on the 5DMkII images.
I have always found the 1.6 crop, (choose your body here), images need way more attention later than FF images.
The two bodies are very different tools for me, each has it's strong points.
I won't use a framing hammer to do delicate finish work that is better suited for a small finishing hammer.
That's why I own and use different hammers, and camera's!
YMMV
A former sports shooter
Follow me at: https://www.flickr.com/photos/bjurasz/
My Etsy store: https://www.etsy.com/shop/mercphoto?ref=hdr_shop_menu
An "accurate" reproduction of a scene and a good photograph are often two different things.
You do realize you're looking at 100% crops of distant foliage, right?
An "accurate" reproduction of a scene and a good photograph are often two different things.
xD = joke face
why wouldn't an APS-C sensor come close? with good light and low iso an aps-c sensor can deliver on par with full frame with the exception of scenes with a large dynamoc range.
14-24 24-70 70-200mm (vr2)
85 and 50 1.4
45 PC and sb910 x2
http://www.danielkimphotography.com
Too much pixel density. I think this test proves the APS-C camera is not totally "on par" with the FF camera, but it comes close.
An "accurate" reproduction of a scene and a good photograph are often two different things.
Just a correction to my previous comment, by "stepping" I meant walking, moving the camera.
I'm not convinced that the live view-manual focus method is such a good idea. It's typically the case that the same person making multiple focus decisions of the same subject over a series of exposures will produce a range of measurement error. Therefore it is almost certain that focus is different for each of the shots.
I repeat that perceived resolution is influenced by non-resolution factors such as magnification, colour and contrast, and that translation of the data into a low resolution display is introducing unknown and variable effects into the viewing environment.
Another thing to consider is that with outdoor test data such as this, incidents such as subject movement (wind in leaves), changes in light (moving clouds, movement of the sun), changes in density of air between camera and subject due to convection, camera movement caused by environmental factors, varying sensitivity of the photographer's eye as a result of being in various stages of adaptation in moving away from and into the LCD for focusing, at least - all are uncontrollable, and will produce effects in the image discernible in 100% crops.
Standardised laboratory light bench tests and comparison of prints give a more valid and reliable result.
Neil
http://www.behance.net/brosepix
okay I was about to post I don't see the difference in the 2 (1st viewing was from a laptop) but viewing from a real monitor, there is a heck of lot of difference in the 100% crops. Can it be that different? For example I see much more CA in the 7D which I assumed is a property of the lens..not the body.
I see you sharpened the 7D image more so then the 5D2 image. I wonder if the IQ has to do with halos?
haha..wanna reshoot with same sharpening?
14-24 24-70 70-200mm (vr2)
85 and 50 1.4
45 PC and sb910 x2
http://www.danielkimphotography.com
I doubt it but I can provide that for you without reshooting, I'll just reprocess the RAW file at lower sharpening. The previous test I did I was informed that the 7D required more sharpening than the 5DII in order to do a fair comparison.
An "accurate" reproduction of a scene and a good photograph are often two different things.
I went back to your 1st thread which I hadn't looked at before. I am a bit surprised still. Yeah..it's pixel peeping but I was expecting closer results then that. Maybe the pixel density on the 7D is approaching some cliff?
14-24 24-70 70-200mm (vr2)
85 and 50 1.4
45 PC and sb910 x2
http://www.danielkimphotography.com
I believe so, with current technology. I think the APS-C sweet spot would have been 14-16mp, a la the Pentax K5 (tempting!!). I think if Canon is going to continue with an 18mp APS-C sensor or increase that mp count, they are going to have to invent a new sensor that is a quantum leap over current sensors. Otherwise if we are truly up against that cliff, they should pull back to a more realistic mp count, as they did with the G11.
An "accurate" reproduction of a scene and a good photograph are often two different things.
Yes, knowing what artifacts the testing is producing so as not to read them as results (eg discriminating between independent and dependent variables, and controlling the former) is what makes doing meaningful objective (scientific) testing such a difficult thing. It's a rare enough thing in hardcore science, let alone in "backyard" projects like this.
There's no reason not to compare the two bodies, and no reason why it cannot be done, but it is practically very difficult. What jmphotocraft is showing here might be the truth of the matter, but the doubts make what he shows basically dissatisfying. It's not clear to me what is actually being tested here (dependent variable), and what data is contaminating real data about it.
If I am not mistaken, DxO have comparisons done in more rigorous fashion which must I think be put alongside projects like what jmphotocraft has done. One of the big insights that the scientific method has given us is that in doing tests we tend to produce the results we expect. Not to provide for this tendency leaves us gullible to our own biases. I guess it comes down to how willing we are to believe ourselves.
Neil
http://www.behance.net/brosepix
Personally (on the 5D at least) I would rather them keep it to 20-25mp and increase the low light and dynamic range performance. Are there really many people clamouring for higher resolution? If they could improve the autofocus performance in low light that would be great as well.
An "accurate" reproduction of a scene and a good photograph are often two different things.
Jack, the kind of playing around that you have been doing is fun and interesting. I have looked closely again and again at the comparison shots. It strengthens my thinking that the 7D is not the upgrade I want to commit to. At the same time I know that my thinking is likely largely irrational. The economy would fall flat on its face if everyone were to think rationally.
Your comparison setup is sophisticated, but I suggest not rigorous enough. There are many differences between the images from the 5D and the 7D that you have posted, some of them due to technological factors, no doubt, and others due to chance, I'm sure (that is there would be differences between repeat images from the same camera under the same conditions), but most of these differences are small and difficult to define, and appear to be somewhat amenable to post processing. The take home impression I get is that it doesn't very much matter which camera when it comes to IQ (and IQ is a pandora's box). Other things unrelated to IQ like speed, AF, video, file size are likely going to be more significant in crystalising buyers' decisions.
So, play around, have fun, by all means. But to paraphrase Socrates, the unexamined gear comparison is not worth believing.
Neil
http://www.behance.net/brosepix
If that's your impression then you would probably be happy with a 7D. Indeed, the IQ is probably indistinguishable at any typical print size. I own a 5DII as well for all the other reasons - the VF, the lens choices, the high ISO performance, the somewhat thinner DOF. As an upgrade from a 40D or 50D, the 7D is the best thing yet. For me coming from a 1DIIN and my 5DII it's a little less exciting, but still good.
If you are a pro or semi-pro sports photographer, the 7D can back up a 1D, but as a main body it broke my heart more times than I thought it should. It got me lots of spectacular images too, and it paid for itself. The crop factor and 18mp allowed me to cover a good portion of the soccer field with just my 70-200. Still, I'm pretty sure I'm going to sell it and buy a used 1DIII.
If you aren't deeply invested in Canon glass, I'd look no further than the D700. On paper it nearly has it all. As long as you can live with 12mp, which you should unless you are regularly printing large. Or if that's out of reach, the K5. Pentax makes great lenses.
If you are invested in Canon lenses, as I am, I think if Canon does not have its head up its butt, there will be a D700 killer in the not so distant future, and it will be called the 3D. THAT should be the one camera to rule them all.
An "accurate" reproduction of a scene and a good photograph are often two different things.
Which camera you require depends on your use. I think the 5D series is mainly aimed at wedding photographer like myself, and it's well suited to that with good low light performance, full frame sensor, reasonable cost, video and good enough AF. It seems to have taken over where the old 5D left off as the 'standard' wedding photography camera. I think the next version will probably be similar in most respects and I suspect they will want to keep the cost as low as possible.
I am a passionate amateur, fairly new to the dSLR world, and recently upgraded from a 50D to a 5DmII. I went back and forth between a 5D and a 1Dm4, and ultimately went with the 5DmII for several reasons. I borrowed a 1D and shot several collegiate soccer games, and found it to be too heavy and bulky. And although the 1D fps is much better (the biggest downside of the 5DmII IMHO), I rarely use continuous shooting. In my experience, good timing produces much better results than just holding down the shutter and letting her rip. Also, in a side-by-side comparison of the 1D and 5D, I liked the image quality better in the latter; nothing scientific, just subjective.
If you can live with less fps, and like the look/feel of a full frame sensor, the 5D is great. I can't vouch for its ruggedness, as mine is new and I don't put it through the same paces as a pro.
Attached is a pic from a night soccer game taken with my 5D and 70-200/2.8. f2.8, 1/500, ISO 6400. No content has been altered; just CS5 noise reduction, cropped, WB and exposure corrected.
http://s939.photobucket.com/albums/ad232/San_Francisco_United/Soccer%207%20November/?action=view¤t=_MG_6451.jpg