Jump to content

Interpreting DXO Mark lens results


steve_wagner1

Recommended Posts

<p>Just wondering how people interpret results like these</p>

 

<ul>

<li>Canon 85mm 1.8 on 5d2 - 59/61, nearly the highest rated combo of any lens/body in all of the site</li>

</ul>

<ul>

<li>Canon 85, 1.8 on 7D - 33/61, pretty dramatic drop</li>

</ul>

<ul>

<li>5d is still quite a bit higher score than 7D - 44/61</li>

</ul>

<p>Among other things the resolution (in lp/mm) really drops off. It seems due to sensor size, not mp, since the 5d has fewer mp than 7d, but scores considerably higher in resolution and overall.</p>

Link to comment
Share on other sites

<p>I have both cameras and find that in the real world the 5DII is a big step up on the 7D in terms of resolution and image quality. You only really see the difference when you enlarge a lot or pixel peep. DXO has the 5DII and 85 F1.8 resolving 66 lpm compared to 47 for the 7D and the 85 F1.8. Similarly with the 50 F1.4 the 5DII resolves 63 lpm vs 50 on the 7D. I personally find that the 7D pixel density really stresses lenses if you look closely. The 5DII sensor is 70 lpm whereas the 7D is 116. If you look at the sensor comparsson between the 5DII and the 7D you find the 5DII has 4dB more signal to noise ratio at any ISO and a dynamic range 1EV better from ISO 600 and it has better Tonal range and colour sensitivity. Interestingly they find that despite fewer pixels the first 5D and 85 f1.8 outresolve the 7D but the 1DIII and the 7D produce similar results.</p>
Link to comment
Share on other sites

<p>Thanks Phillip, the 5d part did pique my interest as I just had my 5d2 stolen and I am considering replacing it with a 5d, which will eventually become a backup for another 5d2, instead of going with a 7d. I am attracted by the lower noise, the DOF, and the no crop factor, among other IQ aspects, and this data regarding resolution of different lenses with the 5d outperforming the 7d across the board is another thing steering me in that direction.</p>
Link to comment
Share on other sites

<blockquote>

<p>the trick to interpreting DXO's results is to not interpret them at all. Ignore them completely.</p>

</blockquote>

<p>Yep, that's the only way to make sense of this:</p>

<blockquote>

<p>5d is still quite a bit higher score than 7D - 44/61<br>

Among other things the resolution (in lp/mm) really drops off. It seems due to sensor size, not mp, since the 5d has fewer mp than 7d, but scores considerably higher in resolution and overall.</p>

</blockquote>

Link to comment
Share on other sites

<p>Not trying to be a smartarse either, but wouldn't it be an idea to put your opening question to DXO themselves, Brett? The site as such is not very good in helping to fathom their results. As for the Canons, as the lens is the same, you are obviously seeing the differences in sensors, results for which have been available for some time. I don't quite understand, why you get so excited now that they add a lens to that.</p>

 

Link to comment
Share on other sites

<p>My pictures of speedway are excellent with my 7D, I doubt very much if the 5D2 could Ai Servo as well as the 7D. My A3 prints look great, and my flickr posts (at 0.5MP) get good feedback.<br>

My landscapes are brilliant with my 7D, I'm sure those from the 5D2 would be even better in the same situation.<br>

But I don't have a 5D2 so what am I to do?<br>

Get out with my camera and enjoy it. Take images I'm very happy with?<br>

Or pixel peep?<br>

Or stop a lens down to f22 to see how my ISO 12'800 works out in a coal shed at midnight?<br>

I haven't taken any images of resolution charts so I guess I'll leave that to DXO.<br>

I know that for a few years Nikon users have gloated about DXO charts so I kind of stopped listening. I know that for the 90s and most of the naughties I was able to gloat about have fast accurate AF. I can also boast about getting a camera above 12MP without re-mortaging my house.<br>

I'm with Nathan. In all seriousness. Does whatever DXO say affect my enjoyment? Nope.<br>

Camera performance is actually rather brilliant these days. From the 7D and from the 5D2 and even from some Nikon cameras too.<br>

Charts and pixel peeping don't really have all that much of a real world bearing when you get that feeling after you've got the shot.<br>

5D2 users want to be better than 7D users, Nikon users want to be better than Canon users. Hey ho. The camera is really the least of it.</p>

Link to comment
Share on other sites

<p>Paul, the 5D2 has better resolution than the 7D. Most of us that own both concur, BUT the 7D is better for wildlife and most sport shooting because of its AF and burst-rate. I end up with more bird in flight "keepers" with my 7D than my 5D2 because the birds are larger in the viewfinder and easier for me to keep the focus point on. Still, there are plenty of times when I'm shooting at dawn or dusk and wishing for the high-ISO performance of the 5D2. Pixel-peep a 5D2 scenic image at any ISO vs. the 7D in the same situation and you'll pick the 5D2 for all your scenics, protraits and travel imaging and anything that's not too fast for the 5D2.</p>

<p>One's a FF and one's a crop sensor. No one expects the same thing out of both, just like no one expected the same thing out of a 35mm film Canon and a medium format Hasselblad. They're simply different tools for different jobs. DxO and we 5D2 owners are not out to diss your 7D, but the fact of life is that its IQ is not as good as Canon's FF bodies. DxO measures this, but those measurements are of RAW, unprocessed images. In this digital age and image is not an image until you've completed the RAW conversion, noise reduction and sharpening, if any. Some would say, you really need to consider the image after it's been corrected for geometric distortion, vignetting, etc., all of which we're routinely doing before presenting out images, even on Flickr. (I'm a member there too, so don't think I'm dissing Flickr).</p>

Link to comment
Share on other sites

<p>Hi David, believe it or not, I think we are actually singing from the same song sheet here.<br>

I have a 7D and an delighted with it. Regardless of DXO mark or anything else.<br>

I could have had a 5D2 if I wanted it, but when I was buying the video mode on the 7D was far better especially for PAL region users (I bought before the 5D2 firmware upgrade) and that is important to me. As I also like nature and sports, then yes as I said in my post, the better AF and speed is very useful.<br>

For landscapes I shoot at ISO 100 or 200 max anyway (old film habits die hard) and whilst if I were to pixel peep I may see a difference with a 5D2, I'm happy enough (more than actually) with what I'm getting.<br>

My point was not to diss 5D2 users, or particularly DXO, just pointing out what is more important:<br>

Taking pictures of actual subjects or of resolution charts?<br>

Taking nice images that look brilliant at real life viewing distances or pixel peeping?<br>

The original question was about the importance of these DXO tests, I think my opinion is pretty clear.<br>

Photographers take photographs. Cameras, lenses, DXO correction and everything else are just tools in their arsenal.<br>

The difference between my ability and somebody elses is going to make a bigger difference than the difference in noise performance between a 7D at 3200 and a 5D2 at 3200. So I just try to be a better photographer rather than pixel peep or pore over DXO graphs.<br>

Enjoy your 5D2 mate, sniffy DXO quoting 5D2 users are out there though (the reason I'm not a member of any camera club, when I had a film 3 it wasn't a film 1V, when I had a 16-35 f2.8L it wasn't a 16-35 f2.8L II and so on "I don't know how you can bear to take images with that thing..." etc) I am glad you are not amongst their number.</p>

Link to comment
Share on other sites

<p>Paul, I was trying to look at your Flickr Photostream to see what you're talking about, but I couldn't find a Paul Russell using a 7D posting there. My Photostream is at <a href="http://www.flickr.com/photos/dcstep/">http://www.flickr.com/photos/dcstep/</a> Could you link me to your Flickr Photostream?</p>

<p>Once you get to a certain level of competency and shoot in situations where high-ISO might come into play, like shooting birds at dawn and needing a shutter speed at 1/1000-second or faster, high-ISO performance does become important. You don't need to pixel-peep to see the difference, particularly when there's a dark background.</p>

<p>A lot of my Contacts on Flickr are fellow bird and wildlife photographers. Many of us find it really interesting to pixel-peep bird feather detail or a bee's finer details in a well executed macro image. Before I do my RAW conversion I Preview my shots full-screen and sometimes push them up to 100% to compare eye sharpness and focus. I DO understand that most viewers will be using lower resolution, but if I have a choice between two equivalent poses then I take the sharpest of the two. The pose, background, setting and story always comes first, but I take the sharpest when all the other objectives have been met.</p>

<p>So, I'm routinely comparing 5D2 images and 7D images on my monitor and often at 100%. It's clear and easy to see in certain situations why a lab would rate the IQ of the 5D2 higher. Despite that, 80% of my images are taken with the 7D, because I find it easier to use for birds and wildlife. So, I think we agree that the right tool for a job is usually more important that absolute IQ (at least when a minimum standard has been met). Still, there's no need to get defensive and ride on you highest horse just because your camera doesn't score as well as an alternative. It's good an useful information to have, at least for some of us.</p>

Link to comment
Share on other sites

<p>I'm not on any high horse David mate. I don't really care if the 5D2 gets a better score under lab conditions before RAW processing. It's not DXO's fault but I fear that their tables have rather became a form of top trumps. Skill is important. How a camera handles and responds are tactile and important. Sensor performance is important. Post processing is important. To focus on one component alone is rather futile. I had rather thought my previous post would have made that clear and would have smoothed your ruffled feathers. That it hasn't means our exchange must end here.</p>

 

Link to comment
Share on other sites

Create an account or sign in to comment

You need to be a member in order to leave a comment

Create an account

Sign up for a new account in our community. It's easy!

Register a new account

Sign in

Already have an account? Sign in here.

Sign In Now
×
×
  • Create New...