Return of Interesting article on DxO about 5D III and D800...

Started Apr 10, 2013 | Discussions thread
Shop cameras & lenses ▾
Forum ProPosts: 10,350Gear list
Re: Return of Interesting article on DxO about 5D III and D800...
In reply to Steen Bay, Apr 14, 2013

Steen Bay wrote:

qianp2k wrote:

Steen Bay wrote:

qianp2k wrote:

Steen Bay wrote:

From DPR's 5D and 7D reviews - horisontal/vertical resolution (extinction resolution) :

5D - 2300/2000 (2500/2500) LPH

7D - 2500/2450 (3100/3050) LPH

I am not sure how DPR calculated into above data?  From the same AOV?  Using the same lens? Under that MTF percentage? 5D is 8-yr-old and DPR Lab has changed its test environment between when they reviewed 5D and 7D.  I wish DPR can add some old cameras such as 5D into new standard lab samples

Like I said, then I think that DPR uses the EF 85/1.8 on Canon FF and the 50/1.4 on Canon APS-C. The studio scene has changed since the 5D review, but think that the same resolution test chart was used for 5D and 7D. The AoV (or rather the framing) was of course the same, and they 'calculated' (or rather 'estimated') the resolution figures by simply looking at the testshots (at least 100% view on the monitor, I suppose). You can take a look at the resolution test shots yourself. They are available for download (Page 31 in the 5D review and page 28-29 in the 7D review).

Sure.  I downloaded both JPEG (if original RAWs are available?), and then compare them side by side.  Honestly my eyes don't see much difference except 7D one is bigger and darker.  As I said we need to compare them side by side at the same size either by upsampling 5D to 7D file szie or downamping 7D file to 5D size otherwise different sizes will cause delusion.   Here is one of snapshots comparing them side by side at their respective full size, and I don't see much difference.  My eyes can count same numbers of lines or cannot count the same fine details that unable to figure out

Well, you'll have to look at the finer lines where the difference will be rather clear (see for example around '24', which represents 2400 LPH).

I am sorry but that not clear to me.  As I said they need to compare at the same size.

No mention DPR used different lenses and their AOV are not exact the same if they shoot from the same distance.  From what I see, 5D file is a bit sharper and 7D file shows noticeable purple CA.

And use the same lens in the exact same AOV.  I believe DXOMark tests are more controlled and more accurately.  DPR used two different lenses to make their tests largely irrelevant.

According to DXOMark by checking Measurement | Sharpness | FieldMap, to my eyes 50/1.4 performs noticeably better than 85/1.8 between 5D vs 7D.  So that shows DPR gave a favor to 7D with 50/1.4 over 5D with 85/1.4!  Therefore DPR tests with two different lenses are really invalid.

DXOMark 5D vs 7D with the same 85/1.8 lens

DXOMark 5D vs 7D with the same 50/1.4 lens

DXOMark 5D+85/1.8 vs 7D+50/1.4

In my real-world tests especially outdoor, I simply don't see my 60D has an advantage in resolving fine details than 5D in shots of the school building.  Honestly please download and check these original CR2 files and analysis by yourself and compare them at the same sizes by either upsampling 5D files to 60D sizes or downsampling 60D files to 5D sizes thru Photoshop Bicubic enlargement/reduction.  They need to be compared at the same sizes, period.

You can see whatever you want to see.  But I have to follow my eyes honestly.

That's the problem here.. according to DxO's P-Mpix scxore the 5D beats the 7D, but if looking at DPR's resolution test chart shots, then it's rather clear that the 7D has a higher resolution.

The key is that if those paper resolution actually can be seen by normal human eyes?  In those two DXOMark test charts with very nice prime lenses 100L and 85/1.8 (no mention inferior lenses), we can see 7D is only better at MTR 15% and below (even lower with inferior lenses) that are very low contrast and virtually invisible by human eyes even under grainless/noiseless lab environment (no mention in imperfect real-world environment .  5D is better above 15% MTF that is more meaningful and matter to human eyes.  It's all about perceptual sharpness or perceptual resolution as DXOmark uses two words interchangeably.

5D vs. 7D is just an example. Think that we'll have a similar 'problem' (acutance/sharpness vs. resolution) if comparing for example 5D3 vs. D800, or if looking at the effect of diffraction when stopping down.

I am sure 41mp Nokia 808 is 41mp on paper. In lab test, it may has 1% MTF data higher than 36mp D800 or 2% higher than 22mp 5D3.  The only matter is whether you eyes can actually see that?    If you eyes unable to tell, then those paper resolution is meaningless that you can think but you cannot see, heeh.

-- hide signature --
-- hide signature --
-- hide signature --
 qianp2k's gear list:qianp2k's gear list
Sony RX100 Canon EOS-1D Mark III Canon EOS 5D Mark II Canon EOS 5D Mark III Sony Alpha 7R +20 more
Reply   Reply with quote   Complain
Post (hide subjects)Posted by
Keyboard shortcuts:
FForum PPrevious NNext WNext unread UUpvote SSubscribe RReply QQuote BBookmark post MMy threads
Color scheme? Blue / Yellow