5DS scores at DXO **now posted**
- Thread starter ahsanford
- Start date
Take a look at Ken Tanaka's comments which are found just under the main body of the article ->
http://theonlinephotographer.typepad.com/the_online_photographer/2015/07/reigning-%C3%BCbercamera.html
That might help answer some of your (and perhaps a very great many other people's) questions.
It pretty much sums up what's "meaningful" and what's not.
For myself, I've found the cheap, small Sony A6000 to be a superior image making tool to the much more massive 5D Mk-anything. It'll be interesting to see how this whole thing plays out in APS-C format cameras when Sony intro's it's A7000 (or whatever they're going to call it) next month.
http://theonlinephotographer.typepad.com/the_online_photographer/2015/07/reigning-%C3%BCbercamera.html
That might help answer some of your (and perhaps a very great many other people's) questions.
It pretty much sums up what's "meaningful" and what's not.
For myself, I've found the cheap, small Sony A6000 to be a superior image making tool to the much more massive 5D Mk-anything. It'll be interesting to see how this whole thing plays out in APS-C format cameras when Sony intro's it's A7000 (or whatever they're going to call it) next month.
RGF said:
Upvote
0
Re: 5DS scores at DXO drop **tomorrow** (i.e. 7/8/15)
If you mean the lens scores in combination with a camera, then that's logical. A camera has DR, lens delivers aspects like sharpness; so as a whole system you have things like DR, ISO performance, sharpness, vignetting, etc. But still, I haven't seen that in their tests either, so I'd appreciate if someone would point me in the right direction of the 'wrongdoings'.
Ok, can you point to where it states that DxO measures 'DR of lenses'? I have never seen that.bdunbar79 said:
If you mean the lens scores in combination with a camera, then that's logical. A camera has DR, lens delivers aspects like sharpness; so as a whole system you have things like DR, ISO performance, sharpness, vignetting, etc. But still, I haven't seen that in their tests either, so I'd appreciate if someone would point me in the right direction of the 'wrongdoings'.
Upvote
0
L
Loswr
Guest
Corneria said:
In DxO's own words a requirement for the Lens Score is, "growing linearly with the sensor dynamic range, measured in f-stops, for a perfect optic."
As you say, they 'score' lenses in combination with cameras. But while you and I know that, I suspect we're in a small minority of people who view their comparisons. Even the way they present their scores is, IMO, deceptive. They show the DxOMark Score on top of a list of lens-specific metrics, giving the impression that the score is a composite of those metrics, when in fact that score is better described as suitability for taking pictures of flat subjects in a dimly-lit warehouse.
DxOMark's Lens Score calculation is even more of a black box than their Sensor Score (more of their Bad Science = BS) but looking over their numbers it's evident that the two most important factors are sensor score and lens transmission. Thus, as below you can compare two excellent lenses, where one is sharper, has slightly higher trasnmission, and has much less CA, and see that DxO gives them the same score.
Attachments
Upvote
0
neuroanatomist said:
Don't understand why the Canon and Nikon both got the same score.
Canon wins or ties on metric
Higher resolution on a lower resolution body.
Better T value (vs published F stop)
Less distortion
Similar vignetting
Lower CA
Oh, I forgot. 5% penalty have the name Canon on the lens or body :
Upvote
0
The other example Neuro always brings up is below. Same camera body, and the big white wins each metric other than transmission (which you could argue should be reported as an accuracy to its listed max aperture rather than it's pure T-Stop value), yet it gets a lower score.
Now, reporting the actual T-stop of the lens is absolutely useful if you are buying a lens, but using transmission to state one lens with a completely different max aperture is better than other is insane.
- A
Now, reporting the actual T-stop of the lens is absolutely useful if you are buying a lens, but using transmission to state one lens with a completely different max aperture is better than other is insane.
- A
Attachments
Upvote
0
Another bizarre one, the "battle of the 500s."
The canon has significantly better sharpness (even with a significant sensor resolution disadvantage), significantly better CA, yet the DR of the D800 body is enough to give the lenses a tie score.
However, the Sony which ties or bests the Nikkor everywhere but the small transmission deficit loses by 3 points in aggregate to despite being used on a platform with roughly the same DR as the D800.
Is t-stop really weighted so significantly that .2 knocks adds 3 points?
The canon has significantly better sharpness (even with a significant sensor resolution disadvantage), significantly better CA, yet the DR of the D800 body is enough to give the lenses a tie score.
However, the Sony which ties or bests the Nikkor everywhere but the small transmission deficit loses by 3 points in aggregate to despite being used on a platform with roughly the same DR as the D800.
Is t-stop really weighted so significantly that .2 knocks adds 3 points?
Attachments
Upvote
0
L
Loswr
Guest
3kramd5 said:
It's not just DR, it's the Sensor Score - the D800 beats the A99 by 6 points. That also speaks to their wonky undisclosed BS (biased scores, bad science)...they state a 15-unit Score difference is about 1-stop of performance. That means the D800 should be ~0.4 stops 'better' than the A99. But, the D800 is 0.3 stops better for color depth, 0.4 stops better for DR, and nearly a full stop better for ISO.
So it seems they don't even follow their own description of how their sensor score works. That's BS.
Upvote
0
Hmm... Ok, I have no clue what they mean by that. Also, I take back my earlier comment about the "score for lens + camera", because they don't have any 'metrics' or any aspects about the camera stated, but they do give a "lens score" where they apparently measure some unknown aspects about the camera. I'd be fine if they would give scores for the camera+lens combination, but they should be more transparent about the camera scores then...neuroanatomist said:
ahsanford said:
Yes, I tried to understand the logics of their overall scores, but it is very vague. I have no idea what they do with the weights of the different 'metrics', but I do remember reading somewhere on their website that they look at the sharpness of the whole spectrum. So if a lens would be marvellous on f/8 but horrible on f/2.8 while another lens would be 'okay' at the whole spectrum, the former would have a higher sharpness score but a lower overall score. Or it was an overage for the sharpness score, I'm not sure.3kramd5 said:
Anyway, personally I don't look at the overall score, for example because of that transmission with equal(?) weight as sharpness. Also, I don't care about vignetting or distortion (except for ultra wide lenses). What I do like DxO for, is their sharpness graphs, those are very handy! I do wish they had a more user friendly menu though. I think www.slrgear.com has the nicest sharpness 'widget' (although I don't think the rest of the website is very user friendly), and they had them for many years.
Upvote
0
raptor3x said:
Doh. Serves me right for trying to juggle screenshots of multiple pages while posting about them on a phone.
That makes it even more inane; the Sony matches or bests the Nikkor in every listed metric (and like the Canon tops the Nikkor in sharpness in spite of a large resolution deficit) and loses by 3 points.
Upvote
0
Can I ask a slightly tangential question? The Lightroom HDR feature saves as a .dng file. I totally get that multiple exposures combined can't be considered truly 'raw' in the sense of unmodified sensor data, but the dog file acts more like a raw than a jpeg in terms of how aggressively you can postprocess it. So how does this work? What's going on? Is it like this Dxo thing, in that it's several raw files saved together, or what? I don't know much about file formats :/
Upvote
0
Just saw this from our good friend Tony;
http://northrup.photo/canon-5ds-r-dxomark-scores-why-you-shouldnt-care/
http://northrup.photo/canon-5ds-r-dxomark-scores-why-you-shouldnt-care/
Upvote
0
Northup's review of the D810 vs. 5DSR was a favorable review for the Canon well before DXO scores came out. He also is not a fan of DXO's scoring methods and preaches to disregard many of it and only focus on certain scores and data points which I totally agree with.
Upvote
0
You obviously didn't see his previous video about the d810 vs the 5dsr. He previously said that he thought the Canon was better and the extra megapixels makes a difference (at least for some people). I know people love to bash TN but I've found that he's one of the least biased people on the internet. And the amount of abuse that him and others like him have to take is amazing. He's much better than those Sony "artisans" who are just painful to watch due to the amount of bs flowing from their mouths.ahsanford said:
Upvote
0
