p.1 #1Canon 5DS dynamic range: zero improvement (RAWs analyzed)
I have analyzed the 5DS DR using the RAWs published by Imaging Resource (as somebody hinted, they can be easily downloaded by replacing JPG by CR2 in the URL; e.g. E5DSRhSLI00100NR0.JPG to E5DSRhSLI00100NR0.CR2).
It seems that the 5DS has yet another boring Canon sensor with, again, the same boring 11 f-stop DR at pixel level on base ISO. The 12.50 DR value is calculated by adding 1.32 to "convert" into the DxO notation (downsizing to 8 megapixels):
ISO 0100 - DR 12.50 at 8MP (11.19 at 50MP) read noise 6.31819 DN (start 6.23) saturation 14733 DN
ISO 0200 - DR 12.50 at 8MP (11.18 at 50MP) read noise 6.62953 DN (start 5.98) saturation 15383 DN
ISO 0400 - DR 12.28 at 8MP (10.96 at 50MP) read noise 7.71204 DN (start 6.12) saturation 15383 DN
ISO 0800 - DR 11.86 at 8MP (10.54 at 50MP) read noise 10.3258 DN (start 5.93) saturation 15384 DN
ISO 1600 - DR 11.18 at 8MP (9.86 at 50MP) read noise 16.5492 DN (start 6.06) saturation 15385 DN
ISO 3200 - DR 10.37 at 8MP (9.05 at 50MP) read noise 29.079 DN (start 6.83) saturation 15386 DN
ISO 6400 - DR 9.58 at 8MP (8.26 at 50MP) read noise 50.119 DN (start 9.23 saturation 15389 DN
ISO 12800 - DR 8.61 at 8MP (7.29 at 50MP) read noise 98.2089 DN (start 18.65) saturation 15395 DN
These results should be readed with slight caution. Dcraw didn't show a black point. By visual inspection of the data neither I have been able to infere it. What is even more disturbing is that exiftool doesn't found the black level tags stored in the EXIF information. A potential explanation is that Canon may have removed it: let's hope this is not the case, since this was one of the good things about Canon RAWs (and that would make very, very unhappy to astrophotograpy folks). Hope that this is due to some strange side-effect at decoding. Also note that the estimated DR would be about 0.2 stops less at ISO 100 in case the black level (not substracted in the above calculations) were 2048.
Dcraw is not able to correctly decode the full picture (it only truly saves half of the data, and using double the correct width and half the correct height). And colors are beyond wrongness.
The read noise value shown between brackets is the average read noise of the 33 horizontal rows (from 3 to 35) in the masked area at the top of the image. This value is nearly constant in Canon sensors regardless the ISO level, except when digital ISO amplification is in use (I suspect that it could be the read noise when the analog ISO amplifier is not yet turned on). Note how it doubles from ISO 6400 to ISO 12800 (9.23 -> 18.65). Yes, ISO 12800 is fake, it really doesn't exists... but what it more interesting is that ISO 6400 seems also to be fake (as if the last native analog ISO were 4000 or 5000).
p.1 #5Canon 5DS dynamic range: zero improvement (RAWs analyzed)
Never satisfied, pre-production demos, seen/used only briefly, not released into the wild for exploration, no 'real' RAW converters, yet all knowing proclaims a load of something, much like words translated/twisted about from a Canon Exec in another 'cautionary post'...really
p.1 #6Canon 5DS dynamic range: zero improvement (RAWs analyzed)
I wouldn't expect anything earth shattering in terms of DR but it does seem a bit premature. I'm more interested in real world use--banding when pushing shadows, for example.
p.1 #7Canon 5DS dynamic range: zero improvement (RAWs analyzed)
StillFingerz wrote:
Never satisfied, pre-production demos, seen/used only briefly, not released into the wild for exploration, no 'real' RAW converters, yet all knowing proclaims a load of something, much like words translated/twisted about from a Canon Exec in another 'cautionary post'...really
Even if I agree with you, do you seriously think final 5Ds will have 14 DR stops ?
Even a Canon explorer of light (you know, those always saying that every Canon product are amazing) said 5Ds DR was like 7D2 ...
Canon have done a lot for photography, pushing sensors development forward when Nikon was far behind in noise department. But now days this is totally different. I must say, anyway, that Canon now definitely produce amazing lenses! Bodies are truly great also, ergonomically, AF, etc. But sensor wise, once you have worked with some Sony sensors (not just play with a few raws), you can't imagine going back to old Canon sensor tech. I was waiting for 5Dsr for years, but shooting with A7R for more than a year now, I'm not interested anymore in Canon sensors. Sad, really, because I would like to (no more adapter, back to AF, one body for all, etc), but no-no.
Now I'm sure 5Ds/r will be amazing for a lot of folks and produce outstanding photographs. Like my 1DsII still does. But it won't be 14stops DR, be it beta or final (if only I could be wrong!).
p.1 #9Canon 5DS dynamic range: zero improvement (RAWs analyzed)
StillFingerz wrote:
Never satisfied, pre-production demos, seen/used only briefly, not released into the wild for exploration, no 'real' RAW converters, yet all knowing proclaims a load of something, much like words translated/twisted about from a Canon Exec in another 'cautionary post'...really
The data is based off the optically-masked data in the raw file. That's been a reliable method for establishing DR on all previous Canon bodies. There are some potential gotchas but the results are reasonable.
p.1 #11Canon 5DS dynamic range: zero improvement (RAWs analyzed)
It all comes down to one basic statement...if the new Canon 50mpx bodies don't do what you want then to do, then simply buy something else.
All this analysis, handringing and teeth-knashing doesn't change my mind or anyone elses for that matter. We have decided to buy these cameras. Real world production cameras in the hands of competent photographers not photographing kids soccer or high-school football tell me more about these cameras then any charts or brick wall photographs.
Regarding the analysis, how does changing the file suffix from .jpeg to .CR2 make these compressed files uncompressed and magically RAW files? I'm probably missing something.
p.1 #12Canon 5DS dynamic range: zero improvement (RAWs analyzed)
mogud wrote:
It all comes down to one basic statement.....if the new Canon 50mpx bodies don't do what you want then to do, then simply buy something else.
All this analysis, handringing and teeth-knashing doesn't change my mind or anyone elses for that matter, who have decided to buy these cameras. Real world production cameras in the hands of competent photographers not photographing kids soccer or high-school football tell me more about these cameras then any charts or brick wall photographs.
Regarding the analysis, how does changing the fil suffix from .jpg to .CR2 make these compressed files to uncompressed and magically RAW files? I'm probably missing something. ...Show more →
They must already be on the server, but not advertised.
p.1 #15Canon 5DS dynamic range: zero improvement (RAWs analyzed)
KiboOst wrote: StillFingerz wroteNever satisfied, pre-production demos, seen/used only briefly, not released into the wild for exploration, no 'real' RAW converters, yet all knowing proclaims a load of something, much like words translated/twisted about from a Canon Exec in another 'cautionary post'...really Even if I agree with you, do you seriously think final 5Ds will have 14 DR stops ?
Even a Canon explorer of light (you know, those always saying that every Canon product are amazing) said 5Ds DR was like 7D2 ...
Canon have done a lot for photography, pushing sensors development forward when Nikon was far behind in noise department. But now days this is totally different. I must say, anyway, that Canon now definitely produce amazing lenses! Bodies are truly great also, ergonomically, AF, etc. But sensor wise, once you have worked with some Sony sensors (not just play with a few raws), you can't imagine going back to old Canon sensor tech. I was waiting for 5Dsr for years, but shooting with A7R for more than a year now, I'm not interested anymore in Canon sensors. Sad, really, because I would like to (no more adapter, back to AF, one body for all, etc), but no-no.
Now I'm sure 5Ds/r will be amazing for a lot of folks and produce outstanding photographs. Like my 1DsII still does. But it won't be 14stops DR, be it beta or final (if only I could be wrong!).
Canon will as it was before give you the advantage of best AF compatibility with EF lenses. From what I am reading so far including in this post, it sounds like that even the 36 MP Sony sensor maintains its advantage of more DR stops even compared to Canon's latest FF sensors. One might be more or less important for different kind of shooting styles. I personally rather sacrifice in AF and get more DR in my RAW files.
Nevertheless, we need to wait for the final 1:1 comparison between both sensors or even with the rumored new Sony FF sensor. But it sounds more and more likely that the DR will remain Canon's Achilles heel. But at least they got the resolution up and going....
p.1 #17Canon 5DS dynamic range: zero improvement (RAWs analyzed)
EB-1 wrote:
DR is not an issue for me, but is the noise about the same as the 7D II? Is ISO 1600 still usable?
EBH
The files will be so big it won't matter, you will be enlarging so much less.
But the upshot is, this is an oversized 7D2 sensor and it won't compete with the Sony sensor if DR is your thing. Canon clearly don't think this is important enough to develop a new sensor for, which seems a bit of a shame.
But they will still go on selling more cameras then everyone else.
p.1 #18Canon 5DS dynamic range: zero improvement (RAWs analyzed)
Lucky for all of us...
1. There are great choices from Canon, Nikon, Sony and others — even extending to less-expensive mini MF systems. Cameras form any of them can produce great photographs, and we can choose whichever one we prefer or think will be ideal for us — it is mostly a matter of weighing the variables in the context of our own photography.
2. It won't be too long before we don't have to speculate about the performance of the 5Ds cameras, since they will likely be out in a few months. Between now and then we will likely see more concrete data about their performance, and speculation will be less and less necessary.
Of course, none of that will end the Forumtography controversies that seem to inevitably be pursued with such extraordinary passion! ;-)
p.1 #19Canon 5DS dynamic range: zero improvement (RAWs analyzed)
cgarcia, thank you for the analysis!
The saturation vaule at ISO 100 compared to the 7D II indicates either 1/3 stops higher Full Well Capacity or 1/3 stops denser Color Filter Array (most likely), or a combination.