bclaff

Lives in United States Metro-West Boston, MA, United States
Joined on Nov 4, 2006

Comments

Total: 67, showing: 1 – 20
« First‹ Previous1234Next ›Last »
In reply to:

bclaff: Does not line up with PhotonsToPhotos earlier dynamic range results.
See https://www.dpreview.com/forums/post/60514464 for some visuals and discussion.

Yes, there is a long technical reason but the short one is that DxOMark Landscape score depends exclusively on read noise and for that model of Leica (and a few others) it's very hard to get that measurement right.
Their value is high which results in too low a dynamic range.
(Not to mention that depending solely on read noise makes no photographic sense!)

Link | Posted on Dec 16, 2017 at 04:33 UTC
In reply to:

Bershatsky: DxO has been garbage for years!

https://bershatsky.com/2015/07/14/dxo-no/

@HowaboutRAW
The Canon EOS 6D Mark II DR is correct. I had similar numbers before DxOMark even tested.
There are problems though; for example the Sony ILCE-A7R2 scored lower than it should have making the ILCE-A7R3 look like more of an improvement than it is.

Link | Posted on Dec 15, 2017 at 22:04 UTC
In reply to:

Bershatsky: DxO has been garbage for years!

https://bershatsky.com/2015/07/14/dxo-no/

@JoFa
Yes, DxOMark measurement protocol is somewhat transparent but not completely so. Also I question their choice of thresholds for some metrics as well as their approach.
At PhotonsToPhotos ( www.PhotonsToPhotos.net ) I perform some similar measurements (and some additional ones) using different criteria than DxOMark.
The scoring algorithm has never been documented but I have a model (originated by DosDan) that predicts score to within 2 points.
Not that it matters, because although the underlying measurements have some merit I (and many others) put no stock in the overall score.

Link | Posted on Dec 15, 2017 at 22:02 UTC

Does not line up with PhotonsToPhotos earlier dynamic range results.
See https://www.dpreview.com/forums/post/60514464 for some visuals and discussion.

Link | Posted on Dec 15, 2017 at 21:51 UTC as 33rd comment | 2 replies

No news here. The Pentax 645Z was tested quite a while ago at PhotonsToPhotos.
( www.photonstophotos.net/Charts/PDR.htm#Pentax%20645Z )
FWIW, note that noise reduction kicks in at ISO 3200.

Link | Posted on Nov 15, 2017 at 02:16 UTC as 45th comment
In reply to:

Mais78: What is this ISO 640 thing....??

Rishi is saying to underexpose but not too much? I.e. not to use ISO lower than ISO 640 if the intention is to push shadows in post? Do I understand correctly?

The part that starts with "But there's simply no excuse to the camera's traditional ISO 6400 method of shooting ISO 6400-appropriate exposure..." is chinese to me.

@Mais78:
I wouldn't call it wrong.
He's pointing out that there's no visible advantage to using analog gain above a certain ISO setting (he says ISO 640 for the Mark III but I'd go a bit higher to perhaps ISO 1000). The approach, which I also favor, has the advantage of greatly reducing any potential loss of highlights in exchange for a minuscule loss of shadow improvement. A few manufacturers take this approach but it's not widespread; so this comment is sort of a crusade to raise awareness and get manufacturers on board.

Link | Posted on Nov 1, 2017 at 18:13 UTC
In reply to:

Mais78: What is this ISO 640 thing....??

Rishi is saying to underexpose but not too much? I.e. not to use ISO lower than ISO 640 if the intention is to push shadows in post? Do I understand correctly?

The part that starts with "But there's simply no excuse to the camera's traditional ISO 6400 method of shooting ISO 6400-appropriate exposure..." is chinese to me.

This is a dual conversion gain sensor; the low conversion gain starts at ISO 100 and the high conversion gain at ISO 640. As a result performance at ISO 500 is actually a little worse than ISO 640.
If you're going to take advantage of ISO Invariance then you would use ISO 640 rather than ISO 100 if you're going above ISO 640 lighting conditions.
FWIW, measurements actually show a slight gain to go to ISO 1000 rather than ISO 640.

Link | Posted on Nov 1, 2017 at 16:02 UTC
In reply to:

Azathothh: OMFG it exceeds the D850 at ISO100!!! BIG LETTERS. BIG ARTICLE.

And the 5dIV exceeds the D850 at ISO200 too! And at every ISO higher than ISO1273. But hey it's a Canon so who cares....

http://www.photonstophotos.net/Charts/PDR.htm

And it's also higher than the A7riii at ISO200. Oh well...

@BlueBomberTurbo
My only experience is with measurements at PhotonsToPhotos.
The variations I see mostly fall into an entirely different category (and are normally quite small).
The methodology is consistent, there is very little sample variation between cameras;.
The largest variations come from other sources like sensor temperature, etc.
These are hard to control outside a laboratory setting.
Most importantly, since the photographer using the camera has no control over these factors that level of testing accuracy is not meaningful.
So, for the measurements at PhotonsToPhotos, they are probably too accurate for photographic purposes and sometimes that leads to needless anguish over small differences.

Link | Posted on Nov 1, 2017 at 15:58 UTC
In reply to:

AngularJS: The DR is less than 12 according to photonstophotos
http://www.photonstophotos.net/Charts/PDR.htm#Nikon%20D850,Sony%20ILCE-7RM2,Sony%20ILCE-7RM3, same level as the A7r2.

@sunnycal
There's a good technical reason which involves something called a Photon Transfer Curve (PTC).
The DxOMark (and dpreview imitation) measurements are taken at the far left end of the curve where small improvements in the circuitry are measurable.
PhotonsToPhotos PDR is taken from a point not so far to the left where the appears to be no measurable difference.
This is easier to see with a graph but that's the best I can do in this reply.

Link | Posted on Nov 1, 2017 at 13:00 UTC
In reply to:

TheEulerID: I would echo the call for a proper gamma (visually lossless) compression system. If done properly, it would mean that file sizes would get smaller as ISO increases too. The idea that the only acceptable solution is computationally lossless files when all that happens is that several bits of noise are preserved defies logic.

Lossless compression is easier to implement and would make a huge difference in file size.
If they also offered visually lossy compression (similar to Nikon) that would make files a bit smaller and would just be icing on the cake.

Link | Posted on Nov 1, 2017 at 12:52 UTC
In reply to:

UllerellU: I want to see the DxO analysis of the X-trans sensors, when I had the Fujifilm X-70 I got the feeling that the dynamic range was better than the Bayer sensors I have, and have had, DxO is an interesting tool, but with the lack of Fujifilm sensors, the incomprehensible (and little credible) bad performance of the Cónon sensors, the abandonment of Pentax and the objectives of the compact cameras and the exessive attention to the smartphones, is losing a lot of credibility, at least for me.

@BlueBomberTurbo
They would not have been able to cite PhotonsToPhotos to support that claim:
http://www.photonstophotos.net/Charts/PDR.htm#FujiFilm%20X-T1,Nikon%20D4S

Link | Posted on Nov 1, 2017 at 03:28 UTC
In reply to:

Azathothh: OMFG it exceeds the D850 at ISO100!!! BIG LETTERS. BIG ARTICLE.

And the 5dIV exceeds the D850 at ISO200 too! And at every ISO higher than ISO1273. But hey it's a Canon so who cares....

http://www.photonstophotos.net/Charts/PDR.htm

And it's also higher than the A7riii at ISO200. Oh well...

Caveat emptor. Perhaps it's the readers responsibility not to agonize over small differences :-)
Putting measurement accuracy aside; I find that 1/3 stop is barely noticeable and really only start to care when differences are closer to or above 1 stop.
PDR measurements are usually quite repeatable and often in a range of about +- 1/12 stop; more than enough I think.

Link | Posted on Nov 1, 2017 at 03:17 UTC
In reply to:

Azathothh: OMFG it exceeds the D850 at ISO100!!! BIG LETTERS. BIG ARTICLE.

And the 5dIV exceeds the D850 at ISO200 too! And at every ISO higher than ISO1273. But hey it's a Canon so who cares....

http://www.photonstophotos.net/Charts/PDR.htm

And it's also higher than the A7riii at ISO200. Oh well...

@armandino
Looks perfect to me.

Link | Posted on Nov 1, 2017 at 02:37 UTC
In reply to:

UllerellU: I want to see the DxO analysis of the X-trans sensors, when I had the Fujifilm X-70 I got the feeling that the dynamic range was better than the Bayer sensors I have, and have had, DxO is an interesting tool, but with the lack of Fujifilm sensors, the incomprehensible (and little credible) bad performance of the Cónon sensors, the abandonment of Pentax and the objectives of the compact cameras and the exessive attention to the smartphones, is losing a lot of credibility, at least for me.

@UllerellU
FWIW, I test Fujifilm X-Trans at PhotonsToPhotos
Start at www.PhotonsToPhotos.net/Charts/PDR.htm
If you have a camera I haven't tested I'd be happy to work with you on it.

Link | Posted on Nov 1, 2017 at 01:27 UTC
In reply to:

Holscen: How does this compare to current and recent canon dslrs?Or is there a chart somewhere showing a dynamic range comparison across all makes? thanks

You can start here:
www.photonstophotos.net/Charts/PDR.htm

Link | Posted on Oct 31, 2017 at 18:22 UTC
In reply to:

mgrum: I don't understand how the article can claim:

"Sony's claim that the a7S had 15 EV dynamic range was patently false"

and then say

"since there's not standard for dynamic range measurement, it's hard to say whether or not anyone's claim is right or wrong - manufacturers can claim whatever they wish"

The 15-stop DR claim for the A7S is cannot be false, for exactly the reasons stated in the article - there's no standard way of measuring DR, so it's perfectly possible to come up with a way that generates the figure of 15 stops for the A7S.

I think the word-smithing is careless.
Any dynamic range claim that doesn't indicate the type of dynamic range cannot be substantiated.
Therefore it is both true and false simultaneously :-)

Link | Posted on Oct 31, 2017 at 16:32 UTC
In reply to:

AstroStan: DR chart above ranges from 13.21 to 13.78 - a whooping 4% difference! Good luck making that matter in the real world.

As in all things, the cost/benefit ratio soars at the high end (hi-fi, cars, etc.). At current prices the A7R III costs 33% more ($1k) than A7R II for a potential (but rarely realized) 4% image improvement.

But of course there are many other improvements in the III. I'm contemplating one for the better UI.

Since those are logarithms so it's actually a 48+% improvement.
But in photography we usually look at the difference in stops; a 0.57 stop improvement.

Link | Posted on Oct 31, 2017 at 16:02 UTC
In reply to:

temporaryid: This is slightly different than what DPReview collaborator Bill Claff found.
http://www.photonstophotos.net/Charts/PDR.htm#Sony%20ILCE-7RM2,Sony%20ILCE-7RM3

He found DR for the A7Riii to be near identical to the A7Rii
The article even says Bill's testing shows the A7Riii maxes out at 13.3EV well BELOW 14EV.

The preliminary PDR for the Mark III essentially matches the Mark II.
That might creep up in the final numbers but PDR is not entirely controlled by read noise, so the figure might not change.

Link | Posted on Oct 31, 2017 at 15:16 UTC
In reply to:

JochenIs: What about the iso-values? Are they measured? It doesnt look like it.

These are the ISO settings not DxOMark "Measured ISO".

Link | Posted on Oct 31, 2017 at 15:14 UTC
In reply to:

eno2: A very good article!

I wish you also test other (older) cameras as well and make a clear database with all of them.

You'll find older stuff at PhotonsToPhotos.
Check out the first section at www.PhotonsToPhotos.net

Link | Posted on Oct 31, 2017 at 15:14 UTC
Total: 67, showing: 1 – 20
« First‹ Previous1234Next ›Last »