Why Rely Solely on DxOMark ?

I have been measuring sensor characteristics since before DxOMark even existed.
One of my tests, Photographic Dynamic Range (PDR), is analogous to DxOMark Landscape Dynamic Range (values differ my a fixed amount).
Two others give data that you can't get there, like read noise and Fixed Pattern Noise (FPN).

All it takes is for you to cooperate with me in gathering certain raw images.
(The number varies depending on how many ISO settings your camera has.)

For Samsung I only have data for the NX1 and the NX10.
(In contrast I have every Nikon DSLR and many Canons.)
So if you own any other model and would like to "pitch in" send me email (at the bottom of my site) or a Private Message (PM).

TIA,

Matt Everglade wrote:

What exactly do we need to send you? I'd like to add an NX30.

I need various specific raw (SRW) files to run through my statistical analysis.

I'll Private Message (PM) you to get the ball rolling.

Thanks!

bclaff wrote:

I have been measuring sensor characteristics since before DxOMark even existed.
One of my tests, Photographic Dynamic Range (PDR), is analogous to DxOMark Landscape Dynamic Range (values differ my a fixed amount).
Two others give data that you can't get there, like read noise and Fixed Pattern Noise (FPN).

All it takes is for you to cooperate with me in gathering certain raw images.
(The number varies depending on how many ISO settings your camera has.)

For Samsung I only have data for the NX1 and the NX10.
(In contrast I have every Nikon DSLR and many Canons.)
So if you own any other model and would like to "pitch in" send me email (at the bottom of my site) or a Private Message (PM).

TIA,

So far I've added the NX1000 and have some inconclusive NX3000 data.

Interactive Chart

Why bother at all? I just take photos.

viking79 wrote:

DxOMark has a level of consistency in results that I haven't seen elsewhere.

I could argue that my results are even more consistent than those at DxOMark; but not in any important way.
My results serve as a valuable "sanity check" for some people.
I test some cameras that they don't.
I test intermediate ISO settings, when available, which they don't.

They perform tests that I don't, and I perform tests that they don't.
I provide some of their data (a la sensorgen) at my site along with my data because both are useful.

Do you account for discrepancies between ISO values indicated by manufacturer, and actual ISO value? I.e., if you don't it could shift the curve left or right and it appears that is happening with cameras like the E-M5, which doesn't really have the same dynamic range at a photograph level as a D3x.

I plot versus ISO Setting rather than "Measured" ISO.
The discrepancies you talk about are normally quite small.
And in any case, dynamic range is not increased/decreased, the midpoint is shifted between how much of the dynamic range is "highlight" versus how much is "shadow".

However, if you are satisfied with DxOMark as the sole provider of sensor characteristics that's OK with me!

sensor characteristics and intellectual reference is good, although it has becoming more of a trend that some people use these information into thinking that it would make their pictures much better. it is partly true, although in the end, the key is post-processing workflow and photographic skills.

Massao wrote:

I've got two out dated samsung. What kind and how many photos are we talking about?

There are three separate tests. Doing one test doesn't mean having to do the others.

All require full resolution raw files. Most people deliver files to my Dropbox.

Two tests are a particular shot at each ISO setting (including intermediate and extended).
So the number of shots depends on the camera (could be 5 could be 25).

The third test is 2 sets of 16 (easy) shots.

Send email or Private Message (PM) if you are interested.

Regards,