I've searched on NikonGear but don't seem to find any comparisons between D2H and D2Hs iso noise comparison (RAW File).. Obviously these cameras came out when the net and blogs and forums were in their infancy.. and very few photographers seem to have bought the D2Hs! So if anyone would like to weigh in on this subject I would be very curious to know about this comparison. D2H cameras are an absolute bargain on Ebay at the moment at around $100 (obviously condintion unknown...).. But for younger photographers they would be an excellent way of getting into DSLR photography (before we all move to Mirrorless.. ;-)
So, Raw noise comparison anyone!?
And if they Raw noise is the same, then did Nikon ever do a Firmware update á la D2xs where the majority of the software improvements from the D2Hs were added to the D2H?
Thanks again
Raw noise is not the same. I co-operated with Bill Claff by supplying the various raw files for his measurements. Go see photonstophotos.net for his results and curves. I can give the short version here.
There is about half a stop advantage to D2HS in raw. We don't know why/how, but the sensor most likely remained exactly the same. The cause for the difference can be things like better quality components (amplifiers, AD converters, perhaps hand picked sensors), perhaps better heat/interference shielding to the components and maybe Nikon was able to tweak the timing and rate of the components. For example running the AD converters at a little lower clock frequency can lower the noise.
This is all guesswork, but plausible and not very expensive for Nikon to pull off.
I can't run them side by side anymore, because the D2H developed the dreaded first shot ERR syndrome. I sold it to a hobbyist who was building something that required a DSLR that needn't be perfect. I think I charged him 70 euros.
I still have the D2HS though.
It is funny how places like Fred Miranda is still full of posts where adult men shout mouths frothing that D2HS has 2 stops advantage in high ISO noise. Talk about confirmation bias. But the difference is real, albeit only 0.5 stops.