Thread: D700 ISO 6400 or 3200 -1EV

> > Problem with your theory is, I was sholwing a capture from CC24 that clearly shows 14 bit being superior. > All it shows is gaps in a histogram. The gaps do not necessarily translate into image degradation. Enough is enough.

georgeRC wrote: > The argument that 12 bits are just as good as 14 bits as long as not all 12 bits are zeroes is mere wishful thinking. The results from less data are noise and posterization. Iliah has a good handle on this. Please see the light (pun intended). > The noise *in* the light (aka photon shot noise) as well as in the camera electronics masks the effects of reduced bit depth, an effect called /dithering/ (look it up...

John Motts wrote: > Fact is, in Adobe software you should set the ISO correctly, not compensate afterwards. Of the shots I have just done, by far the best one in every respect (noise, shadow detail, contrast, everything) was the 6400 one. Quality deteriorated substantially at each step to the point where the 800 ISO was so noisy and contrasty as to be a long way from being useable. > Okay I'm going to correct myself a bit...

Brad Morris wrote: > Your argument doesn't make sense to me. > > ISO 6400 and 3200 with -1EV will provide exactly the same exposure of light on the silicon sensor. The processing downstream from the sensor applies corrections based on camera settings from the given volume of data created at the sensor by the quantity of exposure to light. Given that exposure is identical, S/N will be identical in the raw data cause...

I'm not quite sure what to make of it, but I suspect that that data is helping color depth (probably not the right term). Would you be kind enough to tell me raw converter that is?

Here I have used the technique with two separate raw conversions (one for the highlights and one for the shadows) blended in Photoshop using a layermask and a brush. The ISO 25600 original file have blown highlights since the scene is mostly lit by a candle light from the left and the exposure was too generous on purpose for this test. The crop is marked red. http://farm3.static.flickr.com/2516/4090782170_658a7a7f37_b.jpg A ISO 3200 shot with...

Raul wrote: > > Marianne Oelund wrote: > > D300 12-bit will live forever in infamy, I suppose. > > Kindly feed me more on this: what do you mean. In 12-bit RAW mode, the D300 black-clips the signal by a couple of levels or so. It behaves better in 14-bit mode, so if you're trying to recover detail in very deep shadows, there is an advantage to using 14-bit on this particular camera. For most...

Iliah Borg wrote: > > My question is: have you compared results by RPP, ACR and CNX2 re HLs recovery (when clipping occurs in RAW)? > > ACR is best in reconstructing highlights IMHO, and the easiest by far. On certain occasions I process in RPP all but highlights and blend reconstracted highlights from ACR. It is possible to add better HL recovery to RPP - just a matter of time. However, blown out highlights are...

The argument that 12 bits are just as good as 14 bits as long as not all 12 bits are zeroes is mere wishful thinking. The results from less data are noise and posterization. Iliah has a good handle on this. Please see the light (pun intended). Iliah Borg wrote: > > > Problem with your theory is, I was sholwing a capture from CC24 that clearly shows 14 bit being superior. > > > All it shows is gaps in...

and clear :) >Marianne Oelund wrote: