GordonBGood
Veteran Member
There are several errors in the above, as follows:DNG on the K-30 and Ricoh GR is only 12-bit while PEF from the K-5, K-5IIs, and K-3 are full 14-bit RAWs. For me, the 14-bit files are by far more desirable. With my Nikon I always shoot 14-bit uncompressed.
DxOMark when they tested the K-30 talked about the possibility that the 12-bit DNG cost the K-30 around 1 EV vs the K-5 which uses the same exact Sony Exmor sensor as the K-30 but gets over 14 EV DR vs 13 EV for the K-30.
14-bit files should also provide better color depth at least in theory. Whether it's a big enough difference to matter. But since I take the time to shoot and process RAW, I'd rather shoot 14-bit uncompressed.
DNG in theory should work with more RAW converters, but that's a moot point because DNG in every implementation I've ever used, only a 12-bit RAW file instead of the usual, industry standard 14-bit RAW.
- DNG's can be any required depth, and although the K-x/K-r/K-30/K-50/K-500 cameras output both PEF's and DNG's with a 12-bit depth, the K-5 series, the K-3, and the 645D all output 14-bit depth for both their PEF's and DNG's. There is only a (slight) benefit to the 14-bit raw files for the K-5 and K-3 for ISO's below about 400; unfortunately we have no choice in the raw bit depth as Nikon shooter do so we could save several Megabytes of card space per raw image for higher ISO's. The Nikon uncompressed format has no real benefit as to image quality from either the loss-less and the "virtually loss-less" (slightly lossy) NEF format and the main trade-off is processing time and storage space requirements.
- The sensors are not the same between the K-01, K-30, K-50, and K-500 in that they have different raw pixel dimensions and layout and those sensors can be scanned faster than the sensor from the K-5 series, which higher scanning frequency is likely the main reason for slightly less low ISO Dynamic Range (DR) than the K-5 series more than the lower bit depth.
- Raw files with a 14-bit depth do indeed have somewhat more colour accuracy, but only if one pushes low ISO files by about four EV's or more. As I said, the raw compression has nothing to do with real image quality.
- There is no industry standard 14-bit raw depth, just newer cameras are starting to more commonly use that depth or have it available, even some makes such as Canon that don't have enough DR to be able to effectively use it such that the least significant bits contain nothing except random (and non-random) noise and patterns. As I said, the DNG standard is not limited to 12-bit raw depth and can be any depth the camera manufacturer chooses to output; in the case of Adobe DNG conversions from proprietary formats, the original raw bit depth is preserved in the output DNG file, recorded as 16-bit depth of which the most significant two bits contain zeros in the case of 14 bit files. When these files are loss-lessly compressed, there is no wasted space due to the zeros as they are compressed away.
Regards, GordonBGood