Question about file size and noise

shuttersteph

So, some cameras can capture RAW files at the cameras highest resolution and also you can set it to a smaller RAW file. If you set it to a smaller size RAW file than full resolution, does that mean that there will be less noise on a smaller RAW file? I hope my question makes sense.

Icebear

I don't know from Canon, but I think you're referring to "bit depth" and "compression." Nikon offers "uncompressed" "lossless compressed" and "compressed" of either 12 bit or 14 bit NEF (raw) files.

They describe the "lossless compressed" as "no effect on image quality" and the "compressed" as "almost no effect on image quality."

Obviously a 14 bit file is going to be larger and have more color data than 12 bit files, regardless of the compression algorithm you choose.

If you want absolutely the most data available for processing, you want to avoid 12 bit and "compressed" files, but many, many respectible photographers will tell you that there's virtually no humanly discernable difference anyway.

As far as "noise" I don't think that's an issue at all with respect to compression or bit depth.

Hope this helps. Oh, at the risk of stating the obvious, none of these parameters have any effect on image size. A 6, 10, 12 or 18MP image is just that. I think you are misusing the term "resolution." But maybe Canon has a feature I'm not familiar with. Prolly . . .

ziggy53

There has been some testing of Canon cameras that have sRAW files and yes, there is some apparent binning of the pixels and yes, that does reduce effective random noise. Unfortunately, it appears that a good noise reduction software can do much better with the full resolution file. Also, a reduced resolution file has less real detail.

I believe the real reason for sRAW is to allow more files per storage card when there is no need for the extra resolution of the full-res RAW files but still a need for maximum dynamic range and full RAW processing. An ideal application might be wedding and event candids, which rarely get printed very large.

I like the discussion and tests here:

http://www.cambridgeincolour.com/forums/thread1277.htm

Icebear

So I guess there really IS a difference in Canon & Nikon "treatment" of raw files? I thought I wuz jokin'.

shuttersteph

thanks so much!

BenA2

7D mRaw and sRaw lose dynamic range

ziggy53 wrote: »

I believe the real reason for sRAW is to allow more files per storage card when there is no need for the extra resolution of the full-res RAW files but still a need for maximum dynamic range and full RAW processing. An ideal application might be wedding and event candids, which rarely get printed very large.

Oh, how I wish that were true in practice. When I bought my 7D I was super excited about shooting in mRaw, assuming I'd get all the flexibility of RAW with smaller file sizes, when I didn't need the high resolution. Unfortunately, it didn't work out that way for the 7D. In actual practice, Canon's 7D mRaw and sRaw formulas lose dynamic range. I believe I read somewhere, I think it was on Cinema5D, that this is a result of some sort of odd conversion to and interpolation of chrominance data--I'm not going to pretend to understand it. But, I've certainly experienced it.

I've found that 7D mRaw files have significantly less "overhead" for highlight recovery than full RAW files. So, I did not have anywhere near the exposure latitude I was hoping for with mRaw. I'd say I usually have about 1.5-2 stops of highlight recovery in RAW, but only about 0.5 stops in mRaw. So, you really need to nail those mRaw exposures, but even when you do, I still feel like the files are less tolerant of large RAW conversion adjustments. To me, the 7D mRaw and sRaw files are like high-bit-depth JPEGs you can change the white balance of.