Sony A6000 raw files

[Greg D]

Hi folks.

Does anyone know if the Sony A6000 uses a compressed raw file like the A7 series?  And if so, what would be the implications of that for post-processing?

Thanks for any insights.

-Greg

[ErikKaffehr]

Hi,

Probably they do and there are no consequences for raw processing. There is actually two different kind of compression, one reduces the number of bits per pixel, the other does a block based compression that can cause artefacts in cituations with very sharp gradients.

I would be happier without the second type of compression, but I would guess it is quite harmless.

Jim Kasson has analysed it in detail, I can dig up his findings.

Best regards
Erik

Hi folks.

Does anyone know if the Sony A6000 uses a compressed raw file like the A7 series?  And if so, what would be the implications of that for post-processing?

Thanks for any insights.

-Greg

[Greg D]

...the other does a block based compression that can cause artefacts in cituations with very sharp gradients.

This is what I'm concerned about, I think - banding in skies, etc, which is enough of a problem already!

[ErikKaffehr]

Hi,

You are not going to get any banding in the skies, but you can get artefacts on star tracks in night exposures. The compression used is lossless if there are now extreme gradients in the image. In a sky there are essentially no gradients.



You can also check this posting by Jim Kasson: http://blog.kasson.com/?m=20140217

Best regards
Erik

This is what I'm concerned about, I think - banding in skies, etc, which is enough of a problem already!

[Greg D]

Thanks for this.  The sort of banding (posterization) I'm concerned with is what I see sometimes in solid blue skies that transition gradually from lighter to darker through the frame.  I see it sometimes to a small degree in raw files, but mostly I see it when I convert to jpeg.  So that's why I worry when I see the word "compression".  But apparently not something to be concerned with.

Thanks!

[ErikKaffehr]

Hi,

That kind of banding depends on going from 16 bit data to eight bit data and possibly a less than well calibrated screen. It is really about mapping a small difference in a 16 bit space to a significant difference in an 8 bit space.

This is a reason that Adobe RGB is not recommended for JPEGs in many cases, small changes in data can make a great difference on screen. So if your colour fits within sRGB, there will be no gain in using a wider colour space.

There are a lot of different LDC technologies, TN is commonly used on low end but has often just 6 bits colour space. Color calibration may affect the gamma curve and excessive manipulation may lead to visible posterisation.

I have never seen posterisation on my Eizo FlexScan 241W that use PVA technogy (I think) but often on LCD projectors, just as an example.

Regarding the Sony compression, I would say that Lloyd makes a bit to much noise about it, but I would very clearly prefer to be without it, at least regarding the Delta compression. I feel it is a decently good designed code, but something I would prefer to be without. But I would not hurry to blame all issues on it.

Best regards
Erik

Thanks for this.  The sort of banding (posterization) I'm concerned with is what I see sometimes in solid blue skies that transition gradually from lighter to darker through the frame.  I see it sometimes to a small degree in raw files, but mostly I see it when I convert to jpeg.  So that's why I worry when I see the word "compression".  But apparently not something to be concerned with.

Thanks!

[Tuolumne]

Lloyd Chambers disagrees over the significance and effect of the lossy compression on Sony cameras.  Unfortunately, the more detailed discussion, histogram with periodic gapping, and photo example of subtle banding in the sky are all behind a paywall.  I don't have the technical chops to evaluate this issue but I thought I'd point out a dissenting point of view.