Uncompressed or Compressed raw files?

[Ian]

What are people’s opinions on shooting uncompressed files. I use a Sony A7R iii and it takes the file size from about 40 Mb to 80 Mb, with all the attendant implications. I can’t say I have noticed any difference. Interested to hear other opinions and experiences.

 

[BillW]

Is Sony’s compression lossless or lossy? If it is lossless, I think compression makes sense, unless it slows in-camera processing to your memory card. If it lossy, the question is what do you lose for the reduction in size.

Nikon has a lossless compressed option I generally use.

 

[Steve]

Sony compression is actually lossy. I think it's only a problem if you really need to pull shadows, but I'm not 100% sure. With my Sony cameras, I usually shoot uncompressed unless I need more speed (the a9ii only gives you 20 FPS in 12 bit compressed mode).

Nikon is always lossless compressed for me - no reason not to. I wish Sony would add this option...

 

[Capturingtheunseen.com]

Nikon is always lossless compressed for me - no reason not to. I wish Sony would add this option...

Steve, uncompressed gives you a larger raw file. Isn't that an advantage when cropping?

 

[DRwyoming]

Steve, uncompressed gives you a larger raw file. Isn't that an advantage when cropping?

The lossless raw compression Nikon uses gives the raw converter exactly the same pixel information bit for bit as the uncompressed raw files. There is no information lost during Nikon's lossless raw compression nor any advantage to using uncompressed raw files when shooting Nikon gear. In terms of pixels and visual information the Nikon compressed vs uncompressed are identical so there's no cropping or other IQ advantage to using the uncompressed files. Unfortunately that's not true of all brands including Sony that uses a lossy raw compression algorithm.

 

[Steve]

Steve, uncompressed gives you a larger raw file. Isn't that an advantage when cropping?

As DR says, for Nikon there is no difference.

For Sony, comprised dozens mean they are throwing away pixels, but rather bits. My limited understanding of how Sony's compression works is that it throws out non-essential data that I imagine takes the from of deep shadow noise or things like that. Data that you normally wouldn't need unless you were really pushing the file in post. That's why most of the time you can't see a difference - the important data is relatively unaffected. However, I'm not an expert on Sony compression either, so I may be off base too. Still, the one thing I do know is you end up with just as many pixels either way

 

[DRwyoming]

What's the difference than? Looks to me like everything is identical but filesize?

The only thing that is different is the intermediate file formatting and as you say, the file size.

Basically any form of lossless compression such as zip files identifies identical, repeated bit sequences in the data file and rather than literally store each string of identical bits it stores the first such string and then subsequently stores much smaller pointers back to that string of bits when it encounters them again in the file. The extraction program then reconstructs the original data stream bit for bit. That's what makes these routines 'lossless' as the extracted file matches the original file exactly.

Zip files are a great example of lossless compression and they're used to compress software for transmission all the time. In the case of software it's essential that every bit be reproduced exactly as it was in the original file without any change to the original information or the extracted software won't execute properly. Lossless compression of image files through: LZW compression, zip compression or a proprietary lossless compression/extraction scheme works the same way.

So basically with lossless compression and corresponding extraction in the raw converter you take a big file, losslessly compress it for storage on the card and uploading to the computer and then the raw converter losslessly extracts it prior to the actual raw conversion. So all that's really different is the size of the file during the storage and uploading phases.

That's very different than lossy compression algorithms like jpeg where visual information is discarded. Even then the final jpeg will have the same pixel dimensions as the original image but many of those pixels will be subtly or not so subtly changed which can noticeably degrade image quality. How much degradation depends on: the particular image, the jpeg quality setting while saving and how many times the same file is opened and re-saved in jpeg format since the lossy compression is applied each time the file is opened in an editor and re-saved. But it still doesn't change the pixel count, it just changes the information carried in those pixels.

 

[BillW]

Does lossless compression have any impact on the speed of writing to a card? Can’t say I have noticed it. But wonder if it could be faster (less information to write to card) or slower (need time for compression algorithm to run)? Maybe no practical effect.

 

[Steve]

Does lossless compression have any impact on the speed of writing to a card? Can’t say I have noticed it. But wonder if it could be faster (less information to write to card) or slower (need time for compression algorithm to run)? Maybe no practical effect.

I think it's actually better. It seems like when I'm looking at buffer capacity, it's usually higher with lossless compressed.

 

[Warren D]

Write times are a tad faster for lossless compressed. Usually doesn't matter much to me, I shoot in short bursts and XQD and CFe cards are fast. If I were a machine gunner and filled the camera's buffer, it might matter but I only hit the buffer a few times a year and only with the D850. I shoot lossless, no need not to.

 

[BillW]

Thanks Steve and Warren.

 

[Ian]

There is some good information here, glad I asked!!. The benefits of the forum and tapping the knowledge of some clever people.
In practical terms, for Sony at least, if you have your exposure correct or even close to correct there should be no advantage to uncompressed since you wont be “really pushing the file”. The only advantage therefore might be in very low light photography, with a risk of serious under exposure or very high dynamic range.

 

[Steve]

There is some good information here, glad I asked!!. The benefits of the forum and tapping the knowledge of some clever people.
In practical terms, for Sony at least, if you have your exposure correct or even close to correct there should be no advantage to uncompressed since you wont be “really pushing the file”. The only advantage therefore might be in very low light photography, with a risk of serious under exposure or very high dynamic range.

More or less. Sometimes, if you're in an extreme situation, you can have the exposure correct and still have to wrestle with both bright highlights and shadows. Still, most of the time I'm not in that scenario and have a good exposure, so I don't hesitate to use compressed if I want faster frame rates

 

[gbodave]

I'm a recent convert to the Sony A7R3 and have decided to use uncompressed RAW files. A perfect example of it being a potential benefit in situations you can't control was the other day when I photographed the Wood Duck I posted elsewhere. It kept going from a heavily shaded area of water which was black to the eye and the camera to a more open aspect where the water was much brighter from the reflection of the blue sky. There was no way I was going to be changing the RAW setting every few minutes. Mire fiddling time, less photo time.

I haven't noticed any effect when using burst mode on the medium setting.

In the UK this wide variation in lighting is common a lot of the time but especially now right through to spring.