Shulou(Shulou.com)11/24 Report--

With the official release of Sony α7RV, many netizens have recently consulted me about Sony camera "lossless compression (M)" and "lossless compression (S)"RAW questions, today to make a unified answer.

Most cameras use RGB sensors (ignoring black and white sensors, X3 sensors, etc.), and each pixel grid can only be red, blue, or green. A single pixel does not have complete color information, and it needs to be combined with surrounding pixels, that is, through the "inverse Bayer algorithm" to get the color photos we see.

RAW stores data for each pixel on the sensor. If the software extracts data directly from RAW, the picture is black and white and mosaic. What color the final photo presents is determined by both camera data (RGB filter parameters) and white balance-so true RAW can achieve lossless, wide-range white balance adjustments.

Sony α7IV, α 7RV and α1 cameras provide lossless compression RAW in non-full size. For example, α7RV has about 60 million pixels, M format has about 26 million pixels, and S format has about 15 million pixels. Here is the direct conclusion: non-full pixel RAW is not true RAW.

Full frame:

Lossless compression (L)--is true raw

Lossless compression (M)--not true raw

Lossless compression (S)--not true raw

APS-C CUT:

Lossless compression (M)--true raw (i.e. full-width clipping)

Lossless compression (S)--not true raw

Why am I so sure? Because this kind of non-full pixel RAW is really not new. Canon and Nikon have both been offered on digital SLRs, essentially RGB images (similar to TIFF) after inverse Bayer and thumbnail. There are two obvious shortcomings:

① Because the color is fixed during conversion, the white balance adjustment ability is greatly reduced.

② File size did not decrease year-on-year

Here, ET uses its own α7IV to make an example for everyone: under exactly the same parameters, choose lossless compression L, lossless compression M, compressed (lossy compression) to take 3 photos. Then in Lightroom, set each photo to 2800K, 5500K, 20000K and then output. It can be seen that lossless compression L and compressed after a large adjustment of white balance, the picture does not change except for color; while lossless compression M has a white condition.

In terms of file size, ET has also been measured before: Sony α1 lossless compression M effective pixel is about 42%, file size is about 56%, lossless compression S effective pixel is about 25%, file size is about 50%. Canon, Nikon situation is similar, we can find the answer in the specification table of specific products.

(Canon 5D4) Perhaps because miniatures consume extra processor resources, the Nikon D850 uses non-full size RAW for continuous shooting, and the number of buffers (number of continuous shots) is reduced.

(Nikon D850) In simple terms, ET considers the "loss" of RAW that is not full size greater than the "gain." In fact, Canon, Nikon's current full-frame micro-single camera does not provide this kind of non-full size RAW. If you really care about storage space, you might as well choose lossy compression RAW -although lossy compression, at least it is true RAW.

This article comes from Weixin Official Accounts: Camera Notes (ID: xjbiji), by Teng Feiet

Welcome to subscribe "Shulou Technology Information " to get latest news, interesting things and hot topics in the IT industry, and controls the hottest and latest Internet news, technology news and IT industry trends.