sRAW Study: WB and Recovery

With the introduction of the Nikon D4S and the D810, Nikon introduced the new sRAW format for saving images. While we have already explained the format in detail in our sRAW format explained article, there were many follow-up questions from our readers, some of whom asked us to provide some image samples from RAW, sRAW and JPEG formats to compare things like white balance recovery and highlight / shadow recovery. In this article, we will explore the sRAW format in detail and show sample images from both controlled lab and outdoor environments, demonstrating what sRAW is capable of delivering when compared to the regular RAW format.

First, let’s take a look at how capable the sRAW format is at recovering white balance. For the first test, we photographed a color chart with the D810 in proper white balance, then changed the WB setting to 2500K (cool) and 10000K (warm). The below comparisons demonstrate the results in both sRAW and JPEG formats to show their differences compared to regular RAW files.

sRAW White Balance Recovery: Controlled Environment

Here is a crop from a color chart shot with correct white balance:

Nikon D810 RAW

NIKON D810 + 50mm f/1.4 @ 50mm, ISO 64, 16/10, f/5.6

We will be using the above as a reference to compare to colors produced by sRAW and JPEG. Below is the output of sRAW, shot at 2500K and then recovered to 4900K, which is the correct WB (Left: sRAW, Right: JPEG)

Nikon D810 sRAW 2500K Recovered -0.7 EV Nikon D810 JPEG 10000K Recovered

Although the difference between the first RAW and sRAW files is pretty small, a slight change in colors is definitely there. In addition, there is a drastic change in exposure – when I changed the WB to 4900 in Lightroom, the sRAW image appeared visibly brighter. In order to make a proper comparison, I had to decrease the exposure on the sRAW file by -0.7 stops (explained why further down below). Still, it is pretty clear that sRAW contains more data than JPEG, since the color output of JPEG appears drastically different when recovered similarly from 2500K.

Take a look at the below crops that compare the other end of the color spectrum, at 10000K:

Nikon D810 sRAW 10000K Recovered -0.5 EV Nikon D810 JPEG 2500K Recovered

For this comparison, I had to adjust sRAW by -0.5 stops to make it comparable. Again, sRAW demonstrates much better white balance recovery than JPEG. Take a look at what happened to white and gray shades of color – they appear blue. Many other colors are not reproduced accurately either.

If you compare the RAW and sRAW files, they appear to yield similar colors, although sRAW definitely renders the colors a bit differently due to loss of colors. Remember, we are dealing with 11-bit data compared to 14-bit, so there is a definite loss of data. Comparing sRAW and JPEG, is 11-bit vs 8-bit, which is approximately 8.6 billion colors compared to only 16.8 million.

Still, sRAW is drastically better than shooting in JPEG format, as the above examples demonstrate.

How does this apply to real world shooting? Let’s take a look at the below outdoor samples for a comparison.

sRAW White Balance Recovery: Outdoor Environment

For this study, we will be using the below image as a reference, with white balance set to 5600K:

Nikon D810 Normal Exposure

NIKON D810 + 24mm f/1.4 @ 24mm, ISO 100, 1/100, f/5.6

Now let’s take a look at how it compares to sRAW and JPEG, both of which were set to 2500K, then recovered to 5600K in Lightroom (Left: sRAW, Right: JPEG):

Nikon D810 sRAW 2500K Recovered Nikon D810 JPEG 2500K Recovered

Although the first indoor test showed some differences in color, the above samples demonstrate what happens to both sRAW and JPEG images when wrong white balance is used. Although I tried everything I could to imitate the JPEG output to the original RAW file, none of my attempts resulted in correct rendering of blues and greens in the above scene. It is a given that JPEG is a poor choice when compared to either sRAW or RAW file formats. Now if you compare RAW to sRAW, there is still a visible change due to loss of colors. Pay attention to the change of colors and shades in the front two trees and some greens – they appear warmer on sRAW files, which is false color.

Now let’s take a look at what happens at 10000K, recovered back to 5600K:

Nikon D810 sRAW 10000K Recovered Nikon D810 JPEG 10000K Recovered

Now we see sRAW rendering colors a bit cooler in comparison, which is also the result of loss of colors. The difference is not dramatic, but it is definitely there. The warmer colors were easier to recover for the JPEG format, so it does not appear as bad as before.

sRAW Shadow and Highlight Recovery: Adobe Camera RAW

The above sample images demonstrate that the sRAW format preserves many more colors than the JPEG format, which is definitely good news for those that consider using this format. We definitely see loss of colors when compared to the RAW format, but it is not as bad as JPEG. But what happens when one tries to actually recover both highlight and shadow details? Whether one is a landscape, architecture or lifestyle photographer, being able to recover images is important, especially when shooting in high contrast scenes. Let’s see what happens to images when they are over-exposed by four stops, then recovered in post. Here is a comparison of the three formats (Left: RAW, -4 EV Recovery, Right: sRAW, -4 EV Recovery, Bottom: JPEG, -4 EV Recovery):

Nikon D810 -4 EV RAW Recovery Nikon D810 -4 EV sRAW Recovery Nikon D810 -4 EV JPEG Recovery

Now this is where things get very interesting. The difference in what a 14-bit RAW file can yield compared to an sRAW appears to be drastic when recovering highlights. The JPEG format is obviously the worst here, as it cannot even deal with such a change in exposure, which ends up simply darkening the blown out highlights.

Here is what happens if we push the exposure by -3 stops, then attempt to recover:

Nikon D810 -3 EV RAW Recovery Nikon D810 -3 EV sRAW Recovery Nikon D810 -3 EV JPEG Recovery

The RAW file only loses data in some areas that are exposed beyond 4-5 stops, but most of the colors and details are there. Now compare the result to sRAW – the difference is again very noticeable. And JPEG is yet again not able to deal with such changes in exposure, resulting in a darker image. Here, one could argue that the JPEG output actually looks better than sRAW, particularly when comparing the grass area close to the center of the frame.

Now let’s see what happens if we underexpose images heavily by -5 stops, then attempt to recover the data in post:

Nikon D810 +5 EV RAW Recovery Nikon D810 +5 EV sRAW Recovery Nikon D810 +5 EV JPEG Recovery

This is where I got confused when looking at the above shots. I could not believe that JPEG actually looked better than sRAW! At first, I thought that I did something wrong on the camera, so I went back and re-shot the scene. When I compared my first attempt with the second, the results were the same. I then tried to recover the same details, but this time in Nikon’s Capture NX-D to see if anything would come out any different. Below is what I found out.

sRAW Shadow and Highlight Recovery: Nikon Capture NX-D

When I compared how Nikon’s Capture NX-D software rendered the same image, I was very surprised – the result was very different compared to what ACR did. Take a look at the same -4 EV comparison, this time rendered by Capture NX-D:

Nikon D810 -4 EV RAW Recovery Nikon D810 -4 EV sRAW Recovery NX-D Nikon D810 -4 EV JPEG Recovery

The sRAW format is visibly worse than RAW, but not as bad as it was before, which shows that Adobe Camera RAW is simply not ready to work with this format. Capture NX-D renders the sRAW file much better and the difference is quite obvious.

What about underexposing by five stops and recovering? See the results for yourself:

Nikon D810 +5 EV RAW Recovery Nikon D810 +5 EV sRAW Recovery NX-D Nikon D810 +5 EV JPEG Recovery

Looks much better! Now the output is very comparable to that of the JPEG file. When looking at pixel-level data, the sRAW and JPEG contain visibly more noise than RAW files.


This was an interesting study, because it showed what sRAW can and cannot do when compared to RAW and JPEG files. If your intention is to be able to recover white balance, the sRAW format has a definite advantage over JPEG as demonstrated in the first part of the article. Although there is definite loss of colors, you can recover most of the data when altering white balance, which is good news. Hence, if you want to have smaller RAW files and have the flexibility to change white balance, the sRAW format seems to be a viable option.

However, when it comes to recovering information from shadows and highlights, the sRAW format must be used with caution. First, you should not be using Adobe Camera RAW (Photoshop and Lightroom) for sRAW file conversion, since the Adobe RAW engine ends up heavily under-exposing images and losing more data in highlights, as demonstrated above. Capture NX-D clearly does a better job, but the software is very buggy and has its own set of problems (for example, recovering a normal 14-bit RAW file in ACR yields better results compared to NX-D). Second, you need to carefully evaluate the above image samples and decide if you are ready to potentially lose some of that valuable data that can be vital when recovering shadow and highlight details. The sRAW format seems to be somewhere in-between JPEG and RAW – it is not as bad, but definitely not anywhere as good as what a RAW file can yield.


  1. 1) mark
    August 3, 2014 at 8:40 pm

    Do you think Adobe is working on this? I have no desire to work in NX-D. Is there a difference in how NX-D is rendering full RAW and ACR?

    • August 3, 2014 at 9:01 pm

      Mark, it is surely a disappointing issue. Adobe ACR renders regular RAW files better for highlight recovery, but NX-D is better for everything else. Capture NX-D sucks, even when compared to NX2…very sad that Nikon went that way instead of continuing to work on NX2. Hopefully Adobe will fix these issues soon, although it is hard to say if they ever will, since the latest release of CR and LR 5.6 is supposed to be “final”.

      • 1.1.1) Baf
        August 3, 2014 at 11:21 pm

        Nasim, Final? And what about new cameras and features? Would they add this functions to the other product?

        • Profile photo of Nasim Mansurov Nasim Mansurov
          August 3, 2014 at 11:41 pm

          Baf, no, by final I meant that it is not a release candidate anymore. There will be further versions of Lightroom in the future.

          • Profile photo of Mark Pitsilos Mark Pitsilos
            August 4, 2014 at 3:27 am

            Am I the only one who finds the D810 color profiles in LR 5.6 terrible?

            I switched to Camera Standard and the results were hideous, especially in the out of focus areas. Using Adobe sRGB for the time being.

  2. 2) rauck
    August 3, 2014 at 9:48 pm

    Well, another excellent review of a new feature on Nikon cameras that other review sites just gloss over. Nasim, your site is rapidly becoming one of the best photo resources on the web. Keep it up.

    • August 3, 2014 at 9:58 pm

      Thank you for your feedback Rauck, I really appreciate it!

  3. 3) mark
    August 3, 2014 at 9:50 pm

    What about shooting in RAW and converting to TIFF and then go to ACR? Shooting in TIFF is not an option as it renders an 8 bit file. And the TIFF conversion files are huge. I heard some people have learned to compress but I have yet to see anyone come up with a work-around.

    • August 3, 2014 at 10:02 pm

      Mark, that’s a lot of work, very large files (twice in size for sRAW) and just too much unnecessary headache… If I did not want to go the NX-D route, then the best thing to do at this point is to avoid sRAW.

  4. 4) Harsha
    August 3, 2014 at 10:09 pm

    Thanks for these awesome recent D810 DR related articles.

    On the same DR theme, when you get a chance, it will be great if you can clarify if the supposed DR loss in the 1.2 and 1.5 Crop modes is true –

    It will be particularly useful for people like me who will be debating between the 7fps of D810 in DX mode vs the rumored D9300.

    Thanks again for these informative articles.

    • 4.1) mark
      August 3, 2014 at 10:23 pm

      What’s a D9300?

      • 4.1.1) Harsha
        August 3, 2014 at 10:28 pm

        Supposed successor to the D300s coming out around Photokina time I believe –

        • mark
          August 4, 2014 at 7:22 am

          It’s a crop factor camera ? I am thinking crop factor cameras are going away with decreasing costs of the full frame. I see no manufacture benefitting from two complete lines. Consolidation makes sense and all around mirrorless technology.

    • 4.2) Harsha
      August 3, 2014 at 10:40 pm

      Another inconclusive FX vs DX DR discussion-,Ideal_DX,D800,Ideal_FX

      Like you clarified sRAW format in layman terms in one of your articles, it will be awesome if you you can clarify D810 cropped mode DR in layman terms.

    • August 3, 2014 at 11:44 pm

      Harsha, I guess it comes down to how you evaluate DR. If you look at pixel level, there is no difference in DR between FX and 1.2x or 1.5x crops. But if you down-sample like DxO does, then there will be a difference in both SNR and DR. I would not be as concerned about DR as I would be about SNR though when reducing resolution by cropping.

  5. 5) Lee
    August 3, 2014 at 11:07 pm

    The bottomline of this is to shoot in native RAW format so as to maintain that flexibility in both over/underexposure and WB issues. Memory cards and hard drives are much cheaper nowadays and using sRAW to save storage space doesn’t really seem attractive anymore.

  6. 6) Anirban Basu
    August 4, 2014 at 12:12 am

    Excellent comprehensive test to understand the sRaw functionality. I like to know as how the DR of D810 behaves with ISO.
    You have posted one with ISO64 & ISO 100 & practically there is no difference but what about ISO 640 &800. I know that D800 & D800E DR has reduced drastically after ISO 640/800. What is the break even ISO for D810 then.
    Thanking you in advance.

  7. 7) autofocusross
    August 4, 2014 at 12:54 am

    Excellent article Nasim, shows us the strengths and weaknesses of both sRAW and current software. There is one thing I would ask is can you do a sideways test on sRAW using DXO Optics Pro. I find it works wonders on any images I need to ‘recover’ from wrong exposure or WB settings. I never use AWB so sometimes if you are busy shooting the light can change without you noticing, and you end up slightly hot or cool on WB.

    Optics Pro works wonders, so much so that I rarely use ACR, so it would be interesting to see how that sofware copes with sRAW.

    Thanks again for a very interesting article, thought provoking as usual!

    • 7.1) AlanC
      August 4, 2014 at 8:27 am

      Yes please!

      I find DXO a really great RAW converter and I use it all the time. I find L/R impossibly tedious in the way it needs to handle my files.

      The biggest (only?) drawback to DXO, as mentioned earlier, is the large TIFF files it can produce.

      Another great article Nasim.

  8. 8) Michael Foley
    August 4, 2014 at 1:09 am

    What a terrific article. So informative.

  9. 9) Betty
    August 4, 2014 at 5:35 am

    I am a bit baffled why some people buy sophisticated, high quality cameras like the D800/810 etc. and then because they have limited storage or want ‘speed and convenience’ shoot compressed lower quality formats like sRAW or JPEG and then complain about the inevitable negative trade offs in image quality.
    If you want quality, shoot RAW with the best camera you can afford.

  10. 10) Anders
    August 4, 2014 at 11:54 am

    Thanks for a very informative article.

    I have been testing sRaw today and the output looks nice in good light. The size of the sRaw files were all around 27-28 MB versus 37-42 MB for compressed lossless raw, so this is a space savings of only 10 MB per image.

    I simply can’t see the advantage of this format when taking into acount the small space savings and that the size is half (3680 * x 2456) of normal raw and there is less possibility for recovery. So personally I’m not going to use this format at all.

    • 10.1) mark
      August 4, 2014 at 2:21 pm

      It is some what significant. Thats a 23-27% reduction. Many of my files in working in Photoshop can be as large as 233 MB, depending on what you do such a panoramas or focus stacking as large as 6 GB.So while taking 10 MB off the raw may not ne much can be ac couple of hundred MB in post production.

      • 10.1.1) Anders
        August 4, 2014 at 2:30 pm

        Sure, I can see in those cases, when you add up many images that it can be a significant difference.

  11. 11) David
    August 5, 2014 at 10:53 pm

    Very interesting article. I’m curious as whether this applies to 12-bit RAWs as well as both files are very close in size.

    It would be awesome if you could do a follow-up article with 12-bit RAWs. Come to think of it, with high dynamic range cameras like the D8xxs, it may be time to write a 12-bit vs 14-bit RAW comparison article.

  12. 12) Karen Grigoryan
    August 6, 2014 at 6:45 pm

    Yes! This is what I was waiting for!!! Thank you Nasim! :-)

    How about buffer capacity with sRAW? How many sRAW files can fit there in burst?

  13. 13) Sverre
    August 7, 2014 at 1:27 am

    The results is very surprising compared to what i have seen comparing RAW with lossy DNG’s. It will be very interesting to do a similar test with lossy DNG to see if the problem is the technology Nikon has used in SRAW or bad implementation in the RAW converters.

    I have used lossy DNG’s for my snapshots and find it very convenient. I don’t save any space in camera and need to convert the files (start batch and do something else). But I save a lot of space and Lightroom is much faster.

  14. 14) Kathy Adams Clark
    August 7, 2014 at 3:01 pm

    Thanks for the interesting article, Nasim. We had some disappointing images on a recent photo tour due to the person using sRAW in a Canon 70D. The images can be processed to look okay but they are not as good as they could be. I’m going to suggest people stay away from sRAW for right now.

Comment Policy: Although our team at Photography Life encourages all readers to actively participate in discussions, we reserve the right to delete / modify any content that does not comply with our Code of Conduct, or do not meet the high editorial standards of the published material.

Leave a Comment

Your email address will not be published. Required fields are marked *