I first came across the compressed cRAW files when I reviewed the Canon EOS M50 a few years ago. The new Canon mirrorless cameras also have cRAW available. This compressed raw file will enable you to save even more images on one memory card without loss of quality.
When I reviewed the Canon EOS M50 back in 2018, I used the cRAW file format. It is a variation of the regular raw file format, but it’s about half the file size. Using this compressed cRAW file format, you can store about twice as much images on your memory card.
Also, the new Canon EOS R5 and Canon EOS R6 have the compressed cRAW file format. This can be very handy, especially for the large files of the Canon EOS R5. For the Canon EOS R6, it might be less important, because the sensor has only 20 megapixels, producing files that are between 20 MB and 30 MB. With the cRAW file format, you end up with files between 10 MB and 16 MB.
The file menu of the Canon EOS R5. You can choose between raw and cRAW if you like. The Canon EOS R6 has the same file menu page.
Although I made a comparison between the raw and cRAW file with the Canon EOS M50, I thought it would be interesting to do the test a bit more thoroughly. I not only wanted to see if there was a difference between ISO levels, but also by correcting underexposure, overexposure, and with extreme post-processing.
For this test, I went outside and gathered some autumn leaves. I thought it would be a nice subject for this test. In order to have a highlight in the setup, I placed a LED underneath one oak leaf. There were no additional lights, just the natural window lights. I used the Canon EOS R6 with the RF 24-105mm f/4L IS lens placed on a sturdy tripod. I used the manual exposure setting and manual focus, although the latter is of lesser importance for this test.
I noticed Canon doesn’t allow you to save a raw file on one memory card and a cRAW on another. You have to choose one or the other, which is a pity. It would be great if Canon allowed saving a raw to one card and cRAW for backup to the other one.
1. Compare a Correct Exposed Image
The first test was a simple comparison with the correct exposure, both in raw and cRAW. I decided to shoot a series of images ranging from ISO 100 up to ISO 102,400. I don’t want to bother you with the whole series; thus, I chose to compare only the ISO 100 and ISO 1,600 images next to each other.
The raw images are on the left, the cRAW images are on the right.
First are the ISO 100 images with and without the crop.
Next are the ISO 1,600 images with and without the crop.
The images are straight from the camera, and I used Lightroom to produce JPEG images from both the raw and cRAW files without post-processing. Even with a large magnification, which resembles approximately 100%, the images from the compressed raw show no visible decrease in quality.
2. Comparing Three-Stop Underexposed Images
I made the same image with a deliberate three-stop underexposure. This has been corrected in Lightroom Classic by increasing the exposure slider to resemble a correct exposure. I did this with both the ISO 100 and the ISO 1,600 images to see if this exposure correction had an effect on the quality of the cRAW file compared to the raw file.
Again, the raw images are on the left, the cRAW images are on the right.
First are the ISO 100 images with and without the crop.
Next are the ISO 1,600 images with and without the crop.
3. Comparing Two-Stop Overexposed Images
After I had overexposed two stops, I have tried to correct this in Lightroom Classic. As you can see, the leave that is lit by the LED light is clipped. This part cannot be rescued in post-processing. Let’s have a look at the result, both at ISO 100 and ISO 1,600.
The raw images are on the left, the cRAW images are on the right.
First are the ISO 100 images with and without the crop.
Next are the ISO 1,600 images with and without the crop.
4.Comparing an Extremely Post-Processed Image
I wondered how the quality would hold if I did a more extreme post-processing on these images. For this comparison, I took the images that had a three-stop underexposure and corrected this in Lightroom Classic again. Besides that, I set the shadows to +100 and the highlights to -100. I also changed the black point and white point and increased texture, clarity, dehaze, and vibrance. This was done at the ISO 1,600 raw and cRAW images.
The raw images are on the left, the cRAW images are on the right. I only placed the cropped versions for this comparison.
5. Comparing a Very High ISO
Although I have showed only the ISO 100 and ISO 1,600 images so far, images weare shot up to ISO 102,400, which under normal circumstances, I find not usable for commercial purposes. With the Canon EOS R6, I wouldn’t go beyond ISO 12,800.
In post-processing, I increased exposure by 1.25 stops, decreased highlights by 100, and increased shadows by 100. I set the white point and black point, increased texture, clarity, dehaze, and vibrance. I also changed the white balance.
The raw images are on the left, the cRAW images are on the right. I have only shown the cropped versions. Remember, these are shot at ISO 12,800.
My Conclusion After This Test
By conducting this experiment, I have found cRAW to have little to no decrease in quality, despite the smaller file size. Perhaps a closer look will show some difference, especially if something like a color checker is photographed.
But I found this experiment a more realistic comparison, resembling a real-life situation. I think it is safe to say cRAW shows a negligible decrease in quality. I feel it is very safe to use the cRAW file format when shooting important events like weddings, even with extreme post-processing.
Another thing I noticed while looking at the images is the quality of the ISO 12,800 image. I think the Canon EOS R6 is very usable at this high-ISO level.
Please check the images in this article yourself and make your own conclusion. Tell me what you think. Is cRAW as usable as the regular raw files? Please share your opinion in the comments section down below. I am looking forward to your response.
If you're passionate about taking your photography to the next level but aren't sure where to dive in, check out the Well-Rounded Photographer tutorial where you can learn eight different genres of photography in one place. If you purchase it now, or any of our other tutorials, you can save a 15% by using "ARTICLE" at checkout.
Thanks for this! Really useful. I'd been curious about using it on my EOS R, given that most of my work only ends up on Social Media anyway, I'm going to make use of it a bit more now.
Ah yes, the EOS R also has craw. I cannot remember if I choose that format when reviewing that camera.
For social media purposes it won't make any difference.
Well, personally maybe I would spot the difference if I was really trying very hard.
I appreciate this post. Until just recently, I've only been able to find The Digital Picture's comparison, which reached the same conclusion as you did. It's good to have more than one opinion. I have begun using cRAW myself, because I can't see an appreciable difference between RAW and cRaw. With my EOS R, I compared a 3 stop underexposure correction, and I could just barely see a difference between the two. The RAW wasn't better, it was just slightly different. That's good enough for me.
Yes, I remember seeing that comparison on The Digital Picture.
Very clever to do some testing yourself.
I have always shot cRAW on my Canon -- I believe that Canon implements a lossless compression in their CR2 format?
Lossless compression has no loss of quality, by it's very definition! :D
I don't believe everything a manufacturer says blindly. ;)
But I am glad it is correct, as far as I can conclude from my test.
Well, to me as a computer person, "lossless compression" has a very specific and well defined meaning!
It means that after decompression, all the exact same bits are restored as you had before compression. Not a single bit difference.
Just like WinZip.
If they call it "lossless" and it is anything different, I bet that there would be a huge outcry!
It wouldn't be the first time if a manufacturer would say one thing and do the other.
:)
Lossless is not new and been on the TIFF format for ages.
It's just makes the file a bit longer to open but the quality is 100% identical.
For example a line of black pixels with 3 grey one could be coded as this in RAW
00 00 00 00 00 00 00 00 00 22 22 22 00 00 00 00 = 32 symbols
in lossless compression it will be like this
00x9 22X3 00x4 = 12 symbols
A software will decompressed this when opening the image, but there is no loss of information.
Now, I can't find any Canon statement saying that CRAW IS lossless.
So I will have to assume there is a loss of information.
At the other side, Nikon have been offering lossless compression for ages.
I have been shooting the CR3 files now for two weeks on my R5 for commercial work, and cannot see any decrease in quality. However, importing/rendering previews and exporting are much slower, but that is to be expected when working with any more compressed file type.
I did not think of that. Sounds logical; uncompressing a file takes time indeed.
When I was working on the 4000 images of the weddings I shot with the EOS R5, I was expecting a slower culling due to the 45mp files, but it worked acceptable quick.
Well, that really depends, if a compressed file is slower or faster.
A compressed file takes less time to transfer, or read from disk.
So it depends on the balance between your disk speed and CPU speed, and also on the compression applied. CR2 may be less heavy on the CPU but save less diskspace.
Well, if your disk speed is that slow, you probably wouldn't notice difference between compressed or non. I cant think of any examples where editing compressed file would be faster than a non compressed one, that dont include poor file management/storage solution practices.
Well nowadays, your CPU is almost always that much faster than your storage.
I found shooting medium or small Raw in Canon 5D series (4 in my tests) produced extremely odd colors. The image quality overall was not as mailable in post and the latitude was noticeably reduced.
I found this about after trying to find a solution to send files after a shoot and trying the smaller file format. After years and years of shooting the exact same way, the only thing I changed was the RAW size and the difference was noticeable. It is worth mentioning that it’s pretty hard to tell and perhaps not something most people will ever notice. I was surprised.
Correct me if I'm wrong, but isn't "medium" or "small" RAW about the size of the image itself, in pixels? Not about choosing the file-format, which is "Canon Raw", or "Canon Raw 2", or "Canon Raw 3" -- the latter two being compressed (like ZIP files are compressed).
I never experienced that. Just like Tim van der Leeuw says; it is just less pixels.
Although it would be a nice to test the sraw and mraw next to the raw files in my EOS 5D4. I often wondered if there would be a difference between the three. Some believe there is, some say it doesn't make a difference
Yes you are right, shooting medium or small raw looks incredibly different to my eye as well. Was super disappointed the first time I tried this. I dont know the technically details here, but essentially I think there is some aggressive pixel binning happening in camera, which is never a good thing.
Thanks for this comparison Nando! I know that memory is cheap, but when you count storing your files, then backing them up in 2 different locations, storing half the data by using cRAW would be very appreciated :)
And thank god there are still your articles on this site, because to be honest, lately the quality has been going down the drain (not even talking about the growing click-bait titles).
Thanks, Rayann. I hope I can continu to find nice things to write about. :)
a YouTuber has also test CRAW vs RAW differences, basically, CRAW only affect the extremely dark area, otherwise no much different, I am very surprised that it does not preserve more high light area at all on RAW Video: https://youtu.be/m8zrP3W8WeQ?t=617
craw is just another way to store the data. There is no other difference. That is why highlights aren't preserved. You should read my previous article about the highlight preservation settings on cameras. I did a test for that also
Hi,
I have read with hig interest your article.
As I make mostly night star phtography and panoramic night images of milky way and star trails, and since I use a Canon R5, doing the exposure stiching of several images taken as raw with the 45 megapixels sensor can be a proble, especially to stitch them together afterwords. You need a very powerful computer and still much time to build the full panoramic image.
The C-Raw seams great but I read comments that with underexposed or dark images, bring back details can be a problem with C-Raw versus Raw and for my night photography might be a very bad idea.
What is your opinion?
Thanks in advance.
Andrea
The c-raw is just a raw file that has been compressed. If you open the image in a stitching program, the file will ben decompressed. The data that will be used for the stitching is probably the same for both raw and c-raw.
It will only make a difference when storing the files itself.
I just found out if I convert R5 C-Raw to DNG the file size goes back to normal! UGH!
That is interesting. But it seems logical when I think of it. After all, craw is just a lossless compressed raw file. The DNG is not compressed
https://community.adobe.com/t5/camera-raw/dng-converter-dng-files-are-30...
I just got my new M200 and had to realize, that there is a huge difference between RAW and cRAW. If you enable lens correction, RAW still stores the pictures uncorrected, but cRAW picrures are stored corrected like JPG images. Take a picture from a straight line near one side of the picture. With cRAW and JPG, the line is straight, with RAW it is bended.
This makes cRAW almost useless for me. If I want RAW, I want real RAW and not somehow in camera processed RAW.
Wow that would be a huge difference indeed.
Are you really sure that there are not lens-corrections automatically done for you in whatever RAW converter you use???
I agree. cRAW is absolutely usable for all professional levels on the R5. I see NO degradation of picture quality.
Just curious if CRAW is slower than RAW for the camera body to process particulary when burst firing? I'm using an EOS RP and it's not the fastest for tracking in servo etc - if CRAW uses the cameras CPU to compress does that slow down performance as a whole when shooting?
A good question, to which I don't have a good answer for. I didn't perform any tests for that, but I don't have any problems with shooting bursts of images. Perhaps the difference is negligible?
Thank you for that very usefull comparison, Nando.
From all I've heard and read so far about cRAW, it is *not* a lossless compression. Anyways, I'm pretty convinced, the use of cRAW highly depends on what one is trying to achieve.
Thank you for this informative post! I just upgraded from Canon 6D ii to Canon R5. I shoot a bit of everything, but the main reasons for the new camera are to increase clarity in landscape shots and to be able to crop in aggressively in my soccer photos. Question: Does cRAW retain the full resolution that one would get shooting in RAW? Does the compression of cRAW strip away useful information if I want to crop heavily?
Craw is just like raw, but with compression. The resolution is the same, but the data is compressed. This can result in quality loss. However, I didn't notice any downsides whatsoever.
GTK. Thanks Nando Harmsen