Good examples of RAW's advantages over JPEG?

  • I'm curious to see some real examples of where simply capturing the same photo in RAW (and being processed by someone who can do it justice) has significantly improved the photo at the end of the process.

    I understand what RAW is and why you might want to use it over JPEG, however, I'd like to actually see some examples where it has allowed for a better result. More control over tone, conversion from the more detailed data to 8-bit RGB etc.

    Does anyone have or know of some processed RAW+JPEG shots for exact comparison?

    Just a note to other potential answerers. Please don't look at my answer as reason not to post your own examples. Like Nick, I think it would be very useful to see more visual examples of the benefits of RAW over JPEG. So, if you have an example comparison to offer, please post it!

    I agree. Also, it might be worthwhile capturing in RAW if just for the information that I, for one, thought might have been lost altogether due to the exposure. As a new photographer I wouldn't have even thought about RAW when trying to deal with something like that.

    Define better, better according to the photographer or to everyone else.

    For a specific example when shooting in difficult lighting, please see: Lots of noise in my hockey pictures. What am I doing wrong?

  • jrista

    jrista Correct answer

    11 years ago

    The Value of RAW:

    I think you may be misunderstanding the value of RAW. In the grand scheme of things, from seeing a scene with your eye to printing it, the best you get is what the printer you printed with is capable of, and that tends to be considerably less than what you see, or your camera or your computer is capable of representing.

    The value of RAW is not really in the end result, although it is possible for the end result created with a RAW image to be better than that created with a JPEG. The reason for this has to do with the workflow between snapping a shot and saving or printing a final image. RAW gives you headroom that JPEG can't come close to offering. You have the ability to recover highlights and shadows, apply alternative tone curves, rework old RAW images with newer RAW processing algorithms to get better results, etc.

    You are basically asking what is the value of an original film negative or slide, over a final scanned JPEG copy of that film negative/slide. With the original film, you have plenty of capability to rework and improve, use different printing techniques, etc., whereas with the final JPEG, you got what you got, and not a whole lot more.


    An original JPEG of Lower Yellowstone Falls. The sky was completely blown out, as this was one of the very first few photos I took over a year ago when I first got into photography. I had researched RAW, along with most other camera theory, long before I ever purchased a camera, so I had RAW+JPEG enabled at the time:

    Lower Falls JPEG

    Below is the reworked version from a RAW file. Because of RAW's considerable headroom, I was able to nearly fully recover the horrendously blown-out sky, retone the whole image, and generate three alternative exposures (-1.5 EV, Original EV, +1.5 EV) using Lightroom to create a far sharper, clearer, and richer HDR image:

    Lower Falls RAW Corrected

    It was largely because of the radical improvements I was able to make to this image that I rarely ever shoot in JPEG any more. I opt for RAW the vast majority of the time, and as I am still a student of the artistic aspects of photography, I appreciate the headroom that RAW offers. Most of the time, the final image saved from a RAW file is very similar to that of a JPEG; it's the times when you botched it big and need to massively rework an image that RAW's advantages over JPEG really start to shine. It's all in the workflow, rather than the destination. ;)

    JPEG Example:

    Mark took the time to rework the JPEG sample I posted, to demonstrate what can be done with a JPEG. I think its important to note that a JPEG is not completely unworkable once it is taken - I may have given that impression in my comments above. JPEG images do have some room to be reworked, if needed, however it is more limited than RAW. Mark's reworked copy of the JPEG sample is here:

    Retouched JPEG Example

    A couple things should be noted. For one, he was able to retone the image decently, and it looks similar to the retoned RAW example I posted. The retoning caused the unrecoverable parts of the sky to become yellowed, which I would consider an undesirable outcome. Depending on the software used, that may or may not happen. Something also not visible in the very small JPEG examples are compression artefacts, which have a tendency to become more pronounced as you rework an image, limiting your options.

    Detailed Example:

    Something else that I was able to recover from was a severe degree of softness, caused by the 18mm extreme of the cheap EF-S 18-55mm lens I used when I took this shot. I have some crops below that demonstrate the original image, a sharpened copy of the JPEG using a technique explained by @Guffa here on Photo-SE, and an HDR version that was only possible because with RAW, I could use Lightroom to export two additional alternative exposures 1.5EV from the original. Even using Guffa's excellent sharpening technique, the JPEG can't compare to the ability to create an HDR image from a single poorly-shot RAW image (these images are about 1/3 of their full resolution):


    And another example:


    The HDR examples were not sharpened using any normal sharpening technique; the added sharpness was the result of Photoshop's image alignment during Merge to HDR.

    UPDATE: Two years forward

    It has been over two years since I originally posted this answer. Cameras have changed, tools have changed, and the power of RAW only becomes more evident as time continues to march on. With the advent of Sony Exmor sensors, low-ISO dynamic range in the shadows has become legendary. The Nikon D800 allows unparalleled shadow recovery that exhibits barely any noise at all and good color fidelity. Not owning a D800 myself, I can't provide any of my own samples. Fred Miranda, fame of, has provided one of the best examples of the power of RAW in the form of shadow recovery, comparing the D800 and the 5D III. The results in his examples are stunning to say the least.

    For all the bad rap it tends to get these days in light of the D800, Canon shouldn't be forgotten. Before shadow recovery became a "thing", RAW was all about highlight recovery. Far more levels are allocated to highlights in a RAW image to start with, and the recovery power when dealing with overexposed highlights has always been pretty impressive. I encountered a series of photographs I'd taken of a dragonfly today that were terribly overexposed. I was sure they were all goners, as I'm sure almost anyone would:


    Just about everything in the shot above appears blown. The background, which was roughly an even mid-tone in real life, looks completely white. Imagine my surprise when I decided to at least give some exposure and highlight recovery a try. After -4 EV of exposure recovery and about 60% highlight recovery, I was shocked to see this:

    Mind blowing recovery!

    I've heard of such highlight recovery before, although generally only in discussions regarding medium format digital cameras (particularly Hasselblads, which have legendary highlight recovery.) Even the specular highlights in the dragonfly's wings seem to have retained considerable detail (100% highlight recovery detail below):

    Recovery Detail

    Since the photo was overexposed by about 4 stops, the shadows have full color fidelity, zero color noise, and hardly any random noise. With my previous examples, one of the commenters to this answer was able to do some recovery with a JPEG version of the image. With the original overexposed copy of this new sample, its highly doubtful that any amount of "recovery" could be performed on a JPEG. RAW is simply pure, unadulterated post-processing power... and it keeps getting better.

    +1 nailed it on the head. JPEG is the end, RAW is the beginning, and if you shoot JPEG, you just gave up your control of the end to the decisions of a programmer who may just have a different view.

    Thanks. Well, that's what I was essentially saying. What are some examples of why not allowing the camera to convert to JPEG itself can result in. Great example indeed :)

    Looking at the images I posted, I think the RAW version is actually Pre-HDR. The HDR version looks even better, as there is more tonality in the trees, and the sky is a bit brighter. However, despite that, I think the RAW version I posted still demonstrates its value over JPEG.

    I was able to get nearly all the way there starting with the JPEG, except for the blown out sky.

    @Mark: This small resolution is missing something else that you can't really see, and that is sharpness. The original image was taken with a cheap EF-S 18-55mm, which has terrible softness at the extremes (in this case, 18mm). The RAW version allowed me to generate alternative exposures (-1.5/+1.5 EV), which I was then able to Merge to HDR with Photoshop. The final image was orders of magnitude better than the orignal, which included both the recovered sky and the vastly improved sharpness.

    @Mark: Thanks for the reworked jpeg. I probably should have made one of those myself in the first place, to better demonstrate the point. :)

    Just as an update, I've been shooting RAW only about two weeks after getting my 50D (about 2 1/2 months ago) and would never go back. The amount of image depth you get to use is mountains above JPEG (no pun intended). I've since made one of my friends switch to RAW last night and to get Lightroom and quit using Photoshop.

    +1 but only after I finally found "the times when you botched it big ... that RAW's advantages ... start to shine". One thing that always bothers me is the argument, that one can "recover shadows/highlights" - this is only possible if they were removed by the In-Camera-JPG-processing, which mostly happens in situations with extreme lightings. If you "botched" the photo (e.g.: used a flash to fill in sinlight out but neglected to adjust aperture) RAW won't save you: the corresponding regions are whitened/blackened out on the sensor.

    @Leonidas: Thanks for the comments. There are always times when you botch it so bad that you can't recover anything, however those are the most extreme situations. I do dispute the point about not being able to recover "whitened" areas of a photo. If you look at my samples and Matt Grums samples, there are areas that appear completely blown out to maximum white, however given the headroom that RAW offers, those "botched" white areas ARE recoverable. In my sample, there are a couple patches that couldn't be fully fixed, however RAW allowed me to recover the photo to an acceptable point.

    Why do you create pictures with different exposures fro the same RAW file first, then merge them into HDR? Can't HDR tools work on either the RAW file or perhaps a 16 or 32-bit TIFF generated from it directly? The information is all there.

    As a novice photographer, although I understood the difference between JPEG and RAW at the point of capture, I've never really been able to determine how RAW can be applied to justify the extra step in processing to produce the final image.

    Excellent visual demonstrations and explanation about processing headroom and blown highlight recovery has helped me see why I should consider using RAW (or because SD card capacity is so large, use RAW+JPG). I like the dragonfly photo! I'm amazed at the fact that in the original photo, the blown highlights just appear the same white to the human eye, but to RAW and the processing software, it actually captures the actual differentiated colour that can be manipulated "to bring it back" to what was seen by the human eye. Thanks very much for your time in contributing this explanation.

    Intentionally over-exposing by a lot and then reducing the exposure from the RAW, like with your dragonfly, is actually a technique used to minimize image-noise. It's called **Expose-to-the-Right (ETTR)** - see for example here. This works as long as none of the brightness-values are clipped eg. greater than the greatest value that can be read by your camera's sensor.

    I find it really surprising that you used HDR merging from a single RAW file, as well as the fact that it actually resulted in sharpness. I would've expected it to be useless given that you only have 1 RAW file, and also that it would make things blurrier when you combine things, not sharper. (Or if it did make it sharper, I wouldn't imagine it'd be better than the RAW sharpening filter Photoshop has.) Would you mind explaining a little bit about this? Thanks for the answer! :) +1

License under CC-BY-SA with attribution

Content dated before 7/24/2021 11:53 AM