Do RAW files really allow more RAW dynamic range than shooting in JPG?

Question

I've heard a lot of time people saying that shooting in RAW offers a better dynamic range than shooting in JPG. But in some way I've always felt it as hard to believe.

So, the question is: there is any evidence of this fact?

Follow the (probably wrong) reasoning that I have done, why it's hard for me to believe that RAW has really a great advantage in achieving more dynamic range.

The reason is that if that I know that having 12 bit for each channel (instead of 8) offers the possibility to memorize 8 time more shades, so theoretically it would be possible to save more info in a RAW picture.

But at the same time I also know that the final result of a perfect HDR processing is shown using 8 bit.

So, somehow if a picture is taken to JPG as having some zone burned out (clipped) I wonder why the firmware, having the RAW information with the higher dynamic range, prefers to burn out some zone deleting details instead of doing a simple HDR on the fly to save some detail.

Also considering that the human vision is very similar to a natural HDR (it's really rare that the eye see the sky as being white because too much luminous)

Refer also to this link: Why is Adaptive Dynamic Range incompatible with ISO Expansion?

Answer 1

Yes, the evidence that this is a fact is that RAW images are used to make the JPEGs. It isn't possible for a JPEG to have a wider range than a RAW image because the RAW image is the actual sensor data from which the JPEG is made.

A JPEG is the processed image produced by the camera taking its best guess at how the image should be processed. It discards details as irrelevant that are at the top or bottom of the dynamic range and then makes black and white be the points it thinks are the black and white point from the image. The end result of this is that the JPEG may very well LOOK more vibrant out of the box, but it actually contains far less information because the processing has already been applied, detail has been discarded outside the range of the JPEG and the color depth has been reduced significantly (12 or 14 bit color to 8 bit in most cases, which is over an order of magnitude less color.)

You are confusing the dynamic range of the input with the bit depth of the output. In an HDR image, it is covering an expanded input dynamic range by looking at the detail in both ends of two photos to cover a wide range of input. Properly processed, this can be crushed in to 8 bits because you are making sure to spread out detail well across those 8 bits. HDR doesn't work for every image though. If the dynamic range of a scene isn't very wide, then using HDR doesn't accomplish anything. (Also note that the HDR look can be accomplished with a single RAW file on many modern DSLRs due to the high dynamic range they support.)

What has to happen to get from 12 or 14 bit RAW to a JPEG is that the information that is important has to be tone mapped in to the 8 bit range. The camera can try to do this itself or, for more accuracy, an experienced photographer can do this by hand. Keeping the bit depth high allows this detail to be stored until it can be tone mapped by an artist rather than a computer. This is not the same as dynamic range though.

You could have an image with less dynamic range, but more bit depth. Bit depth determines the accuracy and granularity of color, not the range. Range is the measure from darkest to brightest point. It is dependent on the input. Theoretically, a JPEG could be processed to have the same dynamic range as a RAW file, but generally, the camera is going to discard information that it crushes to white or black which results in the JPEG representing a smaller dynamic range than the RAW image.

Answer 2

The advantage of RAW is that the extra bits of precision are available during post-processing for the photographer to play with, so it's possible to bring out extra detail by increasing exposure in shadows and decreasing exposure in the highlights. Either automatically (using the software's "Auto" function) or selectively. Chopping off the extra 4-6 bits and converting to JPEG in the first place loses those details.

So, yes, when the final JPEG is generated from the RAW, it technically only has 8-bits per pixel, but can show enhanced detail that would have clipped had it been 8-bit in the first place.

Answer 3

I totally approve of the answer given by AJ Henderson, but it glosses over an important additional drawback of JPEG.
(Understandably as the question focuses on the color-range aspect of RAW vs. JPEG.)

Nevertheless I think it is important to mention this as I have found over the years that many people don't realize this little fact...

JPEG also looses fine detail !
It is a lossy compression method. It achieves it compression rate by discarding high-frequency components of the image.
Without going in the really technical details (go read the JPEG specifications (here) if you want a headache) it effectively means that quick color-changes (like on sharp edges) in the picture become slightly blurred.
It can also introduce small artifacts in the fine detail that weren't in the original image.
As a result the jpeg will always have slightly less detail (in color and resolution) as the original RAW.
Even on the highest JPEG quality settings this loss of precision is present.

Answer 4

Yes, the raw data absolutely allows for better end pictures when the scene has high dynamic range.

Theoretical Argument

A 14 bit sensor captures intensity with a resolution of one part in 2¹⁴ = 16384. You are partly right in that largely the same dark to light range is captured, and that ultimately you will display or see the picture with much more limited resolution. Let's say the final picture will be 8 bits/color/pixel, which is only 2⁸ = 256 levels.

The point you are missing is that it is important to be able to choose the final 256 levels from a continuum and not be stuck with the 256 levels evenly distributed from the darkest to the brightest point in the scene. Very often you want the final output levels to be non-uniformly spaced, sometimes significantly so, when mapped back to the original captured data.

Humans don't perceive light intensity linearly; they perceive it logarithmically. This means to get what appears to you as a fixed brightness increment, you actually need a fixed multiple in true light intensity. For example a 1.1x step by going from 50 to 55 will look like about the same increment as going from 200 to 220. Conversely, working this backwards, a +1 step from 200 to 201 will be hardly noticable, but a +1 step from 10 to 11 is quite significant.

A scene with reasonably high dynamic range can easily contain 1000x ratio in brightness from the dimmest part you care to see detail in to the brightest part. However, common display means don't come anywhere near that. Even a good print might be 50:1, or 80:1 under ideal conditions. The same holds true for computer monitors, especially since most are viewed with significant ambient light, therefore limiting how black black can really be shown.

Therefore, to make a final picture viewable on a 50:1 medium that starts out with 1000:1 contrast, for example, there needs to be serious compression of dynamic range. To make this compression look natural and acceptable to human viewers, it has to be done in the human conceptual intensity space. As I mentioned above, the human conceptual intensity space is the logarithm of physical light intensity.

The point of all this is to explain why linearly spacing out the 256 final display intensity steps over the original captured intensity range doesn't work. If you follow all the math, the result is that the limited display levels are grabbed from the original bunched up at the dark end and spread out at the light end. You can't do this and not lose information if you start out with only 256 linearly spread out levels.

There are other post-processing effects and other reasons for wanting to chose the display levels non-linearly from the scene intensities. All these require more captured brightness resolution so that you still have differences and smooth jumps between the limited number of output levels.

Practical Argument

We just had a question where someone posted a night scene (high dynamic range) and wanted to know where all the detail in the dark areas went. See my answer, which shows a great example of what happens when you try to perform non-linear brightness mapping starting with only the same 256 levels you ultimately want to display the result in. Basically, don't let this happen to you.

Answer 5

Today - the differences are narrowing between JPeG and RAW for most images on modern cameras.

My experiences is that while RAW shooting offers more chance to recover poorly exposed or white balanced images in post processing - switching to current generation high quality mirrorless cameras has meant that I now get things right first time in-camera more often and rarely need the greater post control of RAW that was essential in the DSLR days.

So I now use JPG for the majority of my shots, and shoot AEB sequences where extended dynamic range is needed (new mirrorless cameras have way faster shooting of burst sequences that makes capturing HDR handheld a breeze) . Now I reserve RAW for a limited number of images where post-processing is more important: dark scences, white wedding dresses.....scenes where I know that I potentially want a lot of extreme post processing.

This minimises computer time - focusing it to where it will do best - freeing time to hold a camera instead! This is important whether photography is your career or hobby.

Each camera (and make) has its "best" JPEG in-camera settings that you need to know (that is why those "scene modes" exist, so use them!). My experience is that with older generation cameras they were set to "shoot to the right" and often blow JPG highlights to avoid underexposing shadows. Modern software is now so good at controlling this, that my older cameras are now shot with a -1/3 to -2/3 EV exposure setting when in automatic exposure mode. This enables faster shooting with older cameras (their caches used to fill up after just a few shots) while I can easily raise the shadow exposure by a stop or more without loss of image quality in post.

My experience is that Olympus and Fuji JPEG settings are very good SOOC; while my Panasonic, ANd Nikon need more individual controls. Canon amateur spec cameras deliver great JPEGs SOOC but the pro-spec ones, like NIKONs, need more personal input.

In a fast chaging situation, it used to be said that RAW enabled more images to be rescued if shot with sub-optimal exposure or colour balance settings. But I have found that each newer generation camera is delivering better and better results from its auto-scene selecting program mode; so this is a good setting to leave the camera in when "snapshots" might suddenly be required (example, the IAS setting on the current Lumix cameras).

Answer 6

If you will finally display your photo as a JPEG, as I will, the comparison is between the RAW->JPEG conversion your camera will do and the RAW->JPEG conversion you will do in post. The conversion to 8 bit JPEG will happen one way or the other. I shoot a lot of bird photos, which are often badly backlit. I get better details out of the shadows by processing a RAW file in Lightroom than in processing a camera JPEG. If the camera conversion were as smart as I am there would be no advantage to shooting RAW. That day may come, but I saw a big upgrade when I got a birding camera that would shoot RAW-the Nikon P950 compared to the P900. Same lens, same sensor, but RAW capture available. When I have time I use(d) exposure compensation in both cameras, but the P900 seemed to fight it in the JPEG conversion trying to expose the rest of the frame correctly.