Raw revisited … perhaps


Some of you know that I am not a fan of raw. No, it’s not that I have anything against it, at all. It’s just that I’ve never seen an advantage in using it. I’ve shot pictures with it and without it and cannot tell any difference between a raw picture and a 1:4 compressed JPG. None. However, that does not mean that I close my eyes to the possibilities. I’ve never had trouble with banding, shadow detail, highlight detail, etc. Either I’m very lucky, I choose easy subjects to shoot, or perhaps I get my exposures right every time. I don’t know. Personally, I find ‘evangelical articles’ that extol the virtues of raw by regurgitating the same old things over and over again, without providing the least bit of evidence, a bit annoying.

Every once in a while you can come across something that might make you reconsider, at least in part. In reading a post by Mark Graf, I found that I had become interested in HDR. Historically, I haven’t liked HDR because it just seemed ‘weird’ and unnatural. Well, after reading Mark’s post and part of the article that he referred to, I started doing a bit of research. I was curious as to the difference between 12 bit raw and 14 bit and ran across this article. Interesting reading, especially the part about shadow detail and the mouse-over image was interesting in what it showed.

Mind you, I’ve not yet found a compelling article that has convinced me of the merits of raw, save for that it might get you a bit more shadow detail or highlight detail should you happen to botch the exposure. However, doing HDR, that little bit extra just might help, versus using the JPG and having the camera apply its own tone curve and possibly clipping those little bits of information out. I don’t know. Certainly, the higher JPG compression that you use, the more of a difference you’ll see.

As I like to keep my beliefs a bit pliable, I’m not above experimenting with HDR, raw, and 12/14 bit processing. Disk space is pretty cheap; I have a very fast card reader, so copying files is faster. I no longer use Mozy, it just proved to be to slow, instead I have switched to Acronis, so days-long backups are no longer an issue. Yet, I just cannot bring myself to shooting entirely in raw. As I said, if I cannot see the difference in print or on the monitor, there isn’t one. The mathematics of it are, well, just math. :-) Also, we output to 8-bit printers as well as 8-bit JPG to display on the web …

That said, though, I’m going to put my camera on lossless compressed raw for a bit and do some experimenting. The only thing that I do know is that changing to this mode slows my camera WAY down from 8 fps to 2.5fps! So, this is certainly a landscape only thing.

I’ll keep you informed.

About Paul

Comments

13 Responses to “Raw revisited … perhaps”
  1. Paul,

    It all depends on what you do in post-processing, and i don’t mean what you must do, I mean what you want to do. For the way you make your images, it probably makes no difference. I still think that you could profit in situations when light is only so-so, but maybe you can be picky about the kind of light that you photograph in.

    I can’t. I post an image per day, and almost without exception these are images that were taken that day. This makes it sometimes very challenging :) Let’s take for instance this image. The contrast was extreme, and the little bit of sky in the background is what I rescued by reducing exposure to -4.0 in ACR. Basically this is a blend of three versions from the same RAW, with different exposures and different color temperatures. The same can be said of this and that. In general, JPEGs break when you begin to stretch them. Color shifts are such a problem and so are steep curves. As long as you don’t do such things, you should be OK with JPEG. It’s a matter of artistic choice as well. For me the image from the camera is only where it starts. You won’t find many “straight” images on my site. That does not mean that I can’t enjoy them, it’s only that most of the time I don’t make them.

  2. Paul says:

    Andreas, I understand your point. 99% of my images are taken on a tripod and, using the histogram and highlight feature, I already know if the image is out of range for a single shot, so I bracket for both highlight and shadow and then blend the results together in Photoshop. Problem solved.

    Reducing the exposure in ACR is not exactly the same thing, as the picture has already been taken and if you don’t have the highlight or shadow detail, you just don’t have it. Fortunately for you, those images that you mentioned, still had some detail to be had. Possibly, that would have been lost in JPG … possibly, depending upon how close to the ‘ceiling’ it was, but I understand that point as well.

    Also, changing exposure after the fact increases signal/noise ratio, which can be very obvious in the shadows. There’s a lot of digital noise in the shadows. Increasing the exposure in the shadows, via software, also increases the noise, so it behooves me to get various exposures containing as little noise as possible and then blend the best of them together.

    As for the color temperatures, well that’s where raw beats the pants off of JPG, although JPG can be color adjusted, just not as easily or quite as effectively as raw.

  3. Mark says:

    Paul – are you are showing your ‘JPEG evangelical’ side? ;-) Free your mind my friend.

    (p.s. – keep in mind that monitors and printers won’t be 8 bit forever. We are already seeing the emergence of 16 bit printing and larger gamut monitors).

    People are certainly entitled to work with whatever format is best for them, but the differences come down to more than math dude. :-)

  4. Paul says:

    @Mark: LOL!!! Was I being an evangelist? :-) I hope that my mind is free, which is why I was thinking of going to ‘the dark side’ once again, or at least for a few moments. ;-) For now, I’ll still believe what my eyes can see and stick with my ‘math’ theory. :-)

  5. John Watts says:

    Paul,

    Welcome to the “Dark side”, LOL!!!

    Your post is very intriguing, as are the links – - I’m going to give HDR a whirl, and I’ll let you know how it goes…

    Cheers,

    John

  6. Adam Maas says:

    Note it’s not Lossless Compressed that slows the D300 down to 2.5fps, it’s 14-bit RAW mode. I always shoot my D300 in Lossless Compressed RAW mode, either at 6fps in 12-bit or 2.5fps in 14-bit.

    What Lossless Compressed does do is reduce the buffer size compared to lossy compression (which is much faster for the camera to do).

  7. Thomas says:

    I can only second Andreas’ post. Of course, a good JPG is as good as any RAW (or better, as you can actually show it to people). RAW is simply more flexible, it provides me with a safety net that I use quite often… But of course, that’s just me.

    Paul, I wonder what you do when converting to B&W? Do you use the in-camera conversion?

  8. Paul says:

    @Thomas: As for the black and white, I convert in Photoshop. I like having the option to either go color or black & white. Plus, I like having the control of the 3 channels.

  9. Anita Jesse says:

    I have drunk the Kool-Aid. I could never convince myself to go back to jpeg, but then I am not in your league—I need that safety net Thomas mentioned. I don’t do that much converting to b&w, still I didn’t think I would ever want to give up channels. Now with CS3 and smart objects, in addition to the New-to-me-b&w adjustment layer, I may be converted on that issue.

  10. Sure, the tripod changes a lot, and with bracketing you get even more headroom than from RAW, the only problem is, I shoot 70% of my images on the streets of Vienna, on my way to or from work, and there the tripod is not only impractical to carry around, it is also a big problem to bracket at all, because of all the moving people and objects in a city. Sure, Photoshop has nice tools to get rid of all that by now, i.e. to reduce the scene to the static elements, but that would mean bracketing for two dimensions: to get rid of moving objects, and additionally to expand dynamic range. This gets a little out of control, and if I had to do all this regularly, I would be better off with RAW.

  11. reza says:

    yeah, there isn’t much difference between jpeg and raw if you ask me. Visually they are the same on my screen. But when I do begin to work on a photo, a RAW shot seems to be more forgiving. I have no “proof” of this, its just how it feels when I work with it. I can do much more without the photo becoming distorted.

  12. Eric Jeschke says:

    Nice to hear from someone else who is not a RAW fanatic. I sometimes shoot RAW, but like you, I am used to careful exposure from my days shooting slides, so I don’t get too many surprises.

    I do realize that there are high bit depth advantages that are getting increasingly compelling. For now, I can store 3 bracketed JPEG exposures in less than the space of 1 RAW image.

    There seems to be a general refrain of “there is no good reason to NOT shoot RAW” on the web these days, when, in fact, RAW processing does impact your workflow and storage requirements in significant ways!

  13. Paul says:

    @Eric: Ah, another voice of reason. :-) Long ago I got tired of listening to that same refrain repeated over and over that JPG files where only for images that you didn’t care about. Snapshots, etc. I still disagree and much prefer JPG with 1:4 compression. Mind you, these files still come of my card up to 15 MB in size, depending on the complexity of the scene. They can also come out as small as about 4 MB if the scene is basically monochrome.

    I am, however, willing to experiment with RAW from time to time knowing full well that technology does improve over time. But, as yet, I have never ‘seen’ an advantage, even in pictures that I take side-by-side, one RAW, the other JPG. There is never a visible or printable difference. Now, if I decide to mess up the exposure to a certain degree, perhaps more of it can be saved using RAW. Also, as mentioned before, if I mess up the white balance, it’s easier to fix in RAW. Lastly, the other excuse I’ve heard, is that when you want to really push a picture to the limits, you might get banding. Point taken. I’ve seen this happen, but only because I wanted to see if it could happen. I don’t push my pictures that far. So, I guess that it depends on how you shoot and how you post process. I don’t play on the edges, so RAW is a negative thing for me.

    In the end, low compression and careful shooting will yield the same results. Also, my work flow is greatly simplified AND my files are compatible with most anything.