Which file format should you shoot in, when and why? Now that most photography is being captured in a digital format, we have so many more choices when it comes to processing our image files. Granted, the amount of time you spend in front of a computer may be directly taking away time from behind the lens, but the question remains, do you prefer being an chef, or would you rather have someone else cook your images for you?
Cameras capture a large amount of analog information with each shutter actuation which is then processed into a digital image file. Each image file is compressed into a readable format. The amount of compression is where the difference between RAW and JPEG files come into play.
A JPEG file’s compression parameters are determined in camera. If you adjust the sharpness, saturation or contrast via the sliders found in your camera’s menus, you are adjusting how your camera is applying those instructions to compress the image file once you take the shot. If you are shooting in a “scene” mode, like smooth skin, landscape or sepia, that is a set of compression instructions that the camera will apply into its JPEG compression. The tricky part about this, is once those levels are adjusted in the JPEG file, they are pretty set and have little latitude for manipulating those levels without pushing the file to its limits in post production. The main benefit with JPEG files are, due to their compression, they are much smaller files. This allows you to fit quite a few more shots on a memory card (and your hard drive) vs RAW files. The other boon is that they come out of the camera as saturated, sharpened, “finished” picture files. Most of the time, assuming you got your composition, focus and exposure down, they’re keepers.
A RAW file is the information gathered by the sensor prior to compressing it. Because it contains almost all of the information captured by the sensor originally, the file is much larger. You have the ability to manipulate the information quite a bit more than with a JPEG file as the information has yet to be compressed. A RAW file will need to be adjusted for sharpness, saturation and contrast to get it to “look” like a compressed JPEG file. How you apply those, and other adjustments, is more in your control though when shooting RAW as you’re not asking the camera to determine those values for you. I think of it as a giant pot of stew. With a JPEG, I have a good tasting pre-made stew that I need to thaw and heat up. Ready to go. With RAW, I’ve got my carrots, celery, potatoes, sausage, collard greens, onions, garlic, stock and whatever else I feel like throwing into my concoction. I get to choose how the ingredients work together by introducing them in the order and volume I choose. This gives me a more finely tuned stew to my taste, but is much more work to get it to a finished product.
(Very) simply put, the major difference between an 8 bit compressed JPEG file and a 12, 14 or 16bit RAW file is the range for and depth of each of the three color channels (R-G-B). With an 8 bit file, you get 256 levels of luminance per channel (2^8) akin to your histogram showing the ‘range’ of tonality and luminosity from pure black to pure white in each of these channels, with 12 bit it grows to 4096 levels per channel (2^12), 14 bit is 16,384 levels per channel (2^14), while a 16bit RAW file will have 65,536 levels per channel (2^16)! Thought of simply, this would be like taking each of your original 256 values and dividing them 16x, 64x and 256x! This gives much smoother transitions between your highlight to shadow info throughout all three of your captured color channels. Once you’ve compressed a file, you are limited to its bits per channel based on the last compression. Herein is where a continual modification of a JPEG file continually degrades the image as you are essentially re-compressing already compressed data. If you modify a single JPEG file multiple times, with each compression/save you are getting rid of more and more data with each compression/save. This is one major reason that many photographers prefer to use RAW files (and a ‘non-destructive’ workflow). You’re gaining 16, 64 or 256 times the information per pixel essentially depending on the amount of bits your RAW file records (most current dSLR’s have either a 12 or 14bit RAW file, while the only current 16bit RAW file capture that I’m aware of resides in the digital medium format world…)
Now, using the image above as an example, the JPEG version did a good job at getting the colors, saturation and sharpness to where I wanted it. With the RAW file, I decided to go a little overboard with the processing just to show how much more info I could bring out in the eyes and increase the detail in her hair. I also added a vignette to the processed RAW file just for kicks. If I’d played with the JPEG file a bit, I’m sure I could have made it look a bit more like the processed RAW file and because I lit this portrait with 4 different lights, I intentionally maintained as much shadow info as I could which could enable me to utilize the JPEG file to better effect. As it is though, it required me to work with the RAW file to get it to look similar to the JPEG, while maintaining much more info to work with. Is it worth it? To me it is, but to you it may not be necessary.
The first time I shot RAW files, I wondered what I had done and why I had ruined my photo opportunities for the day by capturing flat, unsaturated, soft pictures. Not knowing exactly what a RAW file was at the time, I had just decided to figure it out by doing it the only way I could think to, by taking pictures. As I have evolved my digital workflow, I’ve come to appreciate the room to work so to speak. I quickly learned that not only did shooting RAW allow me the room to fudge my exposure, but saved my tail when it came to certain aspects, namely white balance and overall exposure issues pertaining to shadow and highlight values. With a JPEG file, if the light in your scene makes for an overly yellow/orange tint coming from incandescent bulbs for instance, an incorrectly balanced photo can be tricky to readjust in the computer. While shifting the tint sliders to accommodate for the white balance shift, it can globally alter elements in the photo incorrectly while not taking into consideration how the white balance should be differentiated in the tonal range of the image. You can alter the white balance with a JPEG file, but not as accurately, and certainly not as much. A big area that you see a benefit to RAW vs. JPEG is in both highlight and shadow information. If you try to pull back blown highlights in a JPEG, it just turns from white into gray while in RAW, unless the info is entirely clipped and lost, it is pretty amazing how much info you can pull back into those pixels. Think clouds in the sky going from white blotches in your image to having discernible contrast and dimension as opposed to the white sky just turning a murkier shade of gray. Clipped shadows on a JPEG are all but lost. Try to brighten up the image and watch the black areas in the photo become pocked with noise (graininess). While even a RAW file cannot do much for dark areas where there was very little info to begin with, you are able to bump those shadows up much further before they turn into blotchy noise riddled areas.
When importing a RAW file into your computer, you need a RAW converter. Depending on your camera and it’s proprietary RAW format, you may need to double check your software’s compatibility to convert the files. Most all cameras, when purchased new, ship with software to convert it’s RAW files. Other 3rd party software (like Adobe Camera RAW, Adobe Lightroom, Apple’s Aperture, etc) are constantly updated to be able to handle the RAW files as new cameras come to market. If you use a particular software to import your image files and it doesn’t recognize the RAW files, check the software company’s website for updates. There are other work arounds via DNG converters and the like, but the easy answer is to figure out if your software of choice can convert your RAW files, and if not, shoot JPEG, or JPEG+RAW until you figure out what you want to do with your RAW files. It isn’t entirely unheard of to run 2 or 3 different image importing software programs to handle different RAW files as well as different tasks. I’ve used as many as 4 due to different camera’s not being supported by my first software choice, which I then convert the RAW files to TIFF’s after manipulating to my liking and export them to my main library inside of Aperture keeping the original RAW files unaltered in the event that I want to go back to it for any reason in the future.
So, when if ever, should you switch between JPEG and RAW file formats? Assuming that you have your software workflow figured out to easily convert your RAW files if and when you shoot them, the choice really comes down to the amount of info you want to work with. Personally, I shoot RAW almost all the time. If I am doing any work for hire, or personal projects where I know I will want to plop them into Photoshop or another type of software, I will shoot RAW. My only exceptions have been due to incompatibility issues where I don’t want to go through extra steps to convert the RAW files through a different software because I’m lazy and just snapshooting. All my personal RAW workflow issues aside, if you are happy with the images coming out of the camera in JPEG format, shoot JPEG. Most cameras have pretty amazing JPEG compression capabilities built into the camera and in many cases, look better than what you may get doing all the tweaking yourself. It is all a matter of taste. If you plan to do a lot of modification to your pictures (using Photoshop, et al) then the extra data per pixel in a RAW file might be a better route. If your camera allows you to simultaneously shoot RAW+JPEG, it might be a good way to figure out which is best for you. Regardless, shooting RAW or JPEG it is a good idea to make a copy of the original file so that you don’t lose the image in the event that a year later you realize you’d like a color version of one you’d already converted to black and white. Many software programs offer a “lossless” or “non-destructive” workflow which will automatically leave the original alone and copy the info as you assign it parameters through whatever manipulation you apply. I think it is worth purchasing a non-destructive software like Lightroom or Aperture if taking and manipulating pictures is something you enjoy doing. If you enjoy making the changes yourself and see post-processing as part of the fun, go with RAW files (or shoot a combined RAW+JPEG capture so you have the best of both worlds). One other thing regarding a RAW file that seems to stump quite a few folks is that if you have your camera set up to capture RAW files, and have for instance changed your picture mode to display black and white, or a more saturated version which will show up on the LCD screen, when importing it, it will be the RAW image data without any of those parameters applied. RAW is just that, raw uncompressed data while the image on your LCD screen is a JPEG thumbnail which will show you what you ask it to show you. This is another reason many folks like to shoot RAW+JPEG, you get your RAW file and a pre-compressed JPEG version with whichever parameters you applied in camera. Keep in mind also that RAW files will fill up your hard drive more quickly due to much larger file sizes. But then again, memory is cheap. You can get a 1-TB external hard drive for less than $100 now.
With JPEG you get a nice, easy, finished stew whereas with RAW, you get the ingredients and you spend the time creating the finished product. Both way’s you get to eat, it just depends on how much work you’re willing to do to get the meal fine tuned to your taste.
Thanks for the stew analogy. I love stew. This helped clarify the differences between the formats for me. I always kind of knew what each was capable of, but not really. Sounds like as with most things in life, it is all about priorities.
Thanks for the excellent tutorial. I noticed, I think, that my 7D, which has a RAW/JPEG button on the back, appears to capture the RAW version at the same file size as the JPEG, such as 5.68 megabytes. Actually, for some strange reason, the RAW is slightly less than the JPEG files, like 5.49 for example. That’s a mystery to me, but I suppose the engineer types out there can explain why that is. What I had hoped for was that the RAW button would take a 20 plus something megabyte file and the JPEG would be whatever it is. I can get the 20 plus file size if I set the menu just for RAW, but I liked the idea of the RAW on/off button. Does anyone know any way to make that happen? I think I am setting it up properly in the menu section, so this may be just the way it is, or maybe, heaven forbid, I actually don’t really know what I’m doing. Again.
Dan, if you go into the menu and find the “Quality” field, you can designate 3 different RAW sizes (as well as 6 different JPEG compressions). RAW, S-RAW1 and S-RAW2. My guess is that it is set up on one of the S-RAW settings. You can also set the RAW+JPEG capture in there I would imagine. I’ve not played around with the 7D yet…:) so I’m not entirely sure, but it should be something like that.
S-RAW is just a smaller RAW file which sizes down to I think about 9mpx and 5mpx files respectively which would explain the file sizes. Let me know if that doesn’t work and I can do some research.
Pingback: *Panasonic GF-1, my new obsession. «
Pingback: *The Exposure Trifecta. Or, how to shoot on Manual. «
Not sure I see a lot of difference in the RAW file and the Jpeg files…hair seems to have a bit more tonal differences, and if this is a hair commercial shot, that may mean a lot…!
Thank you for your words, they fall not on deaf ears…may still be pearls before swine however, we will see….smile !
Thank you for taking the time to read and comment. One big point you bring to light is that for many of the shots that any of us take, I doubt many of us would be able to tell the difference and you’re very valid in saying so. Assuming you found this through the thread on flickr, I think one thing that was mentioned there that makes for a very interesting (and equally as valid) point is that most of us are viewing our pictures on LCD screens which are only able to provide us with limited bit depth and overall resolution. Having a young child has shown me just how few pictures I actually print compared to the total amount anymore which, in many cases would never require a 12 or 14 bit per channel file when they’re used for on screen viewing, emailing the g’parents, etc. There is even the argument that if I were to have a lab print these, I’d normally have to compress down to 8bit Jpegs anyway (although if I’d had the 12 or 14 bits/channel to begin with, I’d have the ability to manipulate the exposure, et al prior to that compression). There really are no concrete answers as to which is better than another as both have their place. For me though, I can always compress a larger, deeper file into a smaller, more limited file, but going the opposite direction is difficult 🙂
Pingback: *Everything you need to know about digital photography. Two years of tips, tricks and various freebies, revisited. «
Pingback: *OM-D E-M5 vs G3… what’s this about a new sensor? « Tyson Robichaud Photo-blography
Could you please tell me, how can i get 12- or 14-bit prints from my RAW-files? What kind of printers do large foto labs have?
Prints are going to be more geared around color space vs bit depth as far as I understand it. Most all commercial print labs that I’m aware of require any file to be converted into an 8bit JPEG print ready file. This doesn’t mean having the original info boiled down to an 8bit per channel at the time of capture will produce the same thing. Having the information to work with will enable you to manipulate it (in RAW) to get it where you want it to be as opposed to that info being locked down from the get. Certain printers will have better range of colors being able to be recreated within a particular color space (sRGB, AdobeRGB, Lab color, CMYK, etc) where cheaper printers will be unable to recreate the range of even sRGB in many cases. Basically, the bit depth will determine the amount of information while the color space will determine how that information translates to print. It’s always easy to get rid of info after you’ve adjusted what that info “looks like” but you can’t add info after the fact, at least not organically.
I’m certainly no print expert though and have a lot of room to grow my understanding in that area for sure.
Thanks and hopefully this helps a little.
Pingback: *The Mirrorless Diet, how to lose weight instantly! Part 2 | Tyson Robichaud Photo-blography