Posted by Peter Conway on June 27, 2007 at 2:42 PM
Being new to the world of Digital photography and using a Canon EOS 1DS, I have been shooting in large jpeg format rather than RAW. It just seemed easier to use and process. Am I making a big mistake?
I am shooting wildlife, primarily birds, using a Canon 400 2.8 IS lens often with a 2x extender.
Do I need to just plunge into RAW and learn how to deal with it? What will be the big advantages if I do?
A Raw file is like having a B&W negative. If you have ever had or been in a B&w darkroom you know that you can process the same negative so many different ways to achieve different print looks. Same thing with the Raw file, You can convert that file to have many different looks, You can't do that with a jpeg without losing quality.
The jpeg the camera produces is the finished product, If your white balance, exposure is out the pic is ruined. Sure you should get those right in camera, but the jpeg then can't be changed like a Raw file can. The more manipulation you do to a jpeg the worse the image file will be.
I shoot raw because storage is cheap and I have basically a digital negative of the image
I agree with Keith, The way I see it, with all of the trouble and expense that you go through to get good photos, why not walk away with as much useable information as possible. You can always convert to JPEG later, and dialing in your photos afterwards is half the fun. Just make sure you throw out the clunkers and storage won't be too much of an issue. (Especially if they're mostly clunkers like mine)
I shoot in RAW for virtually all my nature images now. I started out with jpeg because I had no idea how to convert and process the RAW files. I concur with the advice above. With the RAW file you can adjust exposure, white balance, etc. and still have the original RAW file to save as a negative. RAW conversion and processing is not as hard as I originally thought it would be. I started out using Capture One LE for my RAW converter which I like very much. The LE version was fairly inexpensive and did everything I needed. I have since moved up to CS3 with has Adobe Camera Raw with it, so I don't use Capture One any more. For starters Capture One LE and PS Elements would serve you well. You can upgrade from there as needed. I wish I had learned to shoot in RAW earlier, because some of my favorite images are jpegs, and unfortunately I didn't understand that they would degrade after making changes and saving multiple times. One of my favorite images is practically ruined because of it!
Eventually, we will convert you to the joys of RAW
Until then, if you are more comfortable with JPEG, use it. If I remember correctly, your camera has an option for shooting both - use that. You'll get the JPEG you want, but will retain the "negative" for future use. The downside is you'll use a lot more memory, both in terms of cards and computer resources, but I think you'll find it a good choice 2-3 years down the road.
Posted by Peter Conway on June 27, 2007 at 11:18 PM
Thanks everyone. I use Aperture to download my photos into my computer. Then I use "open with external editor" which opens the photo in Photoshop CS. If I use RAW do I just follow the same steps or do I need to do something else? Can you tell I just do things by trial and error?
If you use RAW in CS, you will get a special options/editor that will open prior to opening it in CS. Here, you can adjust things like white balance, exposure, saturation, etc. The more you process the image right out of the raw file, the cleaner it *should* be.
For a quick test that convinced me...open a jpeg and then the saturation dialogue box...tweak the saturation way up...farther than you usually would...it will look like a cartoon. Then, do the same thing with the raw file...tweak it way up and you will see it will hold it's composure much better.
If you d/l your files in Aperture, Aperture will become your raw converter-editor. That's actually a good thing. Canon has tested their own DPP, Aperture, and Adobe Camera Raw and found that the Adobe raw converter is a bit more noisy than DPP or Aperture. When you set up your preferences in Aperture, you tell it what format to open images in when you select Open In External Editor. I think .tif is the default, but it's been a while... Pick a lossless format like .tif, .psd, etc. The .jpg storage algorithm loses information. So... you download it from your camera already short of information, do some post-processing, and save it again. You lose more information from a file that started out with lost information. Open it again at a later date and save it and guess what? More information goes in the toilet. Shoot raw and download your files into Aperture and the pixels in your original shot will never be harmed.
The stuff the wingnut in the third article says about .jpg is true, but it also applies to shooting raw. Get your white balance and exposure right in the camera. Then you don't have to tweak either in your raw processor (Aperture). I understand where he's coming from there. Shooting a wedding, for instance, where I'll d/l several thousand images, I don't want to have to tweak exposure and white balance, or rescue shadows and highlights, on each one to proof them. Get it right the first time and it saves you a ton of PP work whether shooting raw or .jpg. Still, there's no reason to throw away information the sensor records by having the camera do a .jpg conversion on your image files.
Posted by Lance Warley on June 28, 2007 at 10:18 AM
Peter, as a follow-up on Alice's comment, I use PS Elements and the raw editor works fine for me. If nothing else, it is a huge help in pulling blown-out highlights back to details.
It's easy to use. If you wish, I'll be happy to talk you through it on the phone. Just send me a note.
Posted by Peter Conway on June 28, 2007 at 10:20 AM
Thanks for all the help but as always one answer leads to more questions. I shot last night in RAW just to see how it worked. Aperture shows the file as a name.TIF file. Is that correct or should it say RAW? From what yu guys are saying I should make exposure corrections to that file before I do anything else? When I open it to photoshop Aperture makes a duplicate tiff file and sends that to photoshop.
Should I upgrade my CS version to CS3?
thanks in advance!
Posted by Peter Conway on June 28, 2007 at 10:24 AM
Lance, I missed you note perhaps I was posting at the same time. I would like to use Aperture to download since it is so easy to do (and I sorta know how) If I can get the right process down for bringing in RAW and playing with it from there I think I will be in good shape.
thank you!
Posted by David A Bliss on June 28, 2007 at 12:05 PM
Keith, I have to disagree with your blanket statement regarding jpeg.
Working on a first generation jpeg will not cause any more “damage” to the image than working on an imported RAW. When you open a jpeg in PS, you are not working in jpeg. You are working in PS’s (or your software editor of choice) native file type. When you save it, it will save it as whatever file type was originally opened, unless you specify otherwise, or if it can’t save to the original file type (RAW). Something that goes hand in hand with this misconception is... “Working in jpeg will cause you to have more banding and artifacts, because jpeg is compressed.” No, working in jpeg will not cause this, because you do not work in jpeg, you work in PSD (in Photoshop). If you are getting banding or compression artifacts, it is being caused by your workflow, not by the simple fact you shot in jpeg.
If you shoot jpeg, you should open it in the image editing software, then save the file as a tiff, psd, or some other uncompressed file. Then you are never saving over the original jpeg, and it becomes the "digital negative." It would be poor workflow to ever save over the original file. It should always be archived just like a RAW.
Remember, just about any editing can be "destructive," since you are adjusting, converting, adding, or erasing pixels. Even if you start with a RAW file, you can end up with gradient banding and pixel artifacts. Your statement implies that editing an image that was shot with jpeg is more destructive than one shot with RAW. Editing an image that was shot using jpeg will be no more destructive if proper workflow is followed. Remember, as was said before, once you have imported (opened) the file in an image editing program, you are then working in that program's native file type.
Also, it is wrong to say that white balance cannot be adjusted in jpeg. Technically, it is a true statement, since “white balance” is not being adjusted, but it implies that hue cannot be corrected when the image was shot in jpeg.
What's funny is I shoot RAW, but defend jpeg! I swtiched to shooting RAW from jpeg a few years ago because I "supposed" to. I have become so comfortable with my RAW workflow that it doesn't add any more time or effort than shooting jpeg. I truly do like the flexibility of RAW, but it depends on my subject on how much I utilize it.
For landscape and nature, I tend to use RAW processing a lot. I adjust shadows and contrast, and use selective saturation, to try and acheive the Velvia look I am used to.
For portraits I tend to use less RAW processing, and do more of it in PS. Since I have my camera set to the white balance of my strobes, it comes out of camera pretty close, and I prefer to use selective color to fine tune the skin tones. If anything, I process the RAWs somewhat flat, with less contrast, to give me more latitude as I am masking and doing selective processing.
One BIG thing you missed, David, is that JPEG is 8-bit, whereas most (all?) RAW files are usually 12 or 14 bit. This, I think, is what allows more processing flexibility, especially in regard to banding. So some problems could be caused by the fact that you shot in JPEG.
I always work with JPGs, since my camera can't shoot RAW, so I'm very familiar with the limitations imposed by starting with a pre-processed (saturation, sharpening, contrast already applied), 8-bit fiel. I'd definitely be shooting RAW if I could, mainly for two reasons : greater bit depth, and the ability to choose various saturation/sharpening/contrast levels than what my camera applies.
For what it's worth, I only posted the third link above because it's one of the very few I've seen arguing FOR JPG in any way. Seems to me, unless you're time constrained, or really hate doing the conversion, there isn't much reason not to use RAW.
Posted by Peter Conway on June 28, 2007 at 3:06 PM
When one uses Aperture to download RAW files from the camera what is the format of the file in Aperture. Mine says it is a TIF file. Is that correct or should it say it is a RAW file?
Peter, if I remember correctly the 1Ds RAW files were (very strangely) given a .TIF extension, so yes this is correct...if a little confusing! The mk2 bodies and other recent digital EOS bodies give RAW files with a .CR2 extension.
Peter, I'm sure you paid good money for the 1Ds. By using JPG instead of raw you're not really using much of what you paid for.
JPG files can only hold 8-bits (256 shades) per channel. In raw mode your camera is capable of capturing 12-bits (4096 shades) per channel. This means that you are potentially throwing away thousands of color gradations by using JPG, which can make a big difference, especially in large prints where too few shades will lead to banding and other artifacts.
In JPG more you are also giving up the ability to make post-exposure adjustments (e.g. exposure compensation and color temperature) that in some cases can save an image that suffers from exposure or color issues. If you have blocked shadows or overexposed highlights in a JPG, there's nothing you can do to bring them back. With raw capture very often you can salvage details in these areas.
The "down side" is that raw leaves a lot more of the decisions to you and you will likely spend more time processing your images. In JPG mode the camera's computer makes a lot of these decisions for you and you don't get any say in how it makes them.
So you paid for a Porsche, now it's time to take it out of first gear...
The files have a TIF extension to make them compatible with operating systems that do not recognize the raw file types. Part of the file which contains a thumbnail actually is compatible with the TIF format so most operating systems can display a preview of it, but you will need a raw converter to get the full size image. Canon gave up on this strategy in their current cameras since it does get confusing.
Guy
Posted by Peter Conway on June 28, 2007 at 5:31 PM
Thanks all. I am going to play with it again tonight. I will learn to use RAW :-) Which is .TIF in Aperture.
Guy and David, you both mentioned using 16 bits (12 actual bits with the canons) per channel instead of 8, which is a good point, and a whole different discussion I have seen both sides of the argument for using 16 bits/channel instead of 8, and I can't say I have been wholey convinced that using 16 bits produces an significantly better final print. Both monitor display and (most) printers use 8 bits. I am not arguing that 16 bits is not better, or that it doesn't reduce or eliminate banding artifacts (i.e.). I just haven't seen a final print, where the processing was done with good workflow, that could show me a visible difference between the two. I'm always up for being shown though! Also, for the most part, I would guess that most amateurs and weekend warriors are not working in 16 bit.
Guy, you mentioned the one point that I truly think makes raw the file of choice. Raw has a lot more latitude than jpeg to bring back highlights that seem to be blown out. Of course, it is still probably necessary to open the raw file twice, once corrected for highlights and the other corrected for midtones or shadows, and then mask them in PS (or your editing program of choice). I can't speak for CS3, but so far there isn't a way to selectively correct for exposure. It is an all or nothing proposition.
Also, I still disagree with the implication that color correction is not possible in jpeg. While it is much easier to adjust the white balance on the raw file, it is again an all or nothing situation. On many of my shots, I much prefer to selectively adjust the colors using masks, etc.., instead of blanket changing the white balance.
I guess what I am saying is that PS (and many other editing programs) are capable of doing so much more than most people (including myself) use. Using your analogy, if the camera is a porshe, even if you take it out of first gear, it doesn't do you any good if you are driving on a residential street. You gotta open that bad boy up on the freeway!
David, high bit color is really aimed at minimizing loss of detail and tonality in editing. Obviously the final edited image will need to be implicitly converted to 8-bits for printing. Still, in the transition from capture to ready-to-print file, more detail in the original capture will often mean less loss of gradation and fine detail in the end product as well.
Of course color correction is possible in JPG, but at that point you are starting off with less colors and gradation so you run the risk of banding in some color ranges (e.g. if you don't have enough shades of blue and you want to "cool down" the image.)
It is not always necessary to convert a file twice. In the conversion from raw, a tone-mapping curve is applied. If you let the camera use its standard curve in an image with subtle highlight (or shadow) detail, you may end up with multiple values mapped to pure white (or pure black) and lose that detail, whereas in manual conversion you can alter the characteristics of the curve (either using the convenient sliders for contrast, fill light etc.) or even by editing the curve itself which allows you to work on individual ranges of tones independent of others (e.g. you can reduce the contrast in the highlights and keep more detail there, without altering the mid-tones or shadows).
Peter, there definitely is a huge advantage to shooting Raw for all the above mentioned reasons and not the least of my favorites is the ability to shoot two Raw's of the same subject at different exposures and blend them with a mask on your computer to get more shadow/highlight range of detail than could be achieved in a single "compromise" exposure. You can also achieve this to some extent by double processing a single Raw as long as you haven't lost your highlights.
If you shoot .jpeg always use the highest quality that your camera will give you and never re-save a .jpeg after making corrections and adjustments as .jpeg. ALWAYS re-save as .tiff.
Every time you re-save a .jpeg as .jpeg you will degrade the quality because of re-compression. You can work on an open image file as much as you want as long as you can undo your mistakes in the "History" palette AND as long as you do not re-save as .jpeg.
Peter, there definitely is a huge advantage to shooting Raw for all the above mentioned reasons and not the least of my favorites is the ability to shoot two Raw's of the same subject at different exposures and blend them with a mask on your computer to get more shadow/highlight range of detail than could be achieved in a single "compromise" exposure. You can also achieve this to some extent by double processing a single Raw as long as you haven't lost your highlights.
If you shoot .jpeg always use the highest quality that your camera will give you and never re-save a .jpeg after making corrections and adjustments as .jpeg. ALWAYS re-save as .tiff.
Every time you re-save a .jpeg as .jpeg you will degrade the quality because of re-compression. You can work on an open image file as much as you want as long as you can undo your mistakes in the "History" palette AND as long as you do not re-save as .jpeg.
Bob
Posted by David A Bliss on June 29, 2007 at 12:11 PM
Guy, your right of course, using the curves can help to pull out some details without losing others. I was very over simplifying the process.
It does go back to my point, though, that to truly use the power of raw, then a good knowledge of PS (or your editing program...) is required. Not that I see it here as strongly as I do on other forums, but I think it is wrong to tell someone that because they are not shooting in raw, then they aren't doing it correctly. Some people like to spend their time shooting, not processing. I shot Velvia for years (as many of us did), without the benefit of LCD previews, RAW processing, or PS, and managed to get good photos. I think that forcing someone to use raw who doesn't feel comfortable on computers, doesn't understand or want to learn PS, or doesn't want to spend time processing is wrong. If the same attention to detail is given at the time of the shot that was given when using slide film, then incredible, ready to use images can be achieved using jpeg as the medium of choice.
Don’t get me wrong, I am not one of those who say using raw is simply a crutch, and good photographers don’t need it. I use raw, because I love the versatility is provides, but mostly I enjoy the processing. For me it is a lot of fun to get my hands on the picture. Maybe it’s because I never got to spend much time in a darkroom, and this is my way of making up for it!
I do think it is important for beginners to spend their time shooting. And not just shooting, but shooting correctly. They need to learn their equipment, learn composition, and learn how to properly expose an image. They need to learn how to use shutter speeds to their advantage, what aperture to use and why, and, in the digital age, what white balance to use. Jpeg can be very useful for this, because it forces them to learn to do it correctly in camera, instead of being able to correct for sloppy exposures in raw. Sometimes I think using raw right off the bat is like putting the cart before the horse. To use an analogy, it’s like teaching someone to do ornamental woodworking before teaching them to cut straight.
I really don’t care about the raw vs jpeg argument, per se. For me it is about pushing an agenda (which is not to say that is what you are doing, don’t take it wrong ) which can lead people down a path for which they are not yet ready, which will frustrate them, and take their energy from that which they should truly be concentrating. Both jpeg and raw are tools, and have their place.
I have already turned this into a huge post, but I still need to reply to your comments about 16 bit.
When you say you run the risk of banding using 8 bit, that also means you would be running the risk if the raw is imported at 8 bit. So this isn’t a raw vs jpeg issue, it is an 8 bit vs 16 bit issue. (Yes, still tying into the jpeg argument).
Going into the 16 bit discussion, like I said, I have yet to see an example of a final print that had a noticeable improvement in quality by being processed in 16 bit over one processed, with good workflow, in 8 bit. Yes, I have seen the pixel peeping 2X screen shots that show difference in the two, but how many people actually look at the 8x10 or 20x16 prints with a loupe? If you are processing for truly large prints, then I could be persuaded that it would be important to work in 16 bit.
Ok, that’s enough for now. I’m sure most everyone gave up on this post half way through it!
Posted by David A Bliss on June 29, 2007 at 12:56 PM
Robert, I would like to make a couple of points regarding your post, and then I need to get back to work.
Shooting two images of the same scene and blending the exposure can be done with jpeg as well. This is not something that is raw exclusive.
I agree with your comments about jpeg workflow. The original file should be saved untouched, regardless of the file format used, just like raw, be it jpeg, tiff, or any other. Saving over the original file destroys any chance to go back to the original file and reprocess.
Saving a jpeg in of itself does not cause more compression, though. If you open a jpeg, don’t change the pixels in any way, and resave the jpeg using the same (or higher) compression, then the compression algorithm has already been applied, and no further compression will occur.
Jpegs use blocking compression. Say you have an image of a landscape, with, lets say grass in the foreground, and a blue sky in the background. You open the jpeg, and do adjustments on the foreground only, then resave the jpeg. There could be more compression added to the foreground (the area processed), but there would be no additional compression added to the background (which had no changes).
This is not to say that it isn’t better to save the image as an uncompressed file type. I save as psd, simply because that is the PS’s native file type, but tiff will work as well.
David, just to qualify my comments: I am not out to tell anyone how to practice their own craft. What I describe here is what I use in my own process and my personal reasons for doing so, in the hope that this is useful.
I'm a big believer in the end results so to me the only way to properly judge a photograph is to look at the final print. If it is successful, little else matters.
As for examples of the benefit of high-bit files, I have seen quite a few in my own work but I suspect they are not often seen because most people settle on one choice or the other and rarely compare the two in a variety of situations. I believe it's there if you look for it, though I have not seen many side-by-side comparisons.
Guy, through the limited interaction we have had both here and on the RMNP forum, I have the utmost respect for you, and do not believe that you push an agenda, and I apologize for my statements coming across as such. The comments were not directed at you specifically, but to the jpeg vs raw argument in general. Like I said, while I have not seen it here nearly as much, I do see the "it's raw or it's wrong" mantra pushed on many photo forums, complete with misinformation and regurgitated fallacies.
Since joining NPN, your work has been an awe inspiring inspiration, and it would be far from my place to question your abilities or workflow, since you obviously have mastered both. Ok, sorry, enough with the trite hero worship!
I used to work in 16 bit, but through becoming more proficient in my workflow, and learning much more about the processes, I switched back to 8 bit. What I really need to do is choose an image, and process it with exactly the same steps, one in 8 bit, one in 16 bit, to really see differences. Even better, do it with a few images, so see how different shading schemes have different outcomes.
Not gonna add more to the RAW / JPEG, but come on folks; Taking photography advice from Ken Rockwell is like getting economic advice from Bugs Bunny and Daffy Duck.
Can't be bothered to try top go back and answer all the different questions and such but if you want you camera to throw away data before you even get the image to the computer then so be it. I will continue in Raw for the fact that it doesn't try to decide what data to throw away as your camera does when converting a image to a jpeg in camera.
16 bit or 8bit matters not if you hav already let the camera throw away so much data
Discussion
29 repliesI am shooting wildlife, primarily birds, using a Canon 400 2.8 IS lens often with a 2x extender.
Do I need to just plunge into RAW and learn how to deal with it? What will be the big advantages if I do?
thanks in advance,
Peter
http://www.luminous-landscape.com/tutorials/understanding-series/u-raw-files.shtml
http://www.luminous-landscape.com/essays/rawtruth1.shtml
http://www.kenrockwell.com/tech/raw.htm
A Raw file is like having a B&W negative. If you have ever had or been in a B&w darkroom you know that you can process the same negative so many different ways to achieve different print looks. Same thing with the Raw file, You can convert that file to have many different looks, You can't do that with a jpeg without losing quality.
The jpeg the camera produces is the finished product, If your white balance, exposure is out the pic is ruined. Sure you should get those right in camera, but the jpeg then can't be changed like a Raw file can. The more manipulation you do to a jpeg the worse the image file will be.
I shoot raw because storage is cheap and I have basically a digital negative of the image
Until then, if you are more comfortable with JPEG, use it. If I remember correctly, your camera has an option for shooting both - use that. You'll get the JPEG you want, but will retain the "negative" for future use. The downside is you'll use a lot more memory, both in terms of cards and computer resources, but I think you'll find it a good choice 2-3 years down the road.
For a quick test that convinced me...open a jpeg and then the saturation dialogue box...tweak the saturation way up...farther than you usually would...it will look like a cartoon. Then, do the same thing with the raw file...tweak it way up and you will see it will hold it's composure much better.
The stuff the wingnut in the third article says about .jpg is true, but it also applies to shooting raw. Get your white balance and exposure right in the camera. Then you don't have to tweak either in your raw processor (Aperture). I understand where he's coming from there. Shooting a wedding, for instance, where I'll d/l several thousand images, I don't want to have to tweak exposure and white balance, or rescue shadows and highlights, on each one to proof them. Get it right the first time and it saves you a ton of PP work whether shooting raw or .jpg. Still, there's no reason to throw away information the sensor records by having the camera do a .jpg conversion on your image files.
It's easy to use. If you wish, I'll be happy to talk you through it on the phone. Just send me a note.
Should I upgrade my CS version to CS3?
thanks in advance!
thank you!
Working on a first generation jpeg will not cause any more “damage” to the image than working on an imported RAW. When you open a jpeg in PS, you are not working in jpeg. You are working in PS’s (or your software editor of choice) native file type. When you save it, it will save it as whatever file type was originally opened, unless you specify otherwise, or if it can’t save to the original file type (RAW). Something that goes hand in hand with this misconception is... “Working in jpeg will cause you to have more banding and artifacts, because jpeg is compressed.” No, working in jpeg will not cause this, because you do not work in jpeg, you work in PSD (in Photoshop). If you are getting banding or compression artifacts, it is being caused by your workflow, not by the simple fact you shot in jpeg.
If you shoot jpeg, you should open it in the image editing software, then save the file as a tiff, psd, or some other uncompressed file. Then you are never saving over the original jpeg, and it becomes the "digital negative." It would be poor workflow to ever save over the original file. It should always be archived just like a RAW.
Remember, just about any editing can be "destructive," since you are adjusting, converting, adding, or erasing pixels. Even if you start with a RAW file, you can end up with gradient banding and pixel artifacts. Your statement implies that editing an image that was shot with jpeg is more destructive than one shot with RAW. Editing an image that was shot using jpeg will be no more destructive if proper workflow is followed. Remember, as was said before, once you have imported (opened) the file in an image editing program, you are then working in that program's native file type.
Also, it is wrong to say that white balance cannot be adjusted in jpeg. Technically, it is a true statement, since “white balance” is not being adjusted, but it implies that hue cannot be corrected when the image was shot in jpeg.
What's funny is I shoot RAW, but defend jpeg! I swtiched to shooting RAW from jpeg a few years ago because I "supposed" to. I have become so comfortable with my RAW workflow that it doesn't add any more time or effort than shooting jpeg. I truly do like the flexibility of RAW, but it depends on my subject on how much I utilize it.
For landscape and nature, I tend to use RAW processing a lot. I adjust shadows and contrast, and use selective saturation, to try and acheive the Velvia look I am used to.
For portraits I tend to use less RAW processing, and do more of it in PS. Since I have my camera set to the white balance of my strobes, it comes out of camera pretty close, and I prefer to use selective color to fine tune the skin tones. If anything, I process the RAWs somewhat flat, with less contrast, to give me more latitude as I am masking and doing selective processing.
I always work with JPGs, since my camera can't shoot RAW, so I'm very familiar with the limitations imposed by starting with a pre-processed (saturation, sharpening, contrast already applied), 8-bit fiel. I'd definitely be shooting RAW if I could, mainly for two reasons : greater bit depth, and the ability to choose various saturation/sharpening/contrast levels than what my camera applies.
For what it's worth, I only posted the third link above because it's one of the very few I've seen arguing FOR JPG in any way. Seems to me, unless you're time constrained, or really hate doing the conversion, there isn't much reason not to use RAW.
JPG files can only hold 8-bits (256 shades) per channel. In raw mode your camera is capable of capturing 12-bits (4096 shades) per channel. This means that you are potentially throwing away thousands of color gradations by using JPG, which can make a big difference, especially in large prints where too few shades will lead to banding and other artifacts.
In JPG more you are also giving up the ability to make post-exposure adjustments (e.g. exposure compensation and color temperature) that in some cases can save an image that suffers from exposure or color issues. If you have blocked shadows or overexposed highlights in a JPG, there's nothing you can do to bring them back. With raw capture very often you can salvage details in these areas.
The "down side" is that raw leaves a lot more of the decisions to you and you will likely spend more time processing your images. In JPG mode the camera's computer makes a lot of these decisions for you and you don't get any say in how it makes them.
So you paid for a Porsche, now it's time to take it out of first gear...
Guy
Guy
Guy, you mentioned the one point that I truly think makes raw the file of choice. Raw has a lot more latitude than jpeg to bring back highlights that seem to be blown out. Of course, it is still probably necessary to open the raw file twice, once corrected for highlights and the other corrected for midtones or shadows, and then mask them in PS (or your editing program of choice). I can't speak for CS3, but so far there isn't a way to selectively correct for exposure. It is an all or nothing proposition.
Also, I still disagree with the implication that color correction is not possible in jpeg. While it is much easier to adjust the white balance on the raw file, it is again an all or nothing situation. On many of my shots, I much prefer to selectively adjust the colors using masks, etc.., instead of blanket changing the white balance.
I guess what I am saying is that PS (and many other editing programs) are capable of doing so much more than most people (including myself) use. Using your analogy, if the camera is a porshe, even if you take it out of first gear, it doesn't do you any good if you are driving on a residential street. You gotta open that bad boy up on the freeway!
Of course color correction is possible in JPG, but at that point you are starting off with less colors and gradation so you run the risk of banding in some color ranges (e.g. if you don't have enough shades of blue and you want to "cool down" the image.)
It is not always necessary to convert a file twice. In the conversion from raw, a tone-mapping curve is applied. If you let the camera use its standard curve in an image with subtle highlight (or shadow) detail, you may end up with multiple values mapped to pure white (or pure black) and lose that detail, whereas in manual conversion you can alter the characteristics of the curve (either using the convenient sliders for contrast, fill light etc.) or even by editing the curve itself which allows you to work on individual ranges of tones independent of others (e.g. you can reduce the contrast in the highlights and keep more detail there, without altering the mid-tones or shadows).
Guy
If you shoot .jpeg always use the highest quality that your camera will give you and never re-save a .jpeg after making corrections and adjustments as .jpeg. ALWAYS re-save as .tiff.
Every time you re-save a .jpeg as .jpeg you will degrade the quality because of re-compression. You can work on an open image file as much as you want as long as you can undo your mistakes in the "History" palette AND as long as you do not re-save as .jpeg.
Bob
If you shoot .jpeg always use the highest quality that your camera will give you and never re-save a .jpeg after making corrections and adjustments as .jpeg. ALWAYS re-save as .tiff.
Every time you re-save a .jpeg as .jpeg you will degrade the quality because of re-compression. You can work on an open image file as much as you want as long as you can undo your mistakes in the "History" palette AND as long as you do not re-save as .jpeg.
Bob
It does go back to my point, though, that to truly use the power of raw, then a good knowledge of PS (or your editing program...) is required. Not that I see it here as strongly as I do on other forums, but I think it is wrong to tell someone that because they are not shooting in raw, then they aren't doing it correctly. Some people like to spend their time shooting, not processing. I shot Velvia for years (as many of us did), without the benefit of LCD previews, RAW processing, or PS, and managed to get good photos. I think that forcing someone to use raw who doesn't feel comfortable on computers, doesn't understand or want to learn PS, or doesn't want to spend time processing is wrong. If the same attention to detail is given at the time of the shot that was given when using slide film, then incredible, ready to use images can be achieved using jpeg as the medium of choice.
Don’t get me wrong, I am not one of those who say using raw is simply a crutch, and good photographers don’t need it. I use raw, because I love the versatility is provides, but mostly I enjoy the processing. For me it is a lot of fun to get my hands on the picture. Maybe it’s because I never got to spend much time in a darkroom, and this is my way of making up for it!
I do think it is important for beginners to spend their time shooting. And not just shooting, but shooting correctly. They need to learn their equipment, learn composition, and learn how to properly expose an image. They need to learn how to use shutter speeds to their advantage, what aperture to use and why, and, in the digital age, what white balance to use. Jpeg can be very useful for this, because it forces them to learn to do it correctly in camera, instead of being able to correct for sloppy exposures in raw. Sometimes I think using raw right off the bat is like putting the cart before the horse. To use an analogy, it’s like teaching someone to do ornamental woodworking before teaching them to cut straight.
I really don’t care about the raw vs jpeg argument, per se. For me it is about pushing an agenda (which is not to say that is what you are doing, don’t take it wrong
I have already turned this into a huge post, but I still need to reply to your comments about 16 bit.
When you say you run the risk of banding using 8 bit, that also means you would be running the risk if the raw is imported at 8 bit. So this isn’t a raw vs jpeg issue, it is an 8 bit vs 16 bit issue. (Yes, still tying into the jpeg argument).
Going into the 16 bit discussion, like I said, I have yet to see an example of a final print that had a noticeable improvement in quality by being processed in 16 bit over one processed, with good workflow, in 8 bit. Yes, I have seen the pixel peeping 2X screen shots that show difference in the two, but how many people actually look at the 8x10 or 20x16 prints with a loupe? If you are processing for truly large prints, then I could be persuaded that it would be important to work in 16 bit.
Ok, that’s enough for now. I’m sure most everyone gave up on this post half way through it!
Shooting two images of the same scene and blending the exposure can be done with jpeg as well. This is not something that is raw exclusive.
I agree with your comments about jpeg workflow. The original file should be saved untouched, regardless of the file format used, just like raw, be it jpeg, tiff, or any other. Saving over the original file destroys any chance to go back to the original file and reprocess.
Saving a jpeg in of itself does not cause more compression, though. If you open a jpeg, don’t change the pixels in any way, and resave the jpeg using the same (or higher) compression, then the compression algorithm has already been applied, and no further compression will occur.
Jpegs use blocking compression. Say you have an image of a landscape, with, lets say grass in the foreground, and a blue sky in the background. You open the jpeg, and do adjustments on the foreground only, then resave the jpeg. There could be more compression added to the foreground (the area processed), but there would be no additional compression added to the background (which had no changes).
This is not to say that it isn’t better to save the image as an uncompressed file type. I save as psd, simply because that is the PS’s native file type, but tiff will work as well.
I'm a big believer in the end results so to me the only way to properly judge a photograph is to look at the final print. If it is successful, little else matters.
As for examples of the benefit of high-bit files, I have seen quite a few in my own work but I suspect they are not often seen because most people settle on one choice or the other and rarely compare the two in a variety of situations. I believe it's there if you look for it, though I have not seen many side-by-side comparisons.
Guy
Since joining NPN, your work has been an awe inspiring inspiration, and it would be far from my place to question your abilities or workflow, since you obviously have mastered both. Ok, sorry, enough with the trite hero worship!
I used to work in 16 bit, but through becoming more proficient in my workflow, and learning much more about the processes, I switched back to 8 bit. What I really need to do is choose an image, and process it with exactly the same steps, one in 8 bit, one in 16 bit, to really see differences. Even better, do it with a few images, so see how different shading schemes have different outcomes.
16 bit or 8bit matters not if you hav already let the camera throw away so much data