I have always used the 8 bit option in ACR. I shoot with a D200. Is there any advantage to using 16 bit with this camera. I tried it last night, and the image does seem a little "cleaner". Is this just my imagination, or something I should have been doing all a long? I also noticed that it won't let me save a file as a jpeg. Is there conversion I should be doing before saving? And It also slowed down my computer quite a bit. Thanks for any info on the subject.
Chris
Posted by Dan Borzynski on January 21, 2009 at 8:19 PM
Most of the time, working in 8 bit should be fine. 16 bit files give you more of a margin of error or latitude when processing. Where it really will make a dfference is in the following situations:
Large areas of smooth tonality - Generally. this will mean something like a bright, blue sky. Here, you run the risk of possibly causing banding issues in the sky as you make other adjustments.
Large, dark areas where you intend to bring up the tonality - You'll get less noise and less banding, and the color should look more natural.
When you know that you will be doing a lot of post-processing on the image, especially with curves, highlights, shadows, contrast, etc.
The "cleaner look, depending on the individual image, may be because of a lack of noise in darker toned areas. The difference between 8 bit and 16 bit means a lot more space for colors to be selected by the camera and processing software. At the lower half of the color range, 8 bit gives you 2^4, or 16 tonal choices, or 16^3 = 4096 colors. In 16 bit, it is 2^8, or 256 tonal choices, or 256^3 = 16,777,216 color choices. That is a huge difference! Granted, a lot of them are not going to be very differentiable to the eye, if at at, but at least the space is there.
Now, with regards to your computer slowing down, that really shouldn't happen for individual images. If you are working with one image at a time, most current computers really shouldn't even notice the difference.
Hmmm, you mention you use ACR. I haven't used that program - depending on what is going on behind the scenes, maybe that program just doesn't like 16 bit files
The current standard for jpegs is that they are 8 bit only files, with no capacity for layers or 16 bit files.
You really don't want to be using jpegs for the working copy of the image. The are a lossy format, which means that every time you open the image, work on it, and save it, there is information lost. You have no control over this loss of information. For a lot of images, this isn't a huge issue, but if the archived version of the image is a jpeg you have worked on and later you need to come back to it and do something extreme to it (such as trying to print at 16x24), you might run across quality issues.
With the price of large hard drives and solid editting, you might be better served by working with RAW and TIFF or PSD files as your editting and long term storage file formats.
Posted by John Mumaw on January 21, 2009 at 8:53 PM
As Dan noted, the advantage of 16 bit images over 8 bit images lies in the fact that you have much more 'headroom' in which to perform numerous and/or heavy edits before image quality starts to degrade. I seriously doubt you'll notice any difference between an unedited 8 bit file and an unedited 16 bit file just by looking at them.
16 bit images can indeed slow your computer down if it's an older computer. That's because the files are twice as large.
I'm able to save 16 bit images as JPEG files out of ACR by using the 'Save Image' button on the lower left of your screen . I would never do it because it makes no sense, but I just tried it and it works.
You can find the technical details of the difference between 8 bit and 16 bit image files in a fairly simple tutorial here.
Posted by Ed Cordes on January 21, 2009 at 10:47 PM
I agree with all that was said above. However, if you use ACR to load 16 bit files to PS then process and save in 16 bit you will know that you are saving the highest quality file you can. When you need 8 bit jpeg for posting or other reasons go ahead and convert and save as another file name. That way you will always have the best file for future use.
BTW, I save all processed images as TIFF as it is lossless compression. When I need jpeg I convert.
Thank you for all the info so far. I do always save my processed files as TIFFS, and then a web sized jpeg for posting here. I just didn't have the jpeg option with the "save as" method of saving a file. I didn't try "save image" method. I don't see the point of trying to save a file that's twice as big, down to within the posting limits here either. If the 16 bit Tiffs will produce a better print, then I'm all for that!
Thanks,
Chris
Posted by Gene Lugo on January 22, 2009 at 6:07 AM
Something worth considering - are you using compressed or uncompressed NEF files? If you are using compressed NEF, be aware that the compression occurs in the high values, or highlights, which is exactly where you don't want it.
The rationale behind processing in 16 bit vs 8 bit has to do with images that contain colors and saturation points that are at the limits of the working color space. Looking at sRGB vs Adobe RGB vs ProPhoto RGB, each subsequent space represents a considerable expansion, not just in the number of colors that can be manipulated, but their saturation as well. By default, sRGB is 8 bit, while Adobe RBG can be either 8 or 16, and ProPhoto RGB is 16 bit. If a RAW image is viewed and processed in a small color space in 8 bit, you are likely to hit hard limits in color and saturation. Flowers with deep blues and purples and scenes with greens and yellows are particularly impacted. I found this out the hard way, when shooting orchids - some of the results were pale comparisons (no pun) to the original flowers.
Now, all of this is moot if your final output device has a similarly small color space (gamut). If you are printing to an all-in-one printer with three colors and black, it probably won't make much difference what you do - sRGB, 8 bit, jpg file format. However, if your target output is a glicee print on a fine art paper, you will want to be able to work with and view your image in the widest gamut and the highest bit depth possible in order to closely match the capability of the print process and medium. On scenes that have a wide color range with deep colors, the difference between 8 bit sRGB and ProPhoto RGB is startling.
So, some will argue that your working bit depth has to do with your final intent. If you are not sure what that is when you take the picture, then you should use uncompressed NEF files, Adobe RGB in 16 bit at the minimum, but preferably ProPhoto RGB as a workspace. This way, if you have to downsize the workspace or bit depth for a particular destination you will end up with a more reasonable result. At least that's how I understand it - and I apologize in advance if there is any inaccuracy.
Posted by Ed Cordes on January 22, 2009 at 10:17 AM
I agree with Gene. Use ACR to send ProPhoto 16 bit to PS the process and save. Convert when you need it. On my printer Canon Pixma Pro 9500 with 10 pigment inks it makes a large difference in shadow/highlight detail as well as color depth as well as smooth color transitions.
Posted by David Bostedo on January 22, 2009 at 12:53 PM
The rationale behind processing in 16 bit vs 8 bit has to do with images that contain colors and saturation points that are at the limits of the working color space.
One clarification - whether a colorspace (sRGB, Adobe RGB, Prophoto RGB) is 8 or 16 bit does NOT affect the gamut or colors that can exist in the image. What it affects is the fineness of the gradations that can exist. And if you have a very large colorspace in 8 bit, you'd tend to get posterization in smooth transitions. That why the large gamut spaces like Prophoto are 16 bit only. But if you are in AdobeRGB, the saturation points will be the same in 8 or 16 bit.
Posted by Kurt Budde on January 28, 2009 at 12:09 AM
To save as a jpeg from photoshop -
First convert from your preferred color space to sRGB (edit, convert to profile)
Second, image>mode>8bit
Discussion
8 repliesChris
Large areas of smooth tonality - Generally. this will mean something like a bright, blue sky. Here, you run the risk of possibly causing banding issues in the sky as you make other adjustments.
Large, dark areas where you intend to bring up the tonality - You'll get less noise and less banding, and the color should look more natural.
When you know that you will be doing a lot of post-processing on the image, especially with curves, highlights, shadows, contrast, etc.
The "cleaner look, depending on the individual image, may be because of a lack of noise in darker toned areas. The difference between 8 bit and 16 bit means a lot more space for colors to be selected by the camera and processing software. At the lower half of the color range, 8 bit gives you 2^4, or 16 tonal choices, or 16^3 = 4096 colors. In 16 bit, it is 2^8, or 256 tonal choices, or 256^3 = 16,777,216 color choices. That is a huge difference! Granted, a lot of them are not going to be very differentiable to the eye, if at at, but at least the space is there.
Now, with regards to your computer slowing down, that really shouldn't happen for individual images. If you are working with one image at a time, most current computers really shouldn't even notice the difference.
Hmmm, you mention you use ACR. I haven't used that program - depending on what is going on behind the scenes, maybe that program just doesn't like 16 bit files
The current standard for jpegs is that they are 8 bit only files, with no capacity for layers or 16 bit files.
You really don't want to be using jpegs for the working copy of the image. The are a lossy format, which means that every time you open the image, work on it, and save it, there is information lost. You have no control over this loss of information. For a lot of images, this isn't a huge issue, but if the archived version of the image is a jpeg you have worked on and later you need to come back to it and do something extreme to it (such as trying to print at 16x24), you might run across quality issues.
With the price of large hard drives and solid editting, you might be better served by working with RAW and TIFF or PSD files as your editting and long term storage file formats.
16 bit images can indeed slow your computer down if it's an older computer. That's because the files are twice as large.
I'm able to save 16 bit images as JPEG files out of ACR by using the 'Save Image' button on the lower left of your screen . I would never do it because it makes no sense, but I just tried it and it works.
You can find the technical details of the difference between 8 bit and 16 bit image files in a fairly simple tutorial here.
BTW, I save all processed images as TIFF as it is lossless compression. When I need jpeg I convert.
Thanks,
Chris
The rationale behind processing in 16 bit vs 8 bit has to do with images that contain colors and saturation points that are at the limits of the working color space. Looking at sRGB vs Adobe RGB vs ProPhoto RGB, each subsequent space represents a considerable expansion, not just in the number of colors that can be manipulated, but their saturation as well. By default, sRGB is 8 bit, while Adobe RBG can be either 8 or 16, and ProPhoto RGB is 16 bit. If a RAW image is viewed and processed in a small color space in 8 bit, you are likely to hit hard limits in color and saturation. Flowers with deep blues and purples and scenes with greens and yellows are particularly impacted. I found this out the hard way, when shooting orchids - some of the results were pale comparisons (no pun) to the original flowers.
Now, all of this is moot if your final output device has a similarly small color space (gamut). If you are printing to an all-in-one printer with three colors and black, it probably won't make much difference what you do - sRGB, 8 bit, jpg file format. However, if your target output is a glicee print on a fine art paper, you will want to be able to work with and view your image in the widest gamut and the highest bit depth possible in order to closely match the capability of the print process and medium. On scenes that have a wide color range with deep colors, the difference between 8 bit sRGB and ProPhoto RGB is startling.
So, some will argue that your working bit depth has to do with your final intent. If you are not sure what that is when you take the picture, then you should use uncompressed NEF files, Adobe RGB in 16 bit at the minimum, but preferably ProPhoto RGB as a workspace. This way, if you have to downsize the workspace or bit depth for a particular destination you will end up with a more reasonable result. At least that's how I understand it - and I apologize in advance if there is any inaccuracy.
One clarification - whether a colorspace (sRGB, Adobe RGB, Prophoto RGB) is 8 or 16 bit does NOT affect the gamut or colors that can exist in the image. What it affects is the fineness of the gradations that can exist. And if you have a very large colorspace in 8 bit, you'd tend to get posterization in smooth transitions. That why the large gamut spaces like Prophoto are 16 bit only. But if you are in AdobeRGB, the saturation points will be the same in 8 or 16 bit.
First convert from your preferred color space to sRGB (edit, convert to profile)
Second, image>mode>8bit
Then you can save as