16-bit TIFF

[InskiP]

I thought I saw somewhere that when converting RAW images you should convert to
16-bit TIFF, correct? But these images are SO BIG and take forever to edit. Are there certain edits that should be done at 16-bits? At what point is it best to convert to 8-bit, or should I???

[rent]

for archiving purposes and printing, keep it at 16-bit if possible. if you need to use some filters in PS you would have no choice but to convert to 8-bit. for web display or emailing, 8 bit is all you'll need.

-alex

[Jesper]

RAW files contain 12 bits per channel of data, so if you want to take advantage of all of the data that's in your RAW file, you should convert it to a 16 bit per channel TIFF file. (Note that the TIFF format only supports 8 or 16 bits, not 12 bits).

If you convert your RAW file to an 8 bit per channel TIFF, you're throwing away some of the data, and you'll have less dynamic range.

Especially edits like stretching the contrast or correcting the exposure of underexposed images should be done in 16 bits - you'll have more data to work with and you won't see strange things like banding as quickly as you would with an 8 bit image.

You can save your final version in 8 bits per channel - for display on a monitor or printer, 16 bit won't give you any advantage - monitors and printers can't display more than 8 bits per channel anyway.

[jfrancho]

I convert my RAW files directly to 16 bit PSD files. I don't have any clue whether there is any +/- to this, though. It is just the way I've always done it.

[robertwgross]

I agree with Jesper.

The 16-bit file is a good place to do major stuff like major exposure correction, tone curve correction, and that sort. Once all of the major stuff is done, then you can convert to 8-bit and do the rest of the minor stuff, like the Waldo filter, sharpening, noise reduction, cropping, and a few others.

---Bob Gross---

[InskiP]

It sounds like you're saying that once I do the major edits, I can knock it down to 8-bits, correct?
I don't need to save a 16-bit file? What if after I convert to 8-bits, I realize I want to make another edit, can I change it back to 16-bit, then back to 8-bit, or not? It's just that these files are so big, like 46MB, I'd rather not store too many of them if I don't have to. This is basic stuff I'm not sure about???

BTW, I thought noise reduction was a first step???

Thanks

[lostdoggy]

I convert my RAW files directly to 16 bit PSD files. I don't have any clue whether there is any +/- to this, though. It is just the way I've always done it.

I only save in PSD when I want to retain the layers otherwise there aren't much avantage over TIFF16.

[jfrancho]

Once you convert to 8 bit, the extra data is lost, therefore I don't think there is a benefit to goign back to 16 bit. I could be wrong, I just don't think the capability is there to reinterpolate that data. Is reinterpolate a word?

[jfrancho]

I only save in PSD when I want to retain the layers otherwise there aren't much avantage over TIFF16.I think you have a capability to preserve the layers in TIFF, too. I just thinks it makes sense to work in the application's native filetype. I see it as the difference between a CSV and an XLS - they both contain the data, but the XLS has all the formatting, formulas, VBA, etc.

[InskiP]

Well, I'm still using PS7 which I realized doesn't support layers w/ 16-bit files..bummer. I like using adjustment layers. Are there any MUST corrections that should be made in 16-bit before I convert to 8-bit?? What did you guys do before CS(2)???

[tim]

I save the RAW file, it's much smaller than a TIFF or PSD. If the RAW/CR2 format looks like it's going to be dropped in future i'll use the adobe convertor tools to convert to whatever format I need then. By then I guess we'll all have terrabyte hard drives and the size won't be important.

[EOSAddict]

It depends how much PP you do after RAW conversion. Remember the RAW file still has ALL the data in and your converter should hold the 'recipe' to allow you to regenerate a TILL/JPEG anytime. I save to JPEG so I can send/show people easily. I agree with Tim, maybe we will need to convert to DNG format!

[InskiP]

Yes, I will ALWAYS save the RAW version. But I'm torn/confused by what I've read when it comes to PP workflow. I've read that you should work w/ 16-bit files. On the other hand I've been advised to use adjustment layers, basically work w/ layers. But, I don't believe this is possible w/ PS7. So, which is more important? Are there certain adjustments that should absolutely be done at 16-bit before converting and others that aren't as important? Just curious what the workflow used to be before CS?

[InskiP]

You can save your final version in 8 bits per channel - for display on a monitor or printer, 16 bit won't give you any advantage - monitors and printers can't display more than 8 bits per channel anyway.

for archiving purposes and printing, keep it at 16-bit if possible

These are kind of contradictory. Is one reffering to inkjet and the other Pro? Do ALL printers only see 8-bits? What about when using mpix or whcc for printing?

If you knock it down to 8-bits AFTER editing isn't that the same as starting w/ 8-bits
because you're throwing that extra info away?

[robertwgross]

It sounds like you're saying that once I do the major edits, I can knock it down to 8-bits, correct?
I don't need to save a 16-bit file? What if after I convert to 8-bits, I realize I want to make another edit, can I change it back to 16-bit, then back to 8-bit, or not? It's just that these files are so big, like 46MB, I'd rather not store too many of them if I don't have to. This is basic stuff I'm not sure about???

BTW, I thought noise reduction was a first step???

Thanks

It's that "I realize I want to make another edit..." thinking that makes problems when you are trying to keep file sizes manageable.

I shoot RAW. I do some tweaks at the conversion to TIF. I archive the RAW files away onto CDR, so if I ever need to go back to the original, I can. However, TIF is the main file type, at least for me. If I edit in Photoshop, then sometimes I am working in layers, but once it is done, I flatten the layers into a TIF again.

Noise reduction is a first step for some users, but it doesn't have to be. If you keep the ISO low enough and shoot with good exposures, you might not have to do noise reduction at all.

---Bob Gross---

[robertwgross]

Are there any MUST corrections that should be made in 16-bit before I convert to 8-bit??

Some noise reduction programs will not run in 16-bit mode. Some filters will not work in 16-bit mode.

As I stated previously:
The 16-bit file is a good place to do major stuff like major exposure correction, tone curve correction, and that sort. Once all of the major stuff is done, then you can convert to 8-bit and do the rest of the minor stuff, like the Waldo filter, sharpening, noise reduction, cropping, and a few others.

To be more specific:
I tend to do noise reduction in 8-bit mode first. The Waldo filter is almost last. Sharpening tends to be last.

What did you guys do before CS(2)???

Corel Photopaint 8. It still works, and it is a hell of a lot faster to launch than Photoshop.

---Bob Gross---

[tzalman]

If you knock it down to 8-bits AFTER editing isn't that the same as starting w/ 8-bits
because you're throwing that extra info away?

Sure some data is lost or at least made less precise, But if you're using all the 65000+ levels available in 16 bit it will reduce down to 256 levels in 8 bit. If you had edited in 8 bit you might end up with 200 levels.

Another reason for 16 bit. If your working space is ProPhotoRGB or WideGamutRGB you have to use 16 bit. At one time there were people who claimed AdobeRGB needs 16 bit also. although lately thet seem to have been silent.

[JakeC]

I do all my editing in 16bit using noise ninja and CS2. I keep both the RAW files and associated recipes from RSE. I'll save during editting as a .PSD and again just before final sharpening so I've got my 'full resolution copy' Hardrives are so cheap these days and my P4 3.2E system is fast enough to keep up. 150-250mb Panoramas in 16bit get slow with only 512mb RAM but 40-50mb files are fine.

[InskiP]

These are kind of contradictory. Is one reffering to inkjet and the other Pro? Do ALL printers only see 8-bits? What about when using mpix or whcc for printing?


What about this question ;)

Ooops missed part of the post. Anyway one person said keep it 16-bit for archiving and printing. Another person said you can save your final version in 8 bits per channel because monitors and printers can't display more than 8 bits per channel anyway, so 16-bits wont give you any advantage (after editing). Okay now read above question/quote.

BTW What's a Waldo filter?

[jfrancho]

What about this question ;)

Ooops missed part of the post. Anyway one person said keep it 16-bit for archiving and printing. Another person said you can save your final version in 8 bits per channel because monitors and printers can't display more than 8 bits per channel anyway, so 16-bits wont give you any advantage (after editing). Okay now read above question/quote.

BTW What's a Waldo filter?I have various workflows based on intended output. They all start the same: backup, archive (CD-R) and process the raw file (I do my crops here too), save as a 16-bit psd (Adobe RGB), make adjustments in PS - this usually just sharpening, and maybe slight burn and dodge. I also do my monochrome conversions here. If there is some drastic color issue, I'll start back at the raw file and try to fix it there, but sometimes slight adjustments are handled in PS on the psd file. All PS adjustments are done via layers and layer sets. Once this is done, I consider this my "saved version of the picture." It gets backed up in two places and burned to a CD-R. Now the workflows diverge based on output. Whatever photolab I am using dictates the format. mpix.com takes 16 bit tiff files (sRGB); EZPrint takes jpegs (sRGB), same for the grocery store; my friend's Epson will print seemingly whatever I want; etc. Basically, whatever the lab says will give you the best results is what you want to follow. If they are for display over the web, obviously the resolution is decreased, compression increased a little, and converted to sRGB. I never use the "Save for the Web..." though I may try it out, since many here have great results with it. If they will be put in a slide show, I use a different action that works best for that program. If they are put on a disk, to give a customer, it is what they want. I don't know if this answers your question, but basically, the "pictures" get saved/archived as a layered 16 bit psd (Adobe RGB). This is the file that is most important to me. I make sure that "Maximize Compatability" is enabled in PS, and I trust Adobe that I will be able to go back and re-edit/re-output the picture for display. This seems like a good process, maybe someone else has a better way.

[EOSAddict]

Sounds like a great workflow but you must have a whole houseload of storage for non-flattened 16-bit psd files!!

[jfrancho]

Sounds like a great workflow but you must have a whole houseload of storage for non-flattened 16-bit psd files!!300GB RAID 1 (mirrored). I'm going to have to address space sometime spring next year, at the current shooting rate.

[InskiP]

It seems like a lot of the online labs take sRGB, so why would I want to shoot and edit in ARGB?

How do you backup? :o Isn't your archive CD-R's a backup?

[robertwgross]

What's a Waldo filter?

There is a children's game called Where's Waldo. The child looks all over the image to find the face of Waldo hidden in tree leaves or something.

A modern photographer shoots a scene, and he reviews the image on the camera display. It looks good.

Once it gets to the computer, and he starts checking it more closely, he discovers something that does not belong in the image. Some stranger's face, or a hand, or a street sign. There will be something that detracts from a perfect image.

You've found Waldo. Now clone him out. The object is to clone it out so perfectly that nobody can tell you fooled with it.

---Bob Gross---

[jfrancho]

It seems like a lot of the online labs take sRGB, so why would I want to shoot and edit in ARGB?

How do you backup? :o Isn't your archive CD-R's a backup?Adobe RGB is a larger gamut. I use it for the same reason I use 16 bit color. Many will say that the eye can't tell, but I feel whatever you can do to maximise quality without too much impact to time an effort is better. CD-R is really just storage, or archiving. The CDs sit with my pc. I backup to a file server at work, which is also backed up to tapes, which are stored off-site. A backup plan must include some form of continuity if the primary site (my home office) is totally destroyed. I could just as easily store the CDs off site, but I go back to them frequently enough that I consider them an archive.

EDIT: To further elaborate, I backup my entire collection of raw+sidecars, psd files, and jpegs (pretty much the entire working directory) to the server at work every few days, therefore the "backup" is a true backup. This means if something happens to the files, I can back up to the last good version. Any incremental changes are captured in the backups. If I went to the CDs, I would lose the incremental changes made to the files since the time I burned them.