16 or 8 bits


Status
Not open for further replies.

saberlancer

Senior Member
Jan 26, 2002
522
1
18
Singapore
Hi guys,
Was wondering if the pros in here use 16bit throughout their workflow?

Using 16bits really makes the file sizes huge. Not sure if the improvements are worthwhile.

Pls comment.thk you.
 

With Photoshop CS, you can use 16-bit throughout the workflow. I capture photos mostly in large/fine jpeg which is 8-bit. Before I start my editing, I convert the file to 16-bit. (go: Image -> Mode -> 16-bit) I get fewer gaps in the histogram (combing) and therefore less image degradation during editing.

I save the edited file usually in 16-bit 'psd' format. To save as jpeg for webposting or email, you need to convert back to 8-bit first.

Robert
 

Thks Robert,
But the 16bit info is not there in the first place so converting from 8 to 16bit,how does that work? Isn't it just resampling?
 

For my case, it its casual snaps I leave them in jpg and burn em into a CD when I get home. This will be 8bit.

If I am doing something serious, I would shoot RAW and convert to 16bit. You get less image degradation. You can clearly see this when doing a levels in PS CS. The 8bit file will show gaps (ie no image detail) while the 16 bit graph is smoother.

If I already shot in jpg and I need 16bit I will go with robert's workflow which is pretty good too. ;)

After processing, I usually save in psd.
 

Thks Robert,
But the 16bit info is not there in the first place so converting from 8 to 16bit,how does that work? Isn't it just resampling?
You are right that the original info is 8-bit. However, doing your editing in 16-bit will result in less degradation during editing. It's a slick trick!

Try it and see. Do your editing in 8-bit, then again in 16-bit. Compare the histograms.

Does it make a huge difference in the final photo...not really. However, it is so easy to do...why not get the best quality possible.

Robert
 

I would love to do all in 16bit,except that the file/scratch disk really balloons.
Done lots of research, general consensis seems to be that there is little or no noticable image degradation.

So I was curious what people are using. ;)
 

Personally, I used to process in 8-bit and save in 8-bit. When I tried to save in 16-bit I tend to get yellowish casts so ...

Anyway now I process in 12 bits then go to 16 bits in CS then finally out put as TIFF/JPEG depending on the usage of the images :)
 

For me, I do all in 8-bit. Unless the photos need to be enlarged, then I'll process in 16-bit.
 

So are there any benefits in editing in 16bits, then converting to 8 bit to save to Jpeg?

Are there significant changes compared to editing in 8 bits and saving to jpeg?

Also, why is there siginficantly more noise when editing with RAW files?
 

Questions answered above. Why not try it both ways yourself and compare?

To repeat...in 16-bit there will be less image degradation during editing. However, it might not be noticeable in the final image/print.
 

Well i see no difference whatsoever. I mean, who really cares about that little difference. When printing 8x10s, 4x6s, 12x18s, how much better will the picture turn out?

Don't really think it'll make a diff.

What about the RAW noise question?
 

i cant see the difference between 8 or 12 or 16 bits....not to my amatuer eyes anyway....:(
 

The diff is when you're doing major contortions at the extreme ends of the curve, trying to rescue shadows or (more rarely) highlights. Try both 8 and 16bits with a picture where the exposure is ok but there's a patch of shadow where you're trying to 'boost'. Should see a diff btw 8 and 16bits easily.
 

What ST1100 says is right. If you do an extreme (very steep) curve, there will be less visible contouring effect in 16-bits compared to 8-bits in the region of the steep part of the curve. But how often do we perform extreme curves on our images?

For normal photo editting, yes you will see differences in the histograms (combing effects) when you do it in 8-bits and 16-bits, this means image degradation in the images, but can our eyes notice any differences?

Can you tell any differences between a 7-bit grayscale image with an 8-bit grayscale image?

So like in photography, whether to edit in 8-bit or 16-bit depends on what you want. There will be a trade-off between processing time and final result. The question you have to ask is whether the improvement (in any from your eyes) is worth the additional computational processing? Saving in 16-bit for future usage may be advisable, but remember, not only would you take more processing power, you would need more hard-disk space as well!
 

Status
Not open for further replies.