Why 16 BIT RAW conversion?

[Markus61142]

Hi,

i was reading through a lot of RAW workflow docs and was wondering why I should convert my photos into 16BIT format? Whats so wrong about 8 BIT? File size just gets increased and i cannot really see any major difference?

please let me know your experience,

thanks, markus

 

[DonCohen]

Hi Markus,

Hi,

i was reading through a lot of RAW workflow docs and was wondering
why I should convert my photos into 16BIT format? Whats so wrong
about 8 BIT? File size just gets increased and i cannot really see
any major difference?

please let me know your experience,

You may have done some reading here, but here's my take on this issue:

http://www.dlcphotography.net/RawFormatWorkflow/RawFormatHomePage.htm#anchor341628

Also, and more significantly, here is the link to Bruce Fraser's (Photoshop guru) comments on this subject, which I reference at my site:

http://www.creativepro.com/story/feature/7627.html

Hope this helps.

--
Don
http://www.dlcphotography.net

 

[Melanie Kipp]

You're most likely goingto be some slight tweaking, or fine tuning, after you pull the file into photoshop. Using 16 bit data will minimize information loss from your file... do a test:

-Select a raw image file, convert it as 16 bit. Make a levels adjustment to this file in photoshop. Then convert to 8 bit.

-Select another file ( or the same one - if the same one you may need to give the first version a new name first ) and convert the file as 8 bit. In photoshop make the same levels adjustment to the 8 bit version. Save it.

-Now open both. Look at the histogram of the 16 bit converted file. Look at the 8 bit converted file's histogram.

The 16 bit converted version's histogram will be intact. Nice and solid.

The 8 bit's histogram will look like a comb, the spaces representing data that was thrown out, lost, during the levels tweaking.

Hi,

i was reading through a lot of RAW workflow docs and was wondering
why I should convert my photos into 16BIT format? Whats so wrong
about 8 BIT? File size just gets increased and i cannot really see
any major difference?

please let me know your experience,

thanks, markus

--



http://www.caughtintimephotography.com
http://www.pbase.com/melaniekipp

 

[Derekasaurus Rex]

There are lots of threads on this. Here is a recent one:

http://forums.dpreview.com/forums/read.asp?forum=1032&message=9412685

And my answer:

http://forums.dpreview.com/forums/read.asp?forum=1032&message=9412960

Hi,

i was reading through a lot of RAW workflow docs and was wondering
why I should convert my photos into 16BIT format? Whats so wrong
about 8 BIT? File size just gets increased and i cannot really see
any major difference?

please let me know your experience,

thanks, markus

 

[Steven Noyes]

I have wondered the same for awhile as well. Since most (all) pint output devices can't make use of 8 bit AND you use the RAW conversion software to do all of your curve/saturation/contrast/color adjustments.... What is the point of a 16 bit conversion.

Unless you use a linear workflow. But that is a whole another ball game.

Steven

Hi,

i was reading through a lot of RAW workflow docs and was wondering
why I should convert my photos into 16BIT format? Whats so wrong
about 8 BIT? File size just gets increased and i cannot really see
any major difference?

please let me know your experience,

thanks, markus

--
---
New and Updated!!!
Summer 2004: http://www.pbase.com/snoyes/
an_image_a_week_summer
Spring 2004: http://www.pbase.com/snoyes/spring_2004

 

[Bob Watt]

Markus,

If you convert to 16B tif and work the image
over in 16B in PS then save as 8B, you will find that the colors
achieved using levels and the sharpness possible from USM
are a bit better than they would have been starting
from an 8 bit conversion. Not a tremendous difference
but it is noticeable on some kinds of images.

Bob Watt

 

[DonCohen]

Hi Steve,

I have wondered the same for awhile as well. Since most (all) pint
output devices can't make use of 8 bit AND you use the RAW
conversion software to do all of your
curve/saturation/contrast/color adjustments.... What is the point
of a 16 bit conversion.

Photoshop guru Bruce Fraser comments on this here:

http://www.creativepro.com/story/feature/7627.html

And this pre-dates PSCS, where many more PS tools can be used with 16-bit TIF's.

My personal workflow uses C1 to get the exposure 'in the ballpark' but I find I can adjust curves and levels much quicker, and more precisely, in Photoshop. I can also use the History brush to maximize the dynamic range shown in the image. Making all of these adjustments in 16-bit mode keeps the histograms 'smoother' with better tonality, according to Fraser and others.

Yes, you do have to go back to 8-bit eventually, but that is done as the very final step, after all the real processing has been done. The way I think about it is that I want to be able to work with the full range that the camera's sensor captures through all the steps of editing.

If you're not using Photoshop, and relying on C1 for all editing, then I would agree that there's no point in 16-bit TIF. But I think a valid argument can be made to use 16-bit TIF if there will be significant editing done in Photoshop.

My $.02 worth, anyway.

--
Don
http://www.dlcphotography.net

 

[Steven Noyes]

If you use C1 well, it has amazing control over curves and exposure (much better than PS IMO). Also, Bruce's article (if I remember) does some really huge shifts in the exposure/curves to demonstrate his point.

But in the end, unless you are doing huge shifts (and that you should do in ACR or C1) the differences between 8 bit output and 16 bit output are insignificant.

Steven

I have wondered the same for awhile as well. Since most (all) pint
output devices can't make use of 8 bit AND you use the RAW
conversion software to do all of your
curve/saturation/contrast/color adjustments.... What is the point
of a 16 bit conversion.

Photoshop guru Bruce Fraser comments on this here:

http://www.creativepro.com/story/feature/7627.html

And this pre-dates PSCS, where many more PS tools can be used with
16-bit TIF's.

My personal workflow uses C1 to get the exposure 'in the ballpark'
but I find I can adjust curves and levels much quicker, and more
precisely, in Photoshop. I can also use the History brush to
maximize the dynamic range shown in the image. Making all of these
adjustments in 16-bit mode keeps the histograms 'smoother' with
better tonality, according to Fraser and others.

Yes, you do have to go back to 8-bit eventually, but that is done
as the very final step, after all the real processing has been
done. The way I think about it is that I want to be able to work
with the full range that the camera's sensor captures through all
the steps of editing.

If you're not using Photoshop, and relying on C1 for all editing,
then I would agree that there's no point in 16-bit TIF. But I
think a valid argument can be made to use 16-bit TIF if there will
be significant editing done in Photoshop.

My $.02 worth, anyway.

--
Don
http://www.dlcphotography.net

--
---
New and Updated!!!
Summer 2004: http://www.pbase.com/snoyes/
an_image_a_week_summer
Spring 2004: http://www.pbase.com/snoyes/spring_2004

 

[Derekasaurus Rex]

But in the end, unless you are doing huge shifts (and that you
should do in ACR or C1) the differences between 8 bit output and 16
bit output are insignificant.

I agree with your point, but I find there is one important exception: local adjustments. I often find myself applying a digital ND or tweaking something locally using a mask or selection. I find that 16-bit output can be very useful in these circumstances.

 

[DonCohen]

Steve,

If you use C1 well, it has amazing control over curves and exposure
(much better than PS IMO). Also, Bruce's article (if I remember)
does some really huge shifts in the exposure/curves to demonstrate
his point.

But in the end, unless you are doing huge shifts (and that you
should do in ACR or C1) the differences between 8 bit output and 16
bit output are insignificant.

My frustration with C1 in this regard is that you have to use the mouse to adjust curves and levels, and I find this very inefficient and imprecise. I have Photoshop set up so that I can make all my adjustments with the keyboard, and I find this both extremely fast and precise. Plus, I can toggle before/after to more accurately fine-tune any adjustments.

Combined with the benefit of 'local' adjustments to curves/levels, made possible by use of a Wacom Pen and the History Tool, for my needs, Photoshop works much better.

I would agree that the larger the adjustments that need to be made, the more benefit there will be with using 16-bit files. For my own needs and workstyle, I find it both quicker and more precise to adjust curves/levels in Photoshop, and since it doesn't "cost" me anything to do this using 16-bit TIF's, and there are at potential benefits to be had, it just makes sense to me to do it this way.

I'm not trying to convince anybody this is the "right" way to do it, but since there are at least theoretical benefits, and since my personal workstyle is dramatically quicker and more precise using Photoshop, it just works well for me.

--
Don
http://www.dlcphotography.net

 

[Derekasaurus Rex]

My frustration with C1 in this regard is that you have to use the
mouse to adjust curves and levels, and I find this very inefficient
and imprecise. I have Photoshop set up so that I can make all my
adjustments with the keyboard, and I find this both extremely fast
and precise. Plus, I can toggle before/after to more accurately
fine-tune any adjustments.

Combined with the benefit of 'local' adjustments to curves/levels,
made possible by use of a Wacom Pen and the History Tool, for my
needs, Photoshop works much better.

I agree. Like you I am not trying to convince anyone that my way is better, but 16-bit + PS is the heart of my workflow. In fact, I do not make any adjustments during conversion other than WB when necessary.

The ability to make local adjustments and have keyboard control is invaluable to me. Plus I have PS plugins that do a much better job with noise reduction (Neat Image) and sharpening (PhotoKit Sharpener) than C1 or any other converter I have tried.

And since most of my work is done in PS, 16-bit is the only way to go.