8 bit vs 16 bit Tiffs

BF
Posted By
Brendon Fewel
Aug 26, 2003
Views
849
Replies
15
Status
Closed
Help me get this straight please. How much information is lost if I save my raw files (from a Canon s45) as 8 bit tiff vs 16 bit tiff (so I can edit them in PS)? What difference is there; will I miss it? Should I worry about it?

Anyone know? Thanks!

Brendon

How to Master Sharpening in Photoshop

Give your photos a professional finish with sharpening in Photoshop. Learn to enhance details, create contrast, and prepare your images for print, web, and social media.

P
Phosphor
Aug 26, 2003
That depends entirely on the image in question.

If you had some shadow detail or hilight detail that you wanted to bring out later (by adjustments of filters) – then you could lose quite a bit.

If you have the RAW file around for archival purposes, and the image is properly exposed — then you may get by with 8 bits/channel.
BF
Brendon Fewel
Aug 26, 2003
If you can loose so much information, why does PS only let you work with 8 bit tiff images? I thought that people who used PS were heavy into photography; why would they use a program that doesn’t let you use the highest quality images?
JH
Jake Hannam
Aug 26, 2003
I think it is for a couple of reasons. First, many of the effects do not work with 16-bit files (Adobe should be able to fix this but has not so far). Second, 16-bit files are much larger than 8-bit files so they are harder to manipulate. They take up much more hard disk space and memory due to the additional information they include. Third, as Chris Cox says, it depends entirely on the image in question. Many times the file-size tradeoff in terms of memory and disk space used are not worth it. However, with hard drives and memory being so reasonable these days, argument #3 is quickly becoming less important.

Jake
DJ
dennis johnson
Aug 26, 2003
Well, think about it. How many devices, including the monitor you’re using right now, can utilize or display more than the number of colors that can be described by 8-bit color?

You ask, "…why does PS only let you work with 8 bit tiff images?" But Photoshop does give you the ability to edit and manipulate 16-bit images, including 16-bit TIF, so I don’t understand your complaint. Use the bit depth that you find more convenient or applicable to your purposes, and others will use what they need.

The world is wide.
BF
Brendon Fewel
Aug 26, 2003
Everthing you’ve said makes sence, except that I guess that my missing piece of knowlege is how many colors can be described by 8-bit color?

And I didn’t mean to make my last comment sound like a complaint, sorry about that. I phrased it in a chalenging way that I hoped would cause someone to fill in my knowlege gaps. (such as "What are you talking about? 8 bit is PLENTY of colors because…") Thanks for your help!
Y
YrbkMgr
Aug 26, 2003
In 99.99% of the cases, your output devices (monitor, printer, etc) cannot present a discernable difference between the number of colors produced by 16 bit data and 8 bit data. In other words, you won’t see it, in general.

So why have it? There are two reasons that I can think of.

First, 16 bit data contains more information. If you manipulate an image in 16 bit, then convert to 8 bit, the histogram remains relatively smooth (no drop outs). Who cares? Some do – some ascribe to the theory (by necessity or desire) that they want to work with as pure a data set as is possible. Will they see a difference? In most circumstances, no, but the data is about as pure as they can get and that is satisfying. Especially for archivists.

The other reason is that for SOME images one *may* be able to see the differences post manipulation. They will be very subtle, but a trained eye may be able to detect them. It’s rare that this happens, but it happens.

So in consideration of that, think of the processing power required to use a filter on 8 bit v. 16 bit data – it’s an order of magnitude higher. So the question is, for what? To date, it appears that Adobe is of the mind that the performance/benefit ratio isn’t worth it.

Whether that’s actually true or not, seems, from what I’ve read, to be an individual perspective rather than an industry wide one.

Peace,
Tony
P
Phosphor
Aug 27, 2003
Dave – I tend to regard that link as useless. They didn’t do very good testing. (And Dan has an obvious agenda, as usual).
DM
dave milbut
Aug 27, 2003
chris, i didn’t read it all the way through, but i figured it would at least help a beginner. thanks for the heads up though.

dave
P
Phosphor
Aug 27, 2003
Brendon,

you might want to look at something like Asiva <http://www.asiva.com/photo.html> if you want to edit only in 16-bit. I dont quite feel the need for it. I figure I can bring out the detail I want with levels and curves and then switch to 8-bit to do my editing. There arent many ways to output which will take advantage of a pure 16-bit image (the printing gamut is usally smaller than 8-bit).

Ive seen < http://www.inksupply.com/index.cfm?source=html/quadtone.html> for 4 and 6 color printers which expand the b&w printing gamut (sorta ala Ansel Adams effect). I’ve been thinking about trying it, I’m still not convinced even that would need 16-bit.

Hmmm, anyone else noticing the apostrophes disappearing in their posts?
CL
Chuck_Lambert
Nov 14, 2003
I’ve been trying to find out more about bit depth and this forum seems to be the closest to my line of inquiry. My question has to do with tonal range. Does 16-bit offer a wider range than 8-bit?

IOW, if maximum black in 8-bit mode is labeled as "1" and maximum white is "16,000,000," does 16-bit mode start at the same blackness of "1" and simply offer more gradations on its way to the same 8-bit maximum white? Or can 16-bit allow me to access a broader range of tones than 8-bit?

My goal is to be able to work with images that have the same 3200:1 tonal range as a film negative, which would occur at about 12-bpp.

If 16-bit relates to tonal range I should be able to take a picture of a black cat standing in the shadow of a coal bin with the sun in the background glaring into the lens. And then I could pull in detail from sunspots as well as subtle shadings of the cat’s fur. Somehow, I doubt 16-bit works this way.

Go ahead, enlighten me.

Chuck Lambert
L
LenHewitt
Nov 14, 2003
Chuck,

does 16-bit mode start at the same blackness of "1" and simply offer more
gradations on its way to the same 8-bit maximum white? <<

Yes. White is White and Black is Black and evermore will be so
CL
Chuck_Lambert
Nov 14, 2003
I was afraid of that. What a shame.

Thanks for the info.

Chuck Lambert
PC
Philo_Calhoun
Nov 14, 2003
It is easy to do one’s own testing. Basically if the image is well exposed and the manipulations are minimal, it would be exaggerating to say that one can see the difference between the 8 and 16 bit/channel images. With single manipulations, like just curves, I was suprised that I didn’t see that much benefit in 16 bit/channel images in bringing out shadows. But add some sharpening, hue adjustment, etc. and it is easy to find clear benefits to 16 bit/channel images. FWIW, I would not delete RAW files. I generally eventually save back to 8 bit/channel images after I am done, but keep the original RAW in case I wish to redo some part. Generally the RAW files will be considerably smaller than a 16 bit/channel tiff. I agree with Chris that Margolis slants his tests to make his point. However, there are many that overstate the advantages of 16 bit/channel tiffs.
KS
Ken_Storch_(aka_photon)
Nov 15, 2003
As has been touched on, 16bit has it’s greatest advantages in the ‘editing’ of critical files.

I have no doubts whatsoever that 16bit is worth the extra storage, and processing time, for being able to bring subtle, smooth tonal information. This is so even though my print output is done at the 8bit level.

Good luck.

Ken Storch

How to Improve Photoshop Performance

Learn how to optimize Photoshop for maximum speed, troubleshoot common issues, and keep your projects organized so that you can work faster than ever before!

Related Discussion Topics

Nice and short text about related topics in discussion sections