Possibly numpty question about "bits"

mickledore

mickledore

Suspended / Banned
Messages
9,372
Edit My Images
Yes
Just been reading another thread that is going on about 8 bit, 16 bit etc. I've never had any dealings with this, and just edit images as they come out of the camera. I use LR5 and CS6. I've never noticed any problems or reluctance of thye software to process as I desire.

Can someone explain in simple terms what all this bit things mean and how they affect processing?
 
It's a computing term, a bit is either a 0 or 1.
If there's 8 bits of information for each colour channel Red, Green, Blue, then that's 256 possible shades of each channel, roughly 16 million different colours (combination of those). 000 would be pure white and 255 255 255 would be pure black.
 
So if there's more information on 16 bit images why would you want to edit in 8 bit?
 
mickledore said:
So if there's more information on 16 bit images why would you want to edit in 8 bit?
Click to expand...

Less processing power - quicker.

Many of the colours on 16 bit are only theoretical only.
Some filters ,even in full version cs only work on 8 bit photos
 
OK, get that. So why have 16 bit if it's not as good, and how would I know what size my files are? I've had a look but can't see any indication.
 
mickledore said:
OK, get that. So why have 16 bit if it's not as good, and how would I know what size my files are? I've had a look but can't see any indication.
Click to expand...

Its like many things photographic. "Better than..........." just isn't true. More suited to your needs it may be. Some people need 16 bit. Some only need 8.

In CS you can tell which you're in by selecting IMAGE > MODE
 
Last edited:
Lightroom etc doesn't 'care' how big your files are - they're just files. 16 bit is 'better' in some ways - there's advantages and disadvantages, and it's a case of finding the most appropriate set of parameters.
 
Thank you. I import into LR, then move to CS6 if need be. A quick check shows that my images are all 16 bit. I presume this is done on initial import.

This has solved one great big problem that I have. When processing my 16 bit (now that I know what they are) files I cannot open Filter>Filter Gallery.
This has annoyed me because I could do it on my previous version of CS. Now I know why. It was worth the question!

It seems that LR is set to import as 16 bit. Is it worth changing that to 8 bit, or just amend individually as the need arises?
 
Assuming a modern computer, I'd always process in 16 bits and then only move to 8 bits when I export to JPEG.

Why?

What 16 bits gives you is the ability to deal with more dynamic range (detail in the numeric values). If you are processing at 8 bits, you lose a lot of the low level detail - try taking a picture and storing it as both raw and JPEG and then processing both images by trying to get back some detail in the shadow areas. You soon run out of wriggle room in the blacks simply because you have lost the ability to represent the subtle changes at low lighting levels because you have dropped to 8 bits.
 
Editing in 8bit can result in a lot of data loss which can result in color banding and other issues/degradation. Attached is an image of the histograms of an 8bit image. The top histogram shows the data upon opening the image. And the lower histogram shows the data loss after making a basic levels adjustment in PS.
View attachment 28796

I avoid 8bit if at all possible.
 
Another thing, a bit is 2 to the bit power. For example 1 bit color is 2 bits of data, and 2 bit color is 4 bits of data. There is a HUGE difference between 2 to the 8th pwr (8bit) -vs- 2 to the 16th pwr (16bit).
There's even a very large difference between recording files in 12bit -vs- 14bit in the camera... but you *might* not see it.
 
mickledore said:
Thanks again guys. I've been looking at my imports into LR. That wants me to import in 16 bit ProPhoto RGB. That's what I've been using so far and will stick with it. ProPhoto seems to be favoured as mentioned here:-

http://petapixel.com/2014/12/16/photographers-introduction-color-color-space-monitor-calibration/

The only drawback is as mentioned above that some filters won't work in CS6 with 16 bit, but I don't use these a lot so can easily work round that.
Click to expand...
The trick to edit 16bit w/ 8bit filters is to make the image/layer a smart object.
 
Wow! At last I can find filter gallery thanks to your help
 
sk66 said:
Editing in 8bit can result in a lot of data loss which can result in color banding and other issues/degradation. Attached is an image of the histograms of an 8bit image. The top histogram shows the data upon opening the image. And the lower histogram shows the data loss after making a basic levels adjustment in PS.
View attachment 28796

I avoid 8bit if at all possible.
Click to expand...

The easiest place to see the advantage of 16 bits over 8 is with a graduated background. It doesn't take much manipulation of an 8 bit image via levels or curves before it breaks down into distinct bands. The same thing happens on a small scale all over an image.

Processing in 16 bits largely makes this problem go away.
 
mickledore said:
So if there's more information on 16 bit images why would you want to edit in 8 bit?
Click to expand...

You don't generally. However.. depends how you shoot. JPEGs are 8bit. Only RAW, TIFF, PSD etc support 16bit processing. If you do a lot of porcesing you really should shoot in raw. If you need photoshop, export your raw files as 16bit TIFF files.

Yet another reason why JPEG sucks I'm afraid.
 
You must log in or register to reply here.
Back
Top