Closed Thread Icon

Topic awaiting preservation: CS 16-bit and web design Pages that link to <a href="https://ozoneasylum.com/backlink?for=9393" title="Pages that link to Topic awaiting preservation: CS 16-bit and web design" rel="nofollow" >Topic awaiting preservation: CS 16-bit and web design\

 
Author Thread
marf
Bipolar (III) Inmate

From: Canada
Insane since: Oct 2001

posted posted 01-02-2004 19:58

I was wondering that the new option with photoshop CS has 8 bit and 16 bit. When creating web graphics Should I use 16 bit or is it just going to make the image size bigger but not noticibly better quality? Is 16 bit meant for printable graphics? Is 8 bit fine for web graphics?

mahjqa
Maniac (V) Mad Scientist

From: The Demented Side of the Fence
Insane since: Aug 2000

posted posted 01-02-2004 20:06

8 bit is what just about everyone used so far, and it works great. 8-bit means here that 8 bits are used to describe the lightness value of one channel. In RGB mode (used mostly in webdesign) there are three channels for colors: Red, Green and Blue (hence the abbreviation) All these three channels have 8 bits to describe the lightness value: 2^8 for each channel (2^8 = 2*2*2*2*2*2*2*2 = 256, so in total 8-bit color mode is capable of 256*256*256 = 16,777,216 colors. The human eye can't distinguish between increments that small.

Long story short: 8-bit is all you need. Using 16-bit graphics will most likely double the filesize of your images (since the double amount of bytes is used to describe the color value of every pixel) and the human eye can't see the difference anyway.

Alevice
Paranoid (IV) Inmate

From: Mexico
Insane since: Dec 2002

posted posted 01-02-2004 22:00

I may be misinforme, but i though 8bit are teh 256 colors-only images. 16 bit may be actually what are you driving at.

__________________________________


Sexy Demoness cel

poi
Paranoid (IV) Inmate

From: France
Insane since: Jun 2002

posted posted 01-02-2004 22:21

Alevice: We are talking about color ( channel ) depth here, so 8bits stands for 16M of shades.
But when talking about images, people usually count all the channels in the bits depth, that's why a 256 colors image is said to be in 8bits, and 16M colors image is said be in 24bits.

Mathieu "POÏ" HENRI

[This message has been edited by poi (edited 01-02-2004).]

marf
Bipolar (III) Inmate

From: Canada
Insane since: Oct 2001

posted posted 01-04-2004 04:08

Ok thats what i thought. I'll keep using 8 bit for web graphics, but if im making an image to be printed, I will try 16 bit.

Bmud
Bipolar (III) Inmate

From: Raleigh, NC
Insane since: Mar 2001

posted posted 01-04-2004 05:52

That crazy Photoshop CS and it's 64-bit imaging capabilities... WHY!? WHY??? Though it would kick some butt if PNG could be 64 bit. Then it would be rather godly wouldn't it? I see potential for print graphics when increasing bits in channels, yes. I've done one 16-bit-per-channel image and it was cool up until I realized none of the photoshop filters would work anymore... Gotta be 8-bit my friend. The world (and photoshop) sorta demand it until CS or later versions get more wide-spread.

Shine and shine. :: [Cell=992] :: [See my team's 30 second animation]

jstuartj
Bipolar (III) Inmate

From: Mpls, MN
Insane since: Dec 2000

posted posted 01-04-2004 11:42

Well for web graphics, there is no real advantage. 16bit for print work has it's advantages but at the output stage you still loose the extra dynamic range as there are very few 16bit output devices. There are some film recorders used for motion pictures but those are all I am aware of.

The primary usefulness on the print side comes at the Scanning and editing stage. Where you get the advantage of a greater dynamic range, allowing for smoother tone transitions, improved quality when scaleing and resampleing, and more flexablitily while making color corrections. Mathmatically there is an advantage, visually it's debatable and then there's the problem of it being converted to 8 bit in the end anyways. Thus loosing any tonal range benifits on output.

I scan at 600ppi at 100% of final size in 16bit, RGB or LAB. Make major color corrections, edit for dust and make some minor clone tool correction and then archive the image as an uncompress tiff and JPG. I then resample the image to 300ppi convert to 8bit RGB to complete any layer manipulations or specific corrections and archive as a PSD, TIFF, and JPG.

That way the most data possiable is aviable if necessary or at some future time a 16bit printer is an option.

You do have to deal with huge images, most of my scans run around 200 megs, but I find I can edit the just fine with just 512meg, with a fast 10 gig SCSI scatch disk. As for storage All project go on Kodak Gold Archival Quality CD's

J. Stuart J.

The Fragmaster
Nervous Wreck (II) Inmate

From:
Insane since: Apr 2001

posted posted 01-08-2004 00:37

For those who don't understand why 16bit/channel color is important, try this:

1. Start a new document, make it fairly large
2. Fill the background black
3. select a large soft airbrush, set opacity to 10% and color to pure white
4. Make repeated strokes around with the airbrush, getting lots of different shades of grey
5. Open up levels, drag the black slider towards the right until you have a cluster of white and dark greys
6. Feast your eyes on the banding and dithering that one could only get in a 256 color image.

Now I understand that this can be battled by introducing slight bits of other colors, eg. 231-232-231 and so on, but even that has to have a limit as far as color fidelity is concerned. That's also probably why most movie studios use 16bit/channel images for their special effects, that dithering tends to look a lot worse when it's being projected onto a 30 foot screen

« BackwardsOnwards »

Show Forum Drop Down Menu