Reviews, updates and in depth guides to your favourite mobile games - AppGamer.com
|
|
How to work with 16bit values? |
Page: 1 | Reply |
Oct 12th 2006 | #174939 Report |
Member since: Oct 11th 2006 Posts: 2 |
Hello, I like to work in a real 16bit (per channel) RGB mode. When I create a new RGB image with 16bit there are only 8bit values (0-255 and not 0-1023) in the color picker and in my info panel. Is PS CS2 just working internally with 16bit and doesn't provide manual access to all values?? thanks, mH |
Reply with Quote Reply |
Oct 12th 2006 | #174952 Report |
Member since: Mar 27th 2001 Posts: 2237 |
First you really have to have a full understanding of bit depth and bit color. Do a little research. When you see 8 vs. 16 bit, it's referring to the depth of each channel (e.g., R, G, and B). So an 8-bit file is "24-bit color" (8 bit x 3 color channels) and a 16-bit file is "48-bit color". (16 bit x 3 color channels) 24-bit color gives 256 tones in each color which is more than enough for human vision to see pretty continuous tones. The problem is when you start doing image manipulation you are doing math with those pixel values. Every operation you perform loses information from the bottom few bits due to rounding, especially if you make some serious changes to contrast and color balance like you often need to in processing a film scan. With 48-bit color with 16-bit precision you can loose those last few bits to image manipulation and still have more than 8-bits of real data left for output. Performing the same operations in 8-bit quite often would yield to gaps in the spread of values on the histogram which corresponds to posterization (stepping) on the image. Edited cause I'm a bonehead |
Reply with Quote Reply |
Oct 12th 2006 | #174955 Report |
Member since: Oct 11th 2006 Posts: 2 |
thanks tornupinside for replying, i know the terms of 16 and 48bits and we video guys are using the 'per channel' term. some details of my problem: i will have to work (with photoshop CS2) on a image that will be converted from a logarthmic film scan (cin or DPX) to an image in linear colorspace with 10 or 16 bit (for EACH channel). I need to see the real values from 0 to 1023 for the pixels like in my compositing software. photoshop only shows 0 to 255 even if it is a 16bit (per channel) image. How can i see the real values in photoshop? thanks, mH |
Reply with Quote Reply |
Oct 13th 2006 | #174965 Report |
Member since: Apr 27th 2006 Posts: 4 |
[QUOTE=moonhopper]thanks tornupinside for replying, i know the terms of 16 and 48bits and we video guys are using the 'per channel' term. some details of my problem: i will have to work (with photoshop CS2) on a image that will be converted from a logarthmic film scan (cin or DPX) to an image in linear colorspace with 10 or 16 bit (for EACH channel). I need to see the real values from 0 to 1023 for the pixels like in my compositing software. photoshop only shows 0 to 255 even if it is a 16bit (per channel) image. How can i see the real values in photoshop? thanks, mH[/QUOTE] I think, you can not see values in 16bit format in Photoshop. It always conwert all valuse to 0-255 range. Franly speaking, the Photoshop use 16 bit for saving and calculation, but significant are 15 only. The 16th bit uses as "warning for out of gamut". |
Reply with Quote Reply |
Page: 1 | Back to top |
Please login or register above to post in this forum |
© Web Media Network Limited. All rights reserved. No part of this website may be reproduced without written permission. Photoshop is a registered trademark of Adobe Inc.. TeamPhotoshop.com is not associated in any way with Adobe, nor is an offical Photoshop website. |