[Tig] Question about 10bit log to 8bit lin

liao.zd at gmail.com liao.zd at gmail.com
Tue Aug 6 11:34:25 BST 2013


Dan:

Is the 0.6 same thing as 1/1.7? because the 1/0.6 = 1.6667 and 1/1.7 = 0.5882
Thanks.

Liao Zhuodi



liao.zd at gmail.com

From: Dan Tatut
Date: 2013-08-06 18:17
To: tig
Subject: Re: [Tig] Question about 10bit log to 8bit lin
Sohonet www.sohonet.co.uk sponsors the TIG.
Support from Digital Vision www.digitalvision.tv
Support from MediaCloud www.MediaCloud.cc
=====


Hi Liao,

the 0.6 is the film gamma (average value)

Dan

On 8/6/2013 12:10 PM, liao.zd at gmail.com wrote:
> Sohonet www.sohonet.co.uk sponsors the TIG.
> Support from Digital Vision www.digitalvision.tv
> Support from MediaCloud www.MediaCloud.cc
> =====
>
>
> I check the website http://www.cineon.com/conv_10to8bit.php
> There are some formulas about the 10bit log | 8bit lin conversions, like:
>
> OUT = 10 ^ ((IN-Refwhite) * 0.002/0.6) ^ (Dispgamma/1.7) * Gain - Offset
>
> I don't understand what is the factor "0.002/0.6" means? I guess, maybe the factor "0.002" is related with this: "With 10 bits per color over a 2.048 density range, the resulting quantization step size is 0.002 D per code value. " But I am not sure what is the "/0.6" comes from.
>
> And for the Gain and Offset, they make no sense to me even more:
> Gain = 225 / (1-10 ^ (Refblack-Refwhite) * 0.002/0.6) ^ (Dispgamma/1.7))
> Offset = Gain - 225
> Why they both contain a factor "225"? And why the Gain fomulas is like this?
>
>
> Liao Zhuodi
> _______________________________________________
> http://reels.colorist.org
> http://tig.colorist.org/wiki3


-- 
Dan Tatut
VP Business Strategy & Development
MARQUISE TECHNOLOGIES SA
Rue des Avouillons 4
1196 Gland - Switzerland
Tel +41 22 364 54 71
www.marquise-tech.com

_______________________________________________
http://reels.colorist.org
http://tig.colorist.org/wiki3
.


More information about the Tig mailing list