From: | Ben Hutchings |
Date: | 04 Dec 99 at 00:26:25 |
Subject: | Re: Little maths problem |
From: Ben Hutchings <womble@zzumbouk.demon.co.uk>
On Thu, Dec 02, 1999 at 09:50:41AM +1100, Jonathan wrote:
> From: "Jonathan" <jadamcze@utas.edu.au>
>
>
> ----- Original Message -----
> From: Ben Hutchings <womble@zzumbouk.demon.co.uk>
> To: <amiga-c@onelist.com>
> Sent: Thursday, December 02, 1999 9:49 AM
> Subject: Re: [amiga-c] Little maths problem
>
>
> | From: Ben Hutchings <womble@zzumbouk.demon.co.uk>
> |
> | On Thu, Dec 02, 1999 at 09:02:08AM +1100, Jonathan wrote:
> | > From: "Jonathan" <jadamcze@utas.edu.au>
> | >
> | >
> | >
> | > | From: "Alastair M. Robinson" <blackfive@fakenhamweb.co.uk>
> | > | You might also want to experiment with the gamma function (as found in
> | > | your favourite image processing program):
> | > |
> | > | f=exp(gamma*(log(f)-log(500000.0)));
> | > |
> | > | If gamma is set to 1.0, you'll get a linear transfer. Set it lower
> | > | than 1.0 to bias in the direction you want, and higher than 1.0 to
> | > | bias in the other direction.
> | > |
> | >
> | > How does gamma work? (never seen it before)
> |
> | gamma is the usual letter representing this `non-linearity' value in a
> | response curve. gamma values are normally quoted for optical input or
> | output devices, or for some processing/filtering of optical signals.
>
> Yep, know all about it ie Quake ;)
>
> So, what is the actual function?
Something like:
gamma
output = c + f . input
where c is a constant bias and f is a scale factor. There are
probably conventional letters for these that I've forgotten.
Ben.