openexr-devel
[Top][All Lists]
Advanced

[Date Prev][Date Next][Thread Prev][Thread Next][Date Index][Thread Index]

Re: [Openexr-devel] OpenEXR 1.5.0, OpenEXR-Images 1.5.0 released


From: Florian Kainz
Subject: Re: [Openexr-devel] OpenEXR 1.5.0, OpenEXR-Images 1.5.0 released
Date: Wed, 20 Dec 2006 16:35:49 -0800
User-agent: Mozilla Thunderbird 1.0 (X11/20041207)

I don't recall describing encoding as "perceptually logarithmic",
but apparently what I wrote was ambiguous.  I changed the comment
that describes the pLinear flag in ImfChannelList.h to make it clear
(I hope) that pLinear is a hint about how humans perceive the quantity
represented by a given channel, and that the compressor may use this
flag to optimize encoding/quantization of data in this channel.

Florian



Chris Cox wrote:
Florian;

if the data is digitized or encoded in any way - then any description of it 
must include the encodiing.
For example: log compresssion of sound/pressure values can be done in analog 
electronics, but you still have to describe it as log encoded or you'll get 
lousy results when you try to playback or process that signal later.

You really should use light as the baseline (truth) and reference the encoding 
to that.
Saying that something is "perceptually linear" tends to confuse things.

Stating that human perception of light is approximately logarithmic is fine.
But describing an encoding (as you did) as "perceptually logarithmic" is quite 
misleading (as I described previously).

Chris




reply via email to

[Prev in Thread] Current Thread [Next in Thread]