Deutsch   English   Français   Italiano  
<vlhdf5$1mvfl$1@dont-email.me>

View for Bookmarking (what is this?)
Look up another Usenet article

Path: ...!eternal-september.org!feeder3.eternal-september.org!news.eternal-september.org!eternal-september.org!.POSTED!not-for-mail
From: -hh <recscuba_google@huntzinger.com>
Newsgroups: comp.os.linux.advocacy
Subject: Re: Why GIMP Is Better Than Photoshop
Date: Mon, 6 Jan 2025 15:12:21 -0500
Organization: A noiseless patient Spider
Lines: 38
Message-ID: <vlhdf5$1mvfl$1@dont-email.me>
References: <pan$12f0c$b8055f7b$e138cd3a$8c4aa277@linux.rocks>
 <vle9tl$13eiv$3@dont-email.me> <vlerd7$13ci0$4@dont-email.me>
 <vlfbcv$19rah$1@dont-email.me>
MIME-Version: 1.0
Content-Type: text/plain; charset=UTF-8; format=flowed
Content-Transfer-Encoding: 8bit
Injection-Date: Mon, 06 Jan 2025 21:12:22 +0100 (CET)
Injection-Info: dont-email.me; posting-host="f91ff1d41e6b90a2ee46fb99cb6cc1a2";
	logging-data="1801717"; mail-complaints-to="abuse@eternal-september.org";	posting-account="U2FsdGVkX19M22I5VGUN+n3HSvfrsFX4zIl4mE9u2hI="
User-Agent: Mozilla Thunderbird
Cancel-Lock: sha1:ksMM0FtcwHFOXB9yjppvfstdU0Y=
In-Reply-To: <vlfbcv$19rah$1@dont-email.me>
Content-Language: en-US
Bytes: 2786

On 1/5/25 8:24 PM, Lawrence D'Oliveiro wrote:
> On Sun, 5 Jan 2025 15:51:51 -0500, -hh wrote:
> 
>> ... needs to explain how & why it is significant for
>> there to be floating point at all, since the input sensor is integer
>> based ...
> 
> OpenEXR files are commonly used in CG these days, and they have floating-
> point numbers for each pixel component. They also allow for more than 3-4
> pixel components. The values still have their usual meaning, with 0 being
> full black and 1.0 being full white, but the values are allowed to go
> outside this range to avoid clipping of dynamic range.

Dynamic range is why to support more bits, but that's oblique to adding 
more intermediate values by using floating point instead of integer. 
For example, using 16 bits instead of 12 to address the perceivable 
spectrum range only is enough on its own to have more than a full order 
of magnitude greater than perceivable gradations.

> GIMP’s GEGL pixel engine deals natively with such things. Photoshop needs
> to use import/export filters which inevitably lose quality. 

But isn't integral vs modular just a software architectural design 
choice?  Particularly since there's nothing inherent of one over the 
other for questions of if it is designed to be lossless or not/etc.


> This is probably why Adobe products are not used much in VFX work.

Adobe's product here is probably After Effects, and it supports formats 
such as ProRes, where the originals are integer 10 or 12 bits/channel.

For others, see:

<https://www.actionvfx.com/blog/10-best-vfx-software-how-to-choose-what-s-best-for-you>


-hh