RSS/Atom feed Twitter
Site is read-only, email is disabled

incremental scaling

This discussion is connected to the gimp-user-list.gnome.org mailing list which is provided by the GIMP developers and not related to gimpusers.com.

This is a read-only list on gimpusers.com so this discussion thread is read-only, too.

3 of 4 messages available
Toggle history

Please log in to manage your subscriptions.

incremental scaling Alexander Rabtchevich 18 Feb 16:35
4217D8F0.3020505@ozemail.co... 07 Oct 20:17
  incremental scaling Alexander Rabtchevich 20 Feb 12:20
   incremental scaling David Hodson 20 Feb 13:23
Alexander Rabtchevich
2005-02-18 16:35:02 UTC (about 19 years ago)

incremental scaling

It is known incremental scaling produces better results than one-step scaling. Is it possible to add "steps" listbox (with the default value 1) to the future GIMP scaling dialog to provide automatic incremetal scaling?

Alexander Rabtchevich
2005-02-20 12:20:30 UTC (about 19 years ago)

incremental scaling

I've downsampled from 1562x1562 (crop) to 64x64 (avatar). And the incremental scaling produced definetely better results. GIMP 2.2.3, Windows, cubic scaling.

David Hodson wrote:

Alexander Rabtchevich wrote:

It is known incremental scaling produces better results than one-step scaling.

Only if your scaling algoroithm is not very good. Better would be to fix any problems in the scaling algorithm, and scale in a single pass.

David Hodson
2005-02-20 13:23:07 UTC (about 19 years ago)

incremental scaling

> David Hodson wrote:
>
>> Alexander Rabtchevich wrote:
>>
>>> It is known incremental scaling produces better results than one-step >>> scaling.
>>
>> Only if your scaling algoroithm is not very good. Better would be >> to fix any problems in the scaling algorithm, and scale in a >> single pass.

(Sorry, I accidentally replied direct to Alexander first time...)

Alexander Rabtchevich wrote:

I've downsampled from 1562x1562 (crop) to 64x64 (avatar). And the incremental scaling produced definetely better results. GIMP 2.2.3, Windows, cubic scaling.

The problem with scaling down by a large factor is that you need to average out a large number of pixels to get an accurate result. As far as I know, the Gimp rescale function uses the same algorithm for scaling either up or down, so if you're using cubic interpolation it will only sample 16 pixels.

The simplest solution is to blur the image before scaling down. I'd try with a radius about the same as the scaling factor, but that's just a guess.