Legacy ImageMagick Discussions Archive

What is the status of this bug?

Ok. Are there any plans of implementing what it is said on the manual or not?

I can live with whatever is decided. Thank you for your time and attention.

Your gif is a two-frame animation . Each frame has only one color (black or white). So to reset the colormap

convert square1.gif -coalesce -colors 1 square2.gif


Yes, I know. I was constructed to prove that convert -colors 256 is not following what is specified in the FAQ (cf. http://www ...

BTW, I know the color palettes have at maximum 256 colors and that's the point. convert -colors 256 can be used to simply and always optimize gifs (i.e. remove unused or repeated colors from color palettes).

Hi fmw42,

I'm sorry that it was not clear that the second test image was this one:

http://desmond.imageshack.us/Himg406/scaled.php?server=406&filename=square1.gif&res=landing

Unfortunately, the correct ordering doesn't seem to change anything, i.e. duplicated or unused colors on color palettes ...

Color reduction / color palette optimization doesn't seem to be working as advertised on IM 6.7.1 Q16 / Win XP.

Lets take the following image with 603 unique colors:

http://www.imagemagick.org/Usage/anim_opt/speed.gif

Using the following command:

convert -colors 256 speed.gif speed_opt.gif ...

Optimizeframe doesn't seem to be working as advertised either, as the following command:

convert canvas_prev.gif -coalesce -layers OptimizeFrame optframe.gif

for the example provided at:

http://www.imagemagick.org/Usage/anim_opt/

produces a smaller image, but not as small as optframe.gif shown ...

The frame doubling optimization for animated GIFs doesn't seem to be working on IM 6.7.1 Q16 / Win XP. For instance, when I try to use:

convert moving_hole.gif -layers OptimizePlus moving_hole_oplus.gif

on the image moving_hole.gif, the image is here:

http://www.imagemagick.org/Usage/anim_opt ...

Hi glennrp,

Thanks for the explanation.

I have one further question:

Let us suppose that the quantization tables are the same. After decompression, do IM uses the same algorithm as jpegtran to (re)compress image's data or IM algorithm is more powerful and that also contributes to the 2% difference?

Hi anthony,

the study was done stripping all meta-data because it says that the following commands were run:

> jpegtran -copy none -optimize source.jpg result.jpg
> jpegtran -copy none -progressive source.jpg result.jpg

Either IM strips meta-data even better or compress data even better. Either ...

Hi magick,

I saw this interesting study done by Stoyan Stefanov showing that jpeg images compressed by imagemagick are 2% smaller on average than performing the same operation with jpegtran, cf.:

http://www.yuiblog.com/blog/2008/12/05/imageopt-4/

Is this due to imagemagick's slightly lossy ...