Re: seeking for help on MNG creation

> > The main problem is the remainder of the file. Here's a snapshot of a
> > MNGeye dump:
> 
> Ah, that seems to be the kind of information I was expecting to get from
> mngtree. This should get me further. Thanks for your reply.

Check http://www.libmng.com/sven/

I've adapted your file into a MNGeye script that appears to work. What I've 
done is to set do_not_show for the second image and all of it's clones to 1 
(eg. invisible); otherwise they get displayed on the fly. I've also moved the 
cloning bit before the SAVE/TERM/SEEK triplet, so they don't get repeated 
over and over once you set a repetition on the TERM.

Also the image_id on the BACK chunk was listed as 256 (endian byte-
order?!?), so mended that to 1.

I changed the SHOW chunks to list the actual images (clones) to be displayed 
for that specific frame, rather than leaving it empty. This may be a part of the 
problem, but not sure.

I don't yet know why your example doesn't work. I'd have to investigate 
a lot deeper, but with the given script you'll have a working base.

> I am not overly happy with the use of CLON but I am not sure if
> MOVE/SHOW can do what I want to achieve. I need to show different
> parts (that's why I use CLIP) of an image several times. Basically in
> this example, there's one overlay image with seven regions. For each
> window of the building (144 windows), I need to choose one region of
> this overlay image and place it on the screen. All 144 copies need to be
> visible at the same time. Can this be done using MOVE/SHOW?
 
If you put a proper CLIP in between them, sure. And it could save some space 
if there are repetitive sequences (more than 2!) with increasing offsets in x, y 
or both directions. Eg. enveloping the 2nd and onwards inside a LOOP/ENDL 
and making the MOVE&CLIP parms relative. But of course that'll require some 
inventive programming, although not impossible imho.



--
Send the message body "help" to [email protected]