Photon video Cel animator - Microillusions.

[DDNI]

Starting a new thread as the titles of the other ones where I am looking for the Fairy animation are not specific.

Does anyone out there have a copy of Microillusions Photon video Cel animator?

TOSEC lists two disks, but the actual boxed version had three disks. I am hopeful that the missing third disk contains the animation.

[Phantom]

Hi mate.

Possibly have it, need to check. Does your version is 1.8 though?

[DDNI]

Hi any version that has disk 3 is welcomed!
I don't have a copy, just the 2 disks from TOSEC

Hope your check is positive!

[Phantom]

Sure, no problem.

Just let me 2-3 days to check thoroughly.

[Phantom]

I've uploaded in the Zone. Unfortunately I can't test it as I don't have an Amiga here. Please inform me if it works though.

[DDNI]

Hi I just saw your 4th July post!

Many thanks for uploading it! The anim that I seek IS there!
It is called Angel.anim
I am having an issue getting the anim to load though. It stops when loading anim frame 16 complaining of a mangled .iff file.
Could you possibly try redumping disk 3?

[cosmicfrog]

Wow and Big Ribbits too

So Glad you`ve finaly found it

Bloody disks gerrrrrr

[Supamax]

Originally Posted by DDNI

Could you possibly try redumping disk 3?

I suppose Phantom found the disks in some BBS CD (since they are in .dms format), so I don't think he can "redump" them"...

[prowler]

Originally Posted by Supamax

I suppose Phantom found the disks in some BBS CD (since they are in .dms format), so I don't think he can "redump" them"...

I've just downloaded the archive that Phantom uploaded to see if it's possible to fix disk 3.

Edit: I've uploaded an archive containing three valid ADF disk images derived from the DMS images in Phantom's archive.

I've not tried them yet to see if the mangled Angel.anim problem is solved. Will have a look tomorrow.

[Supamax]

Originally Posted by prowler

I've just downloaded the archive that Phantom uploaded to see if it's possible to fix disk 3.

Edit: I've uploaded an archive containing three valid ADF disk images derived from the DMS images in Phantom's archive.

I've not tried them yet to see if the mangled Angel.anim problem is solved. Will have a look tomorrow.

Hi Peter ,
I don't understand: if the .adf files are derived from the .dms files, how can the mangled file problem be solved?

[prowler]

Originally Posted by Supamax

Hi Peter ,
I don't understand: if the .adf files are derived from the .dms files, how can the mangled file problem be solved?

Occasionally, users have reported that when DMS images are use with WinUAE, disk errors of this type result.

I don't know what causes this problem, but sometimes it can be cured by extracting the DMS files to ADF before using them with the emulator.

If it turns out that the Angel.anim file really is corrupt, then I will try to fix it.

[Supamax]

Originally Posted by prowler

Occasionally, users have reported that when DMS images are use with WinUAE, disk errors of this type result.

I didn't know that... ok, that explains all.

If it turns out that the Angel.anim file really is corrupt, then I will try to fix it.

I'll need your help with some disks of mine, with errors... if you'll have time .

[prowler]

Originally Posted by Supamax

I didn't know that... ok, that explains all.

This was the last time it happened:
http://eab.abime.net/showthread.php?t=59102
(See post #3 et seq.)

Originally Posted by Supamax

I'll need your help with some disks of mine, with errors... if you'll have time .

I'll be pleased to help in any way I can, mate. You know where I am.

[Supamax]

Originally Posted by prowler

This was the last time it happened:
http://eab.abime.net/showthread.php?t=59102
(See post #3 et seq.)

VERY good to know...
I have a lot of .dms files from BBS archives etc.

I'll be pleased to help in any way I can, mate. You know where I am.

ok

[Phantom]

Hope that you did fix that little problem.

[cosmicfrog]

Yay Prowlers on the case

[prowler]

Originally Posted by Phantom

Hope that you did fix that little problem.

Not yet, I'm afraid.

I have copied the animations from Art Disks 1 and 2 (Photon Video CEL Animator Disks 2 and 3 ADF images) to my Windows XP machine.

Art Disk 1 contains the following animations:

Horse&Rider.anim (189,514 bytes)
320x200x24bpp, 12 frames

Max.anim (355,086 bytes)
320x200x8bpp, 41 frames

Waterfall.anim (246,630 bytes)
352x480x24bpp, 8 frames

and Art Disk 2 contains the "mangled" animation:

Angel.anim (808,666 bytes)
320x200x24bpp, 26 frames

The animations from Art Disk 1 play flawlessly in the iffanimplayer for Windows, but there is a problem with the Angel animation, as reported by DDNI.

There is a fair degree of similarity between these files. In particular, they all use the same compression method, whereby the first frame (frame 0) is included in full .iff format (this is what you would see if you opened the animation in IrfanView), and subsequent frames are derived from this and 'Delta' (difference) data. The first frame has a 'BODY' header and Subsequent frames' data have a 'DLTA' header.

Thus, Waterfall.anim, for example, has a single 'BODY' block (frame 0) followed by seven 'DLTA' blocks (frames 1-7).

As expected, Angel.anim also has a 'BODY' block (frame 0) and there are 25 'DLTA' blocks (frames 1-25). So far, so good.

The Angel animation displays 14 frames (0-13) in the iffanimplayer, and then crashes. The animation is opened successfully (all 26 frames loaded), but little more than half will display. (The CEL Animator loaded all 26 frames successfully when I tried it in WinUAE and would play the animation up to about the same point, but would then dissolve into a corrupted mess resembling an outdated DirectX driver problem, from which I could escape only with a right mouse button click.)

If the animator is having trouble decoding the 15th frame, you would expect to see a problem with the data between the 14th 'DLTA' in the animation code at offset 0x6D676 and the 15th at offset 0x75A60, but there is nothing obvious in this region.

However, there is a large block of code zeroed out between offsets 0x7E4A0 and 0x80320 (7,840 bytes). The 16th 'DLTA' is at offset 0x808AA, so this is included in the 16th frame.

Remember that DDNI reported that the CEL Animator stopped when loading frame 16, complaining of a mangled .iff file (though I could not reproduce this when I tried it), so this block of zeroes does seem to be implicated.

I have not looked at the structure of Art Disk 2 yet, but I am expecting to see one or more of the Angel.anim file list blocks pointing to empty data blocks caused by disk read errors when the disk was dumped.

If that is the cause, then the only solution will be to obtain a good dump of that disk or, failing that, a good copy of the Angel.anim file from another source with which it would be possible to repair the disk image.

For those interested, I am preparing an archive to upload to the Zone with the animations from the CEL Animator Art Disks 1 and 2 (including the corrupted one), the iffanimplayer for Windows, documentation and batch files to play them on a minimum 1024x768 resolution Windows desktop (though this can be changed).

Edit: The files are now in the Zone!

[prowler]

I have now mapped all the blocks on this disk image.

The Disk.info, .info, Angel.ord and Angel.exp files on Art Disk 2 are intact, but the Angel.anim file, which has 808,666 bytes in a total of 1658 data blocks, has blocks 1011-1076 inclusive (66 blocks, or 32,208 bytes) missing or overwritten. I expect that these will turn out to be the last 66 blocks of the disk image when I have finished my analysis of the file header and list extension blocks, which are intact, fortunately. This region contains a mixture of zeroed out and loose data blocks with missing or zeroed out headers.

The corruption corresponds to the region of the Angel animation file between offsets 0x78550 and 0x8031F inclusive. This falls between the 15th and 16th 'DLTA' headers at offsets 0x75A60 and 0x808AA, respectively, confirming that only the 16th frame is affected.

Sadly, it is not possible to display the undamaged part of the animation following the corrupted section because each frame of the animation is used as the starting point for decoding the next.

The last 66 blocks in this disk image might as well have come from another disk for all the purpose they serve. I know it's a long shot, but I'm going to check the other two disk images in the set for these missing blocks before I give up on it.

[prowler]

My analysis of the Angel animation file header and list extension blocks confirms that the last 66 blocks of the disk image originally contained the missing data.

I have searched the other two disk images in the set for the opening byte sequences from the headers of the missing data blocks:
00000008 00000372 000003F3 000001E8
00000008 00000372 000003F4 000001E8
00000008 00000372 000003F5 000001E8
.
.
00000008 00000372 00000434 000001E8

but no instances of the sequence '00000008 00000372 000003' or '00000008 00000372 000004' could be found.

I will retain the notes I have compiled relating to this disk image in case either the complete animation or the missing data with which to repair it ever become available.

[prowler]

So that I have something to show for all the work I have put into this in the last few days, I have truncated the Angel animation at the end of the last valid frame (the 15th), renamed it to Fairy.anim, and uploaded it to the Zone complete with the iffanimplayer for Windows and a batch file to play it in a continuous loop at a reasonable speed. And, of course, this version of file will play in the Amiga's CEL Animator program too.

It did occur to me that there was just a chance that the 15th delta (i.e., that which describes the difference between the 15th and 16th frames) was identical to one of the others in the animation, but in fact there is a greater difference between those two frames than between any other consecutive pair in the animation, with the possible exception of the last and first frames!

So, truncating the animation at the end of the 15th frame is the best that can be done with it until the missing data is found, unfortunately.

Still, it's better to have tried and failed than never to have tried at all.