Dragon's Lair introdemo for MSX2 - update

OK, I do get MSX-AUDIO samples now. But the quality is a lot less compared to turboR PCM or Covox (SIMPL)!

Ah well, it's still amazingly cool

mars, by the way: note that the default harddisk image name is hd.dsk and you can also just overwrite it with your file. Normally you don't need to adjust the size, because openMSX will increase it automatically to the default 100MB, but because of the aforementioned bug in the Windows version of openMSX, the disk image will be corrupted when openMSX tries to increase the size. So you must indeed set it to the actual size (or smaller).

At least it's set up in My Documents which means it can be much easier accessed than other user-profile-supported applications tend to have... putting the whole lot in Application Data or Local Settings in your profile which is usually a hidden directory...

but I agree it's pretty annoying to find the HD image in a different place than the device itself is located... I'd rather have the whole lot outside of My Documents and more like user defineable as well... for quite some time.

Shagged some girls again, last night, snout? Discussions about 'My Documents' are important, you asswhimp

spl : try it in openMSX !

Yep, it was timed almost right for 60Hz, but you could hear small silent parts in sound, so triple buffering was needed to get rid of these silent parts and to enable 50Hz.

I just tested this first time with real MSX-Audio, and I must say, that it sucks a big time! Even the experimental FM-PAC is much better! Is this really best, that MSX-Audio can do??? Good thing is, that I think, that most of the MSX-Audio cartridges at least in Europe are Phillips models, where you can get a good sound.

I only can see garble

http://img212.imageshack.us/img212/5816/tv20060508005526008wj.png

I use a Turbo-R ST and sunrise IDE

The HDD is just formated, the demo are the only files on disk

I tested with diferent versions of DOS 2

To use multiple harddisks with openmsx just duplicate the ide directory in the extentions.... and put a harddisk image in the my documents dir.

good job, is it in screen 8?

That's not a bad compliment for a screen 2 video.

yea, as long as things keep on moving you can mask these errors..

In this case, each frame should be 256*3*2+16*3 bytes. Rigth ?

Only 16*2 for the palette.

So the datarate is more or less equivalet to a scr5 window of

184x138 (aspect ratio 4/3)

that corresponds to

184/2x138=12.696+32 = 12.728

I see many reasons to develop a pc tool to convert videos to scr2 fomat...

for sure it can,
the conversion now should take in to account a 3rd dimension: TIME

here coding across time you could have a better masking
of artefacts due to colorclash...

remember, on msx1 using TIME you get 100 colors,

let's use the money wasted in the CPU of our PC's
there, the computational load isn't a issue ...

Because I haven't now a IDE, I can't run it on my MSX but THANKS also for the use of ADPCM samples of MUSIC MODULE!
Actually it's not ADPCM. MSX-AUDIO also has a PCM D/A converter. The PCM data register is $15 and can be read while the MSX-AUDIO is recording data or written to for unbuffered sample play. It's 8-bit signed PCM data.

@NYYRIKKI

encoding more frames jointly means also to be able to optimize palette transitions among the frames...
Note also that the colorclash now is masked by the animation.
A coding that takes into account more sucecssive frames could optimize the masking effect improving greatly the final visual result.

I would think that NYYRIKKI is using the VRAM for double buffering. On an MSX1 I guess the update would flicker a bit. But its possible to reduce the flicker by updating the screen a bit random.

I wonder how expensive it would be to create a new ROM mapper that supports say 32MB ROM. It shouldn't be too hard or expensive and then no HD would be needed.

I think it would be extremely cool to see an MSX1 ROM with full screen video and a PSG sample player

Development of such rom mapper would be quite easy with an emulator and then its just a matter of doing real HW once its working.

The double buffering should be no problem on MSX1 either. What is, is the slower VDP access. Outside the vblank you can't use otir, let alone outi. That's going to give a serious dent in the framerate.

I wonder how expensive it would be to create a new ROM mapper that supports say 32MB ROM. It shouldn't be too hard or expensive and then no HD would be needed.

I think it would be extremely cool to see an MSX1 ROM with full screen video and a PSG sample player

Development of such rom mapper would be quite easy with an emulator and then its just a matter of doing real HW once its working

And for games like Dragon's Lair it would lower the systems requirements a lot (no need of HDD). And add the plug-and-play feature!

Matra is selling INK for only 19.99 euros. It must be possible to produce such cartridges at a low cost, ask them!

@wolf_
I am not so convinced, animation can still mask errors a lot, (if you keep the framerate high),
the result could be acceptable, maybe not for any scene, and could improve using interframe
enconding, in oder to maximize the colorclash masking

I see what you mean,
the result cannot be considered acceptable for any practical use,
nevertheless it could be interesting as experiment across visual art,
tecnology and computer science...

OK it is crap

unless you do not fully update the Pattern Generator Table and Pattern Color Table
but you update also the Pattern Name Table,

a good compromise could be that you update the Pattern Name Table, and only some of
the tiles in the Pattern Generator Table and Pattern Color Table, so to get a limited I/O
and still you can try to minimise somehow the distortion...

anyway I cannot see a simple algorithm to solve the joint optimization of the 3 Tables
where you can change only a limited part of the fisrt two...