(Fwd) Re: MISC video spec (640x480 ?)

[wmor1@xxxxxxxxxxxxxxxxxxxxx]

> From:           Penio Penev <penev@venezia>
> 
> > This is the same sport of suggestion I was thinking about when I was 
> > talking about  doctoring the video signal to modulate sound onto it.
> > 
> > If we' re  going to get onto the subject of "just add a sound chip" 
> > (as happened last time)
> 
> What happened last time is that it was pointed out that the NTSC video
> signal does _not_ contain information about sound, so whatever you do to
> it, and however you modulate it, no TV wit a/v jacks (versus RF antenna
> jack) will produce any sound from the video jack.  For this you need to
> connect an additional cable (for mono, or two additional for stereo) with
> the audio signal on them. 

What you Seem to forget is that  video signals go to modulators and 
are modulated with sound signals (the other input into full 
video/sound modulators) onto the baseband signal to produce RF 
signals.

I think we are at opposite ends of the same rope here.  I have been 
talking about doctoring a video input into an RF modulator which 
produces a resultant signal that the TV looks at for the sound 
signal as well.  Then using the interference between the video signal 
and the sound signal in the spectrum  (I think I read it in the MCGraw
Television Handbook, and have observenced it) to produce an effect
(maybe not the greatest effect) from this.  That is as long as the
modulator (or maybe TV) did not try to eliminate it.  While you have
been talking about a video signal that goes to an AV jack which of
course, as you say, won't produce sound (or interference), because the
TV doesn't look at it for sound.  Sorry we have got mixed up here, I 
think we could be wasting our time.
 
> If you want to do sound with a second chip, this might be actually a
> second P21, which can be extorted to sing via its video pin.  I don't know
> yet how good you can make it, but word is that 10 bits at 22KHz mono is
> definitely possible. You'll have to add DRAM, though. 

Well conceptually I was thinking of just two chips NVRAM and Mup21.

> 
> > Still I have  a practicle suggestion, if Jeff is thinking of 
> > interfacing extra chips to the  F21.  The Neograph (www.neograph.com) 
> > magic graph chip offers an 32-bit Vesa local  bus interface, 1 M-byte 
> > on chip  dram and S-VGA chip ll in one.  Using  the on chip dram as 
> > main memory it  should be possible to produce  an two chip computer 
> > (pluss logic), I was actually thinking of using it on the P32 myself, 
> > but it hasn't adventuated.
> 
> As Jeff pointed out, if you wanted to add many chips, this could be 3 F21s
> to handle the R, G, and B, _and_ alleviate the memory bottleneck by doing
> that.  Of course someone might say that F21 is somewhere between vaporware
> nad pre-release, but we all know that soon the story will be different. 

I was just responding to Jeff's suggestion of trying to interface 
other chips (non F21's).

> 
> > Further still to the chips from Silicon Graphics,  Phase 5 
> > (www.phase5.de, or something) in Germany is producing a next  
> > generation type Amiga (Abox), the heart chipset is  a 128-bit device  that 
> > people think will  compete with the 02 graphics wise  (but we shall 
> > wait and see),
> 
> By the time we get to wait and see, the O2 may be obsolete :-) 

Well  my note of caution exactly,  ;) I'm waiting (before I decide to 
take it on) to see if it turns up at all and if so if it turns up to 
spec.

> 
> And, the main O2 feature is not the stunning graphics, but the fact that
> _while_ you render texture-mapped 3D at frame rates you still have
> available 700MB/sec bandwidth between the memory and the processor, _and_
> can decompress and play full frame full motion MJPEG from disk, _and_ you
> can do 10MB/sec ftp, _and_ you can do CD-quality stereo audio in-out,
> _and_ the machine be responsive in a shell.  While doing all of the above
> simultaneously, you still haven't even touched the reserved 267MB/sec
> bandwidth of the PCI64 bus, which isn't needed for any of the above. 

Exactly, 1600Mb's is  the spec for the A/BOX compared to the O2's 
2400Mb's, and they are talking about doing simular things..

> > Next  on the hit list is that Microsoft has apparently donee 
> > something really cool.  They have invented a 3-D graphics standard 
> > caalled Talsiman.  The claim looks like 1.5-2 million pixels per second 

sorry 1.5-2 million polygons per second.

> > on a $300-$500 doller card.  

> $300-$500 doesn't sound like consumer electronics to me.

Sorry,  I just read $200+ for the workstation performance 3-D 
accelerator to the consumer, they were talking about $50 wholesale 
for the Media processor a few years ago. 



----------------------------------------------------------
Wayne Morellini  <wmor1@student.monash.edu.au>

Post Graduate Student Representative.
Rusden Campus, Deakin University, Vic, Australia.

GradDip Media Studies (Current), Bach InfoTech & AD Business(Computing).
----------------------------------------------------------