Digital designs?

[The Dark force of dance]

Y-ellow.
        I'll say this just one more time. Triple DAT _does_not_ have a DSP
on-board.

At 03:53 PM 1/5/98 BST, Martin Fay wrote:

>Doesn't TripleDAT gain a touch of a boost from the DSP on the bundled 
>soundcard? I'm pretty sure I couldn't get near that with Cubase VST 
>(or whatever else)...

With risk at being perceived as an advertisement for TripleDAT and not
wishing really to harp on about this, there are some good reasons why VST
and it's ilk can't do this stuff effectively. It can be summed up in one
word. "MICROSOFT". Microsoft is like music from a major record company.
"Mindless mass for the mindless masses." What file format do they all use?
"WAV FILES" of course. And are wav files suitable for professional digital
audio. Not even close. They all use them of course because people expect
them to use them. Wav files have unfortunately become the defacto standard
on PCs and whilst they are suitable for a range of multimedia presentations,
they are far from suitable for digital audio.

Although Triple can freely convert back and forth to/from WAV files, it's
own file format is very close to an actual digital audio stream. That means
that it contains all the sub-codes which give digital audio meaning. It
differs only in that it is optimized so that unlimited pointers can be run
through it. Quickly picking off points within a single or multiple audio
files. Someone once complained to me that audio was slipping round in one of
these programs. It could have been cuebase or logic. I forget which. This is
because WAV files do not sync at word-clock level. Hell they don't even know
about word clock and yet without word-clock, multi-track digital tapeless
audio is pretty dificult to do. When I think about it, I'm fairly impressed
that these toy audio programs can hold multiple tracks together at all. This
whole approach allows TripleDAT to do things with a minimum of effort.

What I'm suggesting here is that what it takes is to think things through
like that. You have to find the most efficient way of doing things without
compromising performance.

And just for succinctness. The tripleDAT card performs functions beyond mere
sound cards. In spite of providing a digital and analogue audio interface
with switching between 3 digital audio inputs and 2 outputs, the Card also
provides the following. Remote control for just about ever known DAT
machine. This allows it to automatically back up raw data to DAT as well as
totally automate a mix down. It gives the software a dongle-like feature in
that the software will only run if there is a TripleDAT card present.
Although it will talk to ordinary sound cards if they're up to the
challenge. And of course it provides a MIDI interface. Because the
hardware/software will sync at word-clock level. A second card can be added.
This gives you 2 stereo analogue and 2 stereo digital I/O sets. Or you can
add the new T-16 which gives you 8 pairs of audio as well as 2 ADAT
interfaces. All useable at the same time but that's dependent on the speed
of your machine of course. If you're using a Maxtor hard drive forget it.

The cards themselves only provide hardware functions. No DSPs. The
processing is done by the CPU and Co-pro. The whole thing is buffered in
system RAM. TripleDAT asks for half the system RAM up to 8 meg in a 16 meg
machine. If you have 32 meg as do I, then you can specify that it use more
if you wish. Though it has to be said that 8 meg of buffer is usually plenty
to play back 32 tracks. Everything else is done on the fly by the CPU.
Including volume and pan changes. No manipulation of the recorded tracks is
done on a permanent basis unless you specifically copy a track through the
FX processors. This point should not be lost on you. It all but eliminates
dithering. The single biggest complaint about digital audio. The software is
smart enough to know if you've over stepped the power of the machine BTW. If
you add too many effects at one time, the playback/record will stop and you
will get an error message. That's when you know you need a faster machine.
But as I said, On a P166 you can easily do all your vocal processing in one
sweep. Compression, gating, reverb and eq. And you can tweak all of them in
real time till it sounds sweet.

But the system is this capable because they actually thought about it. I'm
sorry to go on about it. Creamware certainly aren't payin' my bills and in
fact they've taken enough money off me over the years. But you did ask. I
don't want to comment on this any more. I think my point has been made and
we've strayed well away from the original subject. All these software
systems are getting better at an exponential rate. Let me put it this way. A
few years ago, David Johnston the author of CoolEdit said to me that. "One
day, computer systems will become so powerful we won't need extra DSP
hardware." I laughed. Now I think I should eat my words. It's looking a hell
of a lot like he was right. But on the other hand. Imagine what you could do
with some suitable add-in hardware.

be absolutely Icebox.
 _ __        _                              
| "_ \      | |         batzman at gist.net.au / aek at gist.net.au
| |_)/  __ _| |_ ____           ALL ELECTRIC KITCHEN               
|  _ \ / _` | __|___ |  The new AEK CD, Elementary Urban sanity    
| |_) | (_| | |_  / /   is out now on Transmission Communications from
|_,__/ \__,_|\__|/ /    http://brisbane.dialix.com.au:80/~com11
                / ,__   and check out http://www2.gist.net.au/~aek/
Goodfortune    |_____|       Let your SANITY wear thin.