Re: Echo II sound cards

[anoneds@netscape.net (Ben Yates)]

mjmahon@aol.com (Michael J. Mahon) wrote in message news:<20040329212800.15247.00000279@mb-m04.aol.com>...
> Ben Yates replied:
> 
> >mjmahon@aol.com (Michael J. Mahon) wrote in message
> >news:<20040329130730.15684.00000286@mb-m04.aol.com>...
> >> Ben Yates wrote:
> 
> <snip>
> 
> >> >I wonder why no one has ever tried to interface a TI speech
> >> >synthesizer to any other type of system. I have other systems and if I
> >> >had time I'd love to try it.
> >> >But still someone would have to write a good text-to-speech system to
> >> >go beyond the built-in words.
> >> 
> >> The Echo _is_ a TI speech synthesizer interfaced to an Apple II,
> >> so I'm not sure what you're getting at here.
> >
> >I mean a "TI speech synthesizer", literally, the box TI made that
> >plugs into the side of the 99/4A. There are a lot of those floating
> >around, and they can be had for cheap. The interfacing is simple. And
> >with this solution, you have the 300 built-in words in the speech
> >roms. And can send direct LPC strings.
> 
> No doubt because it has a specialized interfact to the 99/4 bus,
> and only provides a TI synthesizer chip and a ROM containing
> TI 9900 software to interface to TI BASIC and a 300 word vocabulary.
> 
> This is not of much value on a disk-equipped Apple II, which could
> have a large vocabulary on disk, or an unlimited vocabulary with
> text-to-speech software.
> 

Well, that doesn't sound much different than an Echo II. The TI
interface is pretty each to  connect to. The unit doesn't have a ROM
with 9900 software on it. It is a standalone unit in which the speech
chip is addressed directly. The ROMS contain the built-in speech
words. All the software for text-to-speech is external. The software
for making the unit speak LPC strings is external as well, but it is
simple (much more simpler than a text-to-speech algorithm).

> Echo II cards are also quite readily available and inexpensive, and
> Apple software and large vocabularies are also readily available.

Wasn't aware of that. But I do wonder if the vocabularies are
interchangeable. Meaning LPC streams could be interchanged between the
units?
Always looking to increase my vocabulary, you understand...

[snip]

> 
> BASIC compilers generally make some provision for "attached" machine
> language programs, and Pascal provides excellent tools for incorporating
> assembly language code.  I would be surprised if there were not a Pascal
> "unit" to interface to an Echo card, but if not, it would not be hard to make
> for anyone motivated to do so.

Not all BASIC compilers do. You are right about Pascal, but for some
of us attempting a text-to-speech engine is above our heads.
Especially when we do not desire unlimited speech in an application,
but a set of speech words and phrases that are not built-in.

> 
> It would also be possible to "capture" the LPC streams for desired
> utterances, if that were preferred, but this approach would also require
> M/L interface code at run time to pump the streams to the chip.

But that code is much easier than a text-to-speech engine, correct?
I've done that myself, even in BASIC using POKEs directly to the chip.

> 
> The situation is that speech was never a widely available option on
> Apple II's--meaning more than 10% of the systems--so it did not
> achieve "critical mass", and did not get the software support that
> could have made it more useful.
> 
> At one time, TI hoped to have a self-sustaining business in creating
> new custom "vocabularies" for its synthesizer chip, and so did not
> promulgate software for allowing others to easily generate the LPC
> streams.  In retrospect, this was probably a mistake.

Yes. I remember trying to purchase a TI card for a PC around '89. It
had come out only a couple of years before. It allowed one to create
LPC speech on this PC card. I didn't even know if it created
compatible LPC, but TI simply gave me a list of their non-affiliated
support centers, none of which had a clue!

Ben