Re: Structured Programming using Forth

On Apr 6, 3:24 pm, John Passaniti <n...@xxxxxxxxxxxxxxxxx> wrote:
rickman wrote:
They use external memory, but it is not so they can run a Forth VM. I
don't recall where it read it, but they use the high speed comms to
download snippettes of forth as required. Chuck has even written code
that does this in very few instructions and then runs the code that
loaded starting just after the loader, IIRC.

Jeff's most recent message confirms that at least some at Intellasys are
indeed creating some form of virtual machines. I don't know if this was
just an experiment or a model they might use in some application. For
implementing complex communication protocols (like a TCP/IP stack), it
might make sense.

I seem to recall that the initial application for the SeaForth chip
was home theater. That is something that requires more than 16 or 18
bits typically. I forget how wide the SeaForth word is, but typically
24 bits are used for HiFi audio. Is this chip 24 bit? If not, I
expect double precision math is being used.

Don't confuse the number of bits fed to a D/A converter with the number
of bits needed needed for intermediate calculation. A home theatre
system might (on the low end) only be using 16-bit D/A converters. But
the various processing steps-- such as various matrix-audio decoding,
mixing, filtering, and so on will usually require more bits.

And that potentially is a problem. Using a SEAforth processor as a DSP
in even a consumer-level audio application would most certainly require
multi-precision MACs, and that's going to start to cut into the
processing speed of the system. So you might then use multiple
processors in parallel, but then you have to deal with the
communications between them.

It is very seldom that I understand where you get your information. I
really don't understand you. I point out that the home theater
application requires more than 18 bits and your response is that I
might be confusing the width of the DAC with the size of the
calculation. I clearly said the chip would need to be at least 24
bits which is what is commonly used for high end audio. I don't see
where you say how many bits the SeaForth chip has and yet you somehow
come to the conclusion that it "must" be using multi-precision. How
many bits is the SeaForth chip MAC? I seem to recall that Chuck is
rather fond of 20 bit words which might well suffice in a digital
audio app depending on the processing done. But my point is that your
post does not explain anything, it just jumps to several conclusions.

I get tired of trying to communicate with you. I don't know where you
are coming from and your posts are not really responsive to what I
say.


.

Relevant Pages

  • Re: web camera
    ... fine,and at no point do i lose the picture.even when we accept audio,it,s ... that there is no connectionfor the sound. ... "Chuck" wrote: ... Do you or the other person have a NAT router connecting the ...
    (microsoft.public.windowsxp.network_web)
  • Re: REPEATING AUDIO
    ... Chuck Le Mieux wrote: ... turned off wasn't heard again moving from page to page. ... Now all of a sudden the audio is heard on every page. ...
    (microsoft.public.windowsxp.general)
  • Re: 2m has gone to the dogs.......
    ... - I had to chuck mine, ... the audio, the CPu gave up! ... best scanner ever made had my first one in 1989......quality ...
    (uk.radio.amateur)
  • Re: No Audio!
    ... chuck wrote: ... all play jes' fine on the PC, ... no audio when played on the DVD deck attached ...
    (rec.video.dvd.tech)
Quantcast