DSP sound/audio demos

Thanks for this very useful and interesting information. I hadn't thought of the 
bucket-brigade device in this context, maybe it's the exception that proves the 
rule?

robert bristow-johnson wrote:

...
> 
> might not be the ones you wanted.  are you doing an article for Electronic
> Musician?
> 
Nothing so glamorous, sadly; I am involved in software for electro-acoustic 
composers (the Composers Desktop Project system), and I want to augment our 
existing documentation on the filters with some reference to analog filters, 
etc, given the popular enthusiasm for analog-modelled processing these days. We 
are aiming at schools and colleges as well as e/a composers (and educational 
work is very much part of oure remit), so I need to write in technical terms 
that are accurate so far as they go, but not OTT for school-kids or for 
"non-technical" musical users, so I can write about phase (with lots of 
pictures!); but the continuous-time distinction may be a little too, er, 
academic for these users. It's an ongoing tightrope - we want to increase the 
appreciation of the engineering aspects, and of course not write anything that 
is misleading or downright wrong, but not go so far as to put people off!

Richard Dobson

Richard Dobson wrote:
> 
> I have this irresistible assertion, that analog filters cannot exactly achieve
> linear phase FIR style, so that the FIR filter has no exact counterpart in the
> analog domain. Is this is a correct assertion?

That is incorrect indeed. There is a number of analog filter designs
based on the analog delay lines with the taps. The delay line may be
implemented as a spring with the acoustic wave, for example.

>  My web searches, etc, have found
> things like Bessel filters which are "almost" linear phase in the passband (I
> assume these are favoured for speaker crossover filters).

There are lots of myths, misunderstanding and misconceptions about the
audio. 

 In short, I seem to
> find this "almost" crops up everywhere. In effect this is also asking if an IIR
> filter can be linear phase, and here too I hit "almost" everywhere I have looked
> so far.

An IIR filter never can be an exact linear phase. However it can be as
close to linear phase as you like it to be.
 
> Turning this question around, and making it more general, how accurate a copy
> can the digital filter ever be, of an analog one?

An analog filter has response from 0 to infinity. The digital filter
response is from 0 to fs/2. Due to that reason the digital filter can't
be an exact copy of an analog filter.

> I know about the bilnear
> transform and frequency warping; but what else is there that makes life
> difficult for the designer trying to make a copy of a much-loved and
> idiosyncratic analog filter, that will pass the analog-synthesist-junkie test?

You can't solve the psychology problem with technical methods. 

Vladimir Vassilevsky

DSP and Mixed Signal Design Consultant

http://www.abvolt.com

On Sun, 16 Nov 2003 13:47:18 -0500, robert bristow-johnson
<rbj@surfglobal.net> wrote:

>>(this DSP book by Grover and Deller describes the
>>phase response of the elliptical filter as "drunk fly on cross-country skis
>>in tornado" so don't use elliptical is phase linearity is important to you),

I have a set of "delay-equalized" passive analog lowpass filters
from a company in California called TTE.  They are 5th-order
elliptic filters, and exhibit very nearly linear phase in the
passband and somewhat into the transition band.  The specs say
+0/-0.25 dB passband flatness and within +/-3&#4294967295; of linear phase
from DC to Fc.

Last I checked, TTE no longer made these filters, but they are
interesting examples of what can be done with nothing but
resistors, capacitors, and inductors.

Greg Berchin

Hello Richard,

These devices have been around for awhile. I use to get them at Radio Shack
back during the '70s. If you remember the "echoplex." It was an audio tape
delay whose output was fed back to its input. The read head was on a slider
and you could move it back and forth to change the delay. A friend and I
made some digital versions using the SAD devices. They worked okay. Back at
least the tape loop didn't wear out.

Here's a description of a SAD device:

http://www.geofex.com/sad1024.htm


Clay



"Richard Dobson" <richarddobson@blueyonder.co.uk> wrote in message
news:42Utb.219$537.128@news-binary.blueyonder.co.uk...
> Thanks for this very useful and interesting information. I hadn't thought
of the
> bucket-brigade device in this context, maybe it's the exception that
proves the
> rule?
>

In article <f56893ae.0311160456.20f2d245@posting.google.com>, allnor@tele.ntnu.no (Rune Allnor) wrote:
>Hi All. 
>
.. Stuff deleted ...
>or reinforcing my already firm reputation as "the mad scientist" of the 
>neighbourhood.
>
One of the guys where I work built a refrigerator and another built a rail 
gun.  You have a ways to go to get to "mad scientist" based on DSP.

Rune Allnor wrote:


> So, what DSP demos could be included here?

Reversing the input is fun, also making speech sound like a Cylon by 
modulating it with something like 50 Hz.

Leon

Vladimir Vassilevsky wrote:
> Richard Dobson wrote:
...
> > In short, I seem to
> > find this "almost" crops up everywhere. In effect this is also asking
> > if an IIR filter can be linear phase, and here too I hit "almost"
> >  everywhere I have looked so far.
> 
> An IIR filter never can be an exact linear phase. However it can be as
> close to linear phase as you like it to be.

I have no problem with linear-phase IIR filters - they just aren't
realizable (imagine a symmetric "FIR" with an infinte number of
coefficients). "As close as you like" is the key phrase, however. No
filter is perfect - they all approach some kind of design criteria
"close enough".

> > Turning this question around, and making it more general, how accurate 
> > a copy can the digital filter ever be, of an analog one?
> 
> An analog filter has response from 0 to infinity. The digital filter
> response is from 0 to fs/2. Due to that reason the digital filter can't
> be an exact copy of an analog filter.

This can be problem, but it is not insurmountable. Again think "close
enough".

Perhaps you'll find this useful:
http://www.weiss.ch/eq1/images/brochureEQ1-LP.PDF

If you skip the nice pictures, you'll get some information on the
second page. It also offers an oppinion on why FIR filters are
unsuited for audio equalization.

In another thread I posted this link:
http://www.audiosignal.co.uk/Gerzon%20archive.html

Also an interesting read when discussing linear vs. minimum phase for
audio.

Regads,
Andor

On 16 Nov 2003 04:56:14 -0800, allnor@tele.ntnu.no (Rune Allnor)
wrote:

>Hi All. 
>
>The last few weeks I have been asked by a few people (neighbours, 
>neighbours' kids) what I do for a living. And I find that it's not 
>easy to explain verbally what DSP is about. Terms like "Discrete 
>Fourier Transforms", "Singular Value Decompositions", "Rayleigh waves" 
>and the likes don't do much for lay people, except perhaps establishing 
>or reinforcing my already firm reputation as "the mad scientist" of the 
>neighbourhood.
>
>So, inspired by this and some remarks Al Clark made in the DSP book thread, 
>I've started thinking what demos could be useful to show people what 
>at least parts of DSP is about. I assume I have a laptop PC available 
>where matlab is installed and that has a microphone and a sound recording 
>program. With this system I can record somebody speaking a sentence,
>store the sound to a .wav file that I load into matlab. I then manipulate 
>this sound by means of DSP in matlab, and play the manipulated result over 
>the PC speaker or headphones.
>
>So, what DSP demos could be included here? Remember, I'm not an audio 
>specialist so the most psycho-acoustical stuff would probably best be 
>left out. Implementations should not be too difficult either, or at 
>least good algorithm recipes should be available. Pitch shifting? 
>LPC-errors (somebody told me these things sound like "robot voices" 
>from a 1977 sci-fi movie)? Adding echos/reverbrations? Other things?
>
>Rune

Hello Mad Scientist,

  an audio demo is a great idea.

To add to the other suggtestions here:

I'd start with letting the kids hear a pure tone, 
and see its waveform in time, plus a spectral plot.
Tell them the time-domain plot is just like the 
position of a plucked guitar string.
  
Next let the kids, themselves, change the freq of 
the tone to see what the change looks like in time 
and in the spectral plot.  

Then let the kids hear and see two tones.  Explain 
filtering and then filter out one of the tones, 
letting the kids see and hear the filtered signal.

Let the kids experiment.

Maybe a demo using the telephone DTMF tones would 
be useful.  All the kids will be familiar with 
the DTMF tones.  

You might record a few words of speech and then add
lots of high frequency noise.  Let the kids hear 
and see the noisy speech signal.
Then start filtering out the high freq noise 
and let the kids hear and see the filter output.
Let the kids control how much the noise is attenuated. 
They'll see that reducing the noise more and more 
(watching the noise spectral magnitudes drop) makes 
the speech easier and easier to hear and understand.

After that, you can show the kids how to 
invert a Toeplitz matrix based on cyclotomic 
polynomials in tight Gabor frames.

[-Rick-]

On Sun, 16 Nov 2003 13:47:18 -0500, robert bristow-johnson
<rbj@surfglobal.net> wrote:

  (snipped)
>
>now there are (or used to be) these devices that used to be called "bucket
>brigade delays" (i think the legit name was "charge-coupled devices" which
>operated on sampled analog signals.  that is, the analog signal was sampled
>in time (so it became a discrete-time signal rather than a continuous-time
>signal) but it was never digitized so it remained an analog voltage that
>lived as discrete samples on capacitors that, using some fancy solid-state
>switching technology, got transferred from one capacitor to the next in
>line.  those capacitors could be wired to high input impedance (FET) op-amps
>and the voltages of these could be weighted and summed to a summing
>junction.  even though it's done as analog, it is essentially the same
>mathematical operation that is an FIR filter.  that would be an "FIR style"
>analog filter and could just as well be linear-phase if the weighting
>coefficients where symmetric.

(snipped) 

Hi,
  we used to use linear-array CCDs, along with a laser 
and Bragg cell, to perform spectrum analysis.
(It worked because a glass lens performs a Fourier 
transform, I think.)

Don't remember all the numbers but we could perform 
spectrum analysis of a signal whose bandwidth was, say,  
5 Mhz (this was back when A/Ds could only sample 
at a few MHz/sec rate).  The spec resolution was, say, 
5/4000 MHz.  The neat part was that the spectral 
results where updated at, say, a 10 kHz rate.

All in analog!  

The above numbers may be "way off", but 
at that time, no DSP system could 
even come close to such a wideband, high-speed, 
spectrum analyzer.

See Ya',
[-Rick-]

NOTAM has made an interesting CD (software + sounds + tutorial texts) that may 
get close to what you are suggesting:

http://www.notam02.no/DSP/index-e.html

It is also called "DSP for Children":

http://www.notam02.no/~joranru/DSPforChildren.html


It has a clear electro-acoustic music bias (sonic art, soundscapes, etc), but 
does deal with quite a lot of "the basics". It can be ordered via the Electronic 
Music Foundation:

http://www.emf.org/

This is very much the sort of thing we are interested in at CDP, and I am 
therefore especially interested in what this list has to say about the CD.


Richard Dobson

Rick Lyons wrote:

...
> To add to the other suggtestions here:
> 
> I'd start with letting the kids hear a pure tone, 
> and see its waveform in time, plus a spectral plot.
> Tell them the time-domain plot is just like the 
> position of a plucked guitar string.
>   
> Next let the kids, themselves, change the freq of 
> the tone to see what the change looks like in time 
> and in the spectral plot.  
> 
...