[Audacity-nyquist] Merry Christmas from Germany...

[<edg...@we...>]

Hi everybody

I'm the self-soldered guitar effects guy from germany.

First of all thanks to Roger for his work on this great
analog-synth like toolkit (in fact it is much more)
and thanks to Dominik for making me discover this by 
playing around with Audacity (also Audacity is much more
than only a play-toy) I'm watching Audacity since Summer
2000 and ever wondered about the very professional basics.
Now that i've read Roger's web-sites I think I've 
understood. It's more like university research for audio
programming.

I'm very interested in realtime filters and dynamics and 
here in germany we just do experiments to make audacity
get some real-time "effects" by using several tracks at
once together with multichannel nyquist modules. At this
time there is already a nyquist 3/5/7-channel realtime
equalizer out on test in germany. Markus Meyer knows about
but is very busy so I do not know wether he had the time
to make you know. If anybody's interested please send a
mail to edg...@we... - I will send back the zip-file
(but first I have to translate the explanation files).

I also "discovered" the snd-abs and snd-sqrt problem but
David was faster than me. (Thanks to David for his mail - 
as you can see i'm also on the nyquist list now).

I really need the snd-abs function for very fast snd-follow
with raise- and fall-times of about 1 Millisecond only (!).

Thanks to Roger for offering a lisp-replacement but lots
of snd-fetch (or similar) is much to slow for a broadcast
solution because I have to normalize files of several hours
length. (We just reconstruct the broadcast studio of an
independent (freeware) radio station here in germany).
(www.querfunk.de). I can help myself with a vst-limiter
but what I really need is a "broadcast audio level for
stupids"-function. As this is not possible without clipping
I have to simulate "eardrum distortion" by a very fast
compressor and therefore I have to simulate "doubleway
rectification" (I'm a hardware guy). 

Explanation for non-hardware people:

I have to make all negative samples become positive samples
so the resulting envelope from snd-follow will have the
double resolution and enables twice-as-fast calculations.
I've tested this with multiplication (mult s s) - the trick
works but to compensate the miscalculation by mult I whould
need the snd-sqrt function but this is also missing....

It's not too hurry - Nobody can expect that freeware
programmers work 48 hours a-day, they sometimes have to 
do things like working for money and most of them also have
a family to care about.

Unfortunately I'm not good enough in C (what is no excuse for
not to learn more) but in fact I have to confess that I'm 
just reading the Touretzky (once again) to better understand
the xlisp-extensions. 

Do not misunderstand - do not change the manual! - I hate
manuals where I get explained how much "better" and "faster"
everything is now - over and over again. In fact nothing is
faster because I have to read such a crap everytime I'm
searching for help. Leave it like it is. For people who don't
know enough about the basics (like me) there are enough good
books out there in the Internet (The Touretzky is for free).

This was a very long christmas story (people from other 
religions: please make me know what you're celebrating
because I wanted everybody to be included - i'm only just
a too big ignorant to know).

So I want everybody to have a very nice christmas 
(or whatever - see above) - and a happy new year.

edg...@we...

P.S - If anybody wants to play around with the equalizers
please mail - if no mistakes were found we can give it to
the users list (the equalizers are freeware - you can give
it to anybody you like - but please first test it).

__________________________________________________________
Mit WEB.DE FreePhone mit hoechster Qualitaet ab 0 Ct./Min.
weltweit telefonieren! http://freephone.web.de/?mc=021201