Sdif files (made with spear) Reader
Hi,
I am trying to use sdif files created with Spear for additive resyntsis purposes. I need at least 1024 oscillators for that resynthesis (sinusoids is not enough: just 256 oscs).
I was thinking to use oscbank (with up to 1024 oscs), but I need a sdif files made with Spear reader, in order to get values for individual partials. Do you know if there is an external to deal with Spear-created sdif files?
Thank you very much.
Carlo Laurenzi
Quote: carlo-laurenzi wrote on Fri, 14 September 2007 11:05
----------------------------------------------------
> Hi,
> I am trying to use sdif files created with Spear for additive resyntsis purposes. I need at least 1024 oscillators for that resynthesis (sinusoids is not enough: just 256 oscs).
> I was thinking to use oscbank (with up to 1024 oscs), but I need a sdif files made with Spear reader, in order to get values for individual partials. Do you know if there is an external to deal with Spear-created sdif files?
> Thank you very much.
>
> Carlo Laurenzi
----------------------------------------------------
I just put version 1.7.6 of sinusoids~ on the CNMAT downloads page, it handles 1024 sines. Only the Mach-O version is up right now, windows and cfm will come shortly. See:
BTW: I've made very good sounding models with far fewer sines.
mzed
----- Original Message -----
From: "mzed"
To:
Sent: Friday, September 14, 2007 9:57 PM
Subject: [maxmsp] Re: Sdif files (made with spear) Reader
>
> I just put version 1.7.6 of sinusoids~ on the CNMAT downloads page, it
> handles 1024 sines. Only the Mach-O version is up right now, windows and
> cfm will come shortly. See:
>
> http://www.cnmat.berkeley.edu/MAX/downloads/
>
> BTW: I've made very good sounding models with far fewer sines.
>
> mzed
>
>
> --
> || michael f. zbyszynski -- molecular gastronimist
> || http://www.cnmat.berkeley.edu/
> || http://www.mikezed.com/
>
Thank you very much!! It's just what I was lookong for.
What I need to do is to build a program that automatically analyzes an audio
piece of music (up to 3 hours), writes the results in an Sdif file, then
makes some spectral transformations (cut some frequency bands, excite
others, apply spectral envelopes, etc..) and finally resynthesizes the file
in wav format, ready to be played.
Since my audio material is quite complex, I need the best resolution I can
get (in this case, 1024 oscillators, with your newly updated sinusoids~).
I will use Spear to get the Sdif files out of my audio material, and I will
write a parser that can spit out values good for working with sinusoids.
Can you suggest another way to do all this? Am I missing something?
Thanks
Carlo Laurenzi
> I just put version 1.7.6 of sinusoids~ on the CNMAT downloads page, it handles 1024 sines. Only the Mach-O version is up right now, windows and cfm will come shortly.
You're probably going to knife me for this... but any chance of a 1024
resonators~? Don't feel bad if not. I've made due so far. Quite
well even. Thanks for the 1024 sinusoids~. Getting lots of mileage
out of the new distribution, so thanks again.
Keith
Hi Michael,
since I need to apply spectral transformations to quite complex sound
material converted into sdif format before resinthesizing it in wav, I need
that Threefates~ also is able to handle 1024 partials. Threefates will drive
sinusoids (with 1024 oscs).
Could you make an upgrade of Threefates also?
Thank you very much,
Carlo Laurenzi
P.S.
What about "SDIF-stft~" e "SDIF-istft~"?
Do you know something about them?
----- Original Message -----
From: "mzed"
To:
Sent: Friday, September 14, 2007 9:57 PM
Subject: [maxmsp] Re: Sdif files (made with spear) Reader
>
> Quote: carlo-laurenzi wrote on Fri, 14 September 2007 11:05
> ----------------------------------------------------
>> Hi,
>> I am trying to use sdif files created with Spear for additive resyntsis
>> purposes. I need at least 1024 oscillators for that resynthesis
>> (sinusoids is not enough: just 256 oscs).
>> I was thinking to use oscbank (with up to 1024 oscs), but I need a sdif
>> files made with Spear reader, in order to get values for individual
>> partials. Do you know if there is an external to deal with Spear-created
>> sdif files?
>> Thank you very much.
>>
>> Carlo Laurenzi
> ----------------------------------------------------
>
> I just put version 1.7.6 of sinusoids~ on the CNMAT downloads page, it
> handles 1024 sines. Only the Mach-O version is up right now, windows and
> cfm will come shortly. See:
>
> http://www.cnmat.berkeley.edu/MAX/downloads/
>
> BTW: I've made very good sounding models with far fewer sines.
>
> mzed
>
>
> --
> || michael f. zbyszynski -- molecular gastronimist
> || http://www.cnmat.berkeley.edu/
> || http://www.mikezed.com/
>
>
>
Im curious, how do you feed the SDIF spear files to sinusoids? Ive tried using the spear text files but it seems the data is not formatted correctly, and also It seems a very complex task to play a sequence of spectral information in proper speed. Im just curious if you have had useful results integrating spear and sinusoids, and if you would be willing to share a file parser to that effect.
Anyone used the Metasynth spectrum synth? (not the image synth) Sounds a bit like spear but has this great tone quality I cant put my finger on.
>Ive tried using the spear text files but it seems the data is not formatted correctly<
In my experience - feeding a max-patch with SPEAR txt-files by hand - the SPEAR txt.flies ARE formatted correctly, the problem is, in my opinion, that you get one "physical" overtone represented by two or even more "lines" in the SPEAR-Analysis. Thus the textfile looks confusing to you, or your patch.
I'm thinking , for some months, of a way to "clean" this files, to get a spectrum, with a constant number of continous lines. Probably with any statistical clustering methods...
Gerald
I am working just now on a parser for SPear files to feed sinusoids. I have
not finished yet.
What I saw is that using "Sdif-tuples" and "zl group" I can automatically
store Spear data in a coll object in the correct order of frames and
partials.
If the parser I am doing will work properly I will share it on the forum.
Carlo
----- Original Message -----
From: "Nicholas C. Raftis III"
To:
Sent: Sunday, September 16, 2007 10:49 AM
Subject: [maxmsp] Re: Re: Sdif files_ 2 - for Mzed
>
> Im curious, how do you feed the SDIF spear files to sinusoids? Ive tried
> using the spear text files but it seems the data is not formatted
> correctly, and also It seems a very complex task to play a sequence of
> spectral information in proper speed. Im just curious if you have had
> useful results integrating spear and sinusoids, and if you would be
> willing to share a file parser to that effect.
>
I tried making something like that recently but getting coll to dump
a long list that is stored in some of the indexes didn't work. How do
you get around that?
Peiman
On 16 Sep 2007, at 12:16, Carlo Laurenzi wrote:
> I am working just now on a parser for SPear files to feed
> sinusoids. I have not finished yet.
> What I saw is that using "Sdif-tuples" and "zl group" I can
> automatically store Spear data in a coll object in the correct
> order of frames and partials.
> If the parser I am doing will work properly I will share it on the
> forum.
>
> Carlo
>
>
>
> ----- Original Message ----- From: "Nicholas C. Raftis III"
>
> To:
> Sent: Sunday, September 16, 2007 10:49 AM
> Subject: [maxmsp] Re: Re: Sdif files_ 2 - for Mzed
>
>
>>
>> Im curious, how do you feed the SDIF spear files to sinusoids?
>> Ive tried using the spear text files but it seems the data is not
>> formatted correctly, and also It seems a very complex task to play
>> a sequence of spectral information in proper speed. Im just
>> curious if you have had useful results integrating spear and
>> sinusoids, and if you would be willing to share a file parser to
>> that effect.
I don't think that metasynth has a partial tracker like spear.
p
On 16 Sep 2007, at 09:49, Nicholas C. Raftis III wrote:
> Metasynth
I'll mention this again, but it doesn't do SDIF files, only works on
SPEAR text-formatted files, and it *definitely* isn't
ready-for-prime-time. Michael and I worked it up as a class
exercise for fun a few years ago:
and here's an earlier class with some 'standalone' unix-execs that
show how the data is masssaged:
CodeWarrior files, but I've also noticed that the translation to
Xcode isn't too horrible.
Quote: carlo-laurenzi wrote on Sun, 16 September 2007 05:16
----------------------------------------------------
> I am working just now on a parser for SPear files to feed sinusoids. I have
> not finished yet.
> What I saw is that using "Sdif-tuples" and "zl group" I can automatically
> store Spear data in a coll object in the correct order of frames and
> partials.
...and then you can connect two objects which can only 512 something so that you have 1024 something without waiting for an update to the object.