linuxsampler-devel

[Linuxsampler-devel] giga velocity setting algorithm

[<ken...@ya...>]

hi,

i am not sure whether here is the right place to ask
but your help would be deeply appreciated.

if i try to set the velocity per midi note manually,
the output is not obvious to me. is there any
reference about the giga velocity setting algorithm?
so that i can use sensibly velocity setting.

Many thanks and best regards,
kenneth lee




_______________________________________________________________________
Do You Yahoo!?
Get your free @yahoo.com.hk address at http://mail.english.yahoo.com.hk

Re: [Linuxsampler-devel] giga velocity setting algorithm

[Mark K. <mar...@co...>]

On Sun, 2003-11-30 at 18:45, Kenneth Lee wrote:
> hi,
> 
> i am not sure whether here is the right place to ask
> but your help would be deeply appreciated.
> 
> if i try to set the velocity per midi note manually,
> the output is not obvious to me. is there any
> reference about the giga velocity setting algorithm?
> so that i can use sensibly velocity setting.
> 
> Many thanks and best regards,
> kenneth lee
> 

Kenneth,
   Are you talking about setting the velocity mapping in GigaSampler? Or
in some Linux tool?

   I just did some mapping work on GSt.

Mark

Re: [Linuxsampler-devel] giga velocity setting algorithm

[<ken...@ya...>]

Mark,

i mean using Linux tools. i am using Lilypond to
produce piano music sheet. pmidi to play midi files
which are generated from Lilypind throught timidity++
(now LinuxSampler). The midi files produced by
Lilypond is good enought to proof-hearing but not
adequate for listening. Therefore, some touchup is
required for each file. What i am trying to do is to
write a simple pre-lilypond processor which could be
used to produce lilypond file for generating piano
music sheet. Also, the processor should be able to
produce midi file with specific attributes per note.
Based on the above mentioned, i am tring to play
around with the duration and velocity of the midi
file. At this stage, it seems to me that the result of
different velocity setting is not obvious to me,
therefore, i am asking for advice.

Highly likely that there are some Linux tools already
available for the above mentioned, please kindly let
me know.


Many thanks!

kenneth


 --- Mark Knecht <mar...@co...> wrote: > On
Sun, 2003-11-30 at 18:45, Kenneth Lee wrote:
> > hi,
> > 
> > i am not sure whether here is the right place to
> ask
> > but your help would be deeply appreciated.
> > 
> > if i try to set the velocity per midi note
> manually,
> > the output is not obvious to me. is there any
> > reference about the giga velocity setting
> algorithm?
> > so that i can use sensibly velocity setting.
> > 
> > Many thanks and best regards,
> > kenneth lee
> > 
> 
> Kenneth,
>    Are you talking about setting the velocity
> mapping in GigaSampler? Or
> in some Linux tool?
> 
>    I just did some mapping work on GSt.
> 
> Mark
>  

_______________________________________________________________________
Do You Yahoo!?
Get your free @yahoo.com.hk address at http://mail.english.yahoo.com.hk

Re: [Linuxsampler-devel] giga velocity setting algorithm

[Mark K. <mar...@co...>]

I have not heard of anything like this. Sorry.

On Sun, 2003-11-30 at 23:40, Kenneth Lee wrote:
> Mark,
> 
> i mean using Linux tools. i am using Lilypond to
> produce piano music sheet. pmidi to play midi files
> which are generated from Lilypind throught timidity++
> (now LinuxSampler). The midi files produced by
> Lilypond is good enought to proof-hearing but not
> adequate for listening. Therefore, some touchup is
> required for each file. What i am trying to do is to
> write a simple pre-lilypond processor which could be
> used to produce lilypond file for generating piano
> music sheet. Also, the processor should be able to
> produce midi file with specific attributes per note.
> Based on the above mentioned, i am tring to play
> around with the duration and velocity of the midi
> file. At this stage, it seems to me that the result of
> different velocity setting is not obvious to me,
> therefore, i am asking for advice.
> 
> Highly likely that there are some Linux tools already
> available for the above mentioned, please kindly let
> me know.
> 
> 
> Many thanks!
> 
> kenneth
> 
> 
>  --- Mark Knecht <mar...@co...> wrote: > On
> Sun, 2003-11-30 at 18:45, Kenneth Lee wrote:
> > > hi,
> > > 
> > > i am not sure whether here is the right place to
> > ask
> > > but your help would be deeply appreciated.
> > > 
> > > if i try to set the velocity per midi note
> > manually,
> > > the output is not obvious to me. is there any
> > > reference about the giga velocity setting
> > algorithm?
> > > so that i can use sensibly velocity setting.
> > > 
> > > Many thanks and best regards,
> > > kenneth lee
> > > 
> > 
> > Kenneth,
> >    Are you talking about setting the velocity
> > mapping in GigaSampler? Or
> > in some Linux tool?
> > 
> >    I just did some mapping work on GSt.
> > 
> > Mark
> >  
> 
> _______________________________________________________________________
> Do You Yahoo!?
> Get your free @yahoo.com.hk address at http://mail.english.yahoo.com.hk
> 
> 
> -------------------------------------------------------
> This SF.net email is sponsored by: SF.net Giveback Program.
> Does SourceForge.net help you be more productive?  Does it
> help you create better code?  SHARE THE LOVE, and help us help
> YOU!  Click Here: http://sourceforge.net/donate/
> _______________________________________________
> Linuxsampler-devel mailing list
> Lin...@li...
> https://lists.sourceforge.net/lists/listinfo/linuxsampler-devel

Re: [Linuxsampler-devel] giga velocity setting algorithm

[Christian S. <chr...@ep...>]

Hi Kenneth,

not really sure what you actually want to hear, but these are the velocity 
algorithms that are currently used in LinuxSampler (/src/gig.h):

#define GIG_VELOCITY_TRANSFORM_NONLINEAR(x,dynamic,scale)	((1.0-1.0/pow(x,1.0/
(129.0-x))) * (1.0+scale/20.0) + (5.0-dynamic)*pow(x/300.0* (1.0
+2.0*scale/128.0),2))

#define GIG_VELOCITY_TRANSFORM_LINEAR(x,dynamic,scale)		((1.0
+scale*3.0/128.0)/110.0*x+(5.0-dynamic)/5.0+(5.0-dynamic)*scale)

#define GIG_VELOCITY_TRANSFORM_SPECIAL(x,dynamic,scale)		((1.0
+9.0*scale/129.0)*(1.0-1.0/pow(x,1.0/(129.0-x))+pow(3.0*x/pow(129,2),2)
+pow((5.0-dynamic)*x/500.0,2)))

These are only approximations to the ones from Gigasampler. 'x' is the MIDI 
velocity value, 'dynamic' is the dynamic range parameter and 'scale' the 
curve scale parameter from the gig format. Not sure if that helps...

CU
Christian

Es geschah am Montag, 1. Dezember 2003 08:40 als Kenneth Lee schrieb:
> Mark,
>
> i mean using Linux tools. i am using Lilypond to
> produce piano music sheet. pmidi to play midi files
> which are generated from Lilypind throught timidity++
> (now LinuxSampler). The midi files produced by
> Lilypond is good enought to proof-hearing but not
> adequate for listening. Therefore, some touchup is
> required for each file. What i am trying to do is to
> write a simple pre-lilypond processor which could be
> used to produce lilypond file for generating piano
> music sheet. Also, the processor should be able to
> produce midi file with specific attributes per note.
> Based on the above mentioned, i am tring to play
> around with the duration and velocity of the midi
> file. At this stage, it seems to me that the result of
> different velocity setting is not obvious to me,
> therefore, i am asking for advice.
>
> Highly likely that there are some Linux tools already
> available for the above mentioned, please kindly let
> me know.
>
>
> Many thanks!
>
> kenneth

Re: [Linuxsampler-devel] giga velocity setting algorithm

[Mark K. <mar...@co...>]

Groovy! Should I be downloading a new version and enjoying both velocity
and envelopes soon? ;-)

- Mark

On Mon, 2003-12-01 at 15:03, Christian Schoenebeck wrote:
> Hi Kenneth,
> 
> not really sure what you actually want to hear, but these are the velocity 
> algorithms that are currently used in LinuxSampler (/src/gig.h):
> 
> #define GIG_VELOCITY_TRANSFORM_NONLINEAR(x,dynamic,scale)	((1.0-1.0/pow(x,1.0/
> (129.0-x))) * (1.0+scale/20.0) + (5.0-dynamic)*pow(x/300.0* (1.0
> +2.0*scale/128.0),2))
> 
> #define GIG_VELOCITY_TRANSFORM_LINEAR(x,dynamic,scale)		((1.0
> +scale*3.0/128.0)/110.0*x+(5.0-dynamic)/5.0+(5.0-dynamic)*scale)
> 
> #define GIG_VELOCITY_TRANSFORM_SPECIAL(x,dynamic,scale)		((1.0
> +9.0*scale/129.0)*(1.0-1.0/pow(x,1.0/(129.0-x))+pow(3.0*x/pow(129,2),2)
> +pow((5.0-dynamic)*x/500.0,2)))
> 
> These are only approximations to the ones from Gigasampler. 'x' is the MIDI 
> velocity value, 'dynamic' is the dynamic range parameter and 'scale' the 
> curve scale parameter from the gig format. Not sure if that helps...
> 
> CU
> Christian
> 
> Es geschah am Montag, 1. Dezember 2003 08:40 als Kenneth Lee schrieb:
> > Mark,
> >
> > i mean using Linux tools. i am using Lilypond to
> > produce piano music sheet. pmidi to play midi files
> > which are generated from Lilypind throught timidity++
> > (now LinuxSampler). The midi files produced by
> > Lilypond is good enought to proof-hearing but not
> > adequate for listening. Therefore, some touchup is
> > required for each file. What i am trying to do is to
> > write a simple pre-lilypond processor which could be
> > used to produce lilypond file for generating piano
> > music sheet. Also, the processor should be able to
> > produce midi file with specific attributes per note.
> > Based on the above mentioned, i am tring to play
> > around with the duration and velocity of the midi
> > file. At this stage, it seems to me that the result of
> > different velocity setting is not obvious to me,
> > therefore, i am asking for advice.
> >
> > Highly likely that there are some Linux tools already
> > available for the above mentioned, please kindly let
> > me know.
> >
> >
> > Many thanks!
> >
> > kenneth
> 
> 
> 
> -------------------------------------------------------
> This SF.net email is sponsored by: SF.net Giveback Program.
> Does SourceForge.net help you be more productive?  Does it
> help you create better code?  SHARE THE LOVE, and help us help
> YOU!  Click Here: http://sourceforge.net/donate/
> _______________________________________________
> Linuxsampler-devel mailing list
> Lin...@li...
> https://lists.sourceforge.net/lists/listinfo/linuxsampler-devel

Re: [Linuxsampler-devel] giga velocity setting algorithm

[Christian S. <chr...@ep...>]

Es geschah am Dienstag, 2. Dezember 2003 06:45 als Mark Knecht schrieb:
> Groovy! Should I be downloading a new version and enjoying both velocity
> and envelopes soon? ;-)

First off I'm not sure if it got around here, so just a little hint: we have a 
CVS log running on http://www.linuxsampler.org showing the latest commits to 
the CVS repository, means whenever somebody commits something it will 
immediately show up on the website.

Regarding the velocity: I've already commited an initial version supporting it 
last saturday.

CU
Christian

Re: [Linuxsampler-devel] giga velocity setting algorithm

[Mark K. <mar...@co...>]

On Tue, 2003-12-02 at 04:51, Christian Schoenebeck wrote:
> Es geschah am Dienstag, 2. Dezember 2003 06:45 als Mark Knecht schrieb:
> > Groovy! Should I be downloading a new version and enjoying both velocity
> > and envelopes soon? ;-)
> 
> First off I'm not sure if it got around here, so just a little hint: we have a 
> CVS log running on http://www.linuxsampler.org showing the latest commits to 
> the CVS repository, means whenever somebody commits something it will 
> immediately show up on the website.
> 
> Regarding the velocity: I've already commited an initial version supporting it 
> last saturday.
> 
> CU
> Christian
> 

Christian,
   Hi. I hadn't looked at the web site since about 10 days ago.

   I downloaded from CVS and started building, but it failed:

g++ -DHAVE_CONFIG_H -I. -I. -I..    -pedantic -g -O2 -c -o
linuxsampler.o `test -f 'linuxsampler.cpp' || echo './'`linuxsampler.cpp
source='audioio.cpp' object='audioio.o' libtool=no \
depfile='.deps/audioio.Po' tmpdepfile='.deps/audioio.TPo' \
depmode=gcc3 /bin/sh ../depcomp \
g++ -DHAVE_CONFIG_H -I. -I. -I..    -pedantic -g -O2 -c -o audioio.o
`test -f 'audioio.cpp' || echo './'`audioio.cpp
audioio.cpp: In member function `int AudioIO::Initialize(unsigned int,
unsigned 
   int, unsigned int, unsigned int)':
audioio.cpp:89: error: invalid conversion from `uint' to `unsigned int*'
make[2]: *** [audioio.o] Error 1
make[2]: Leaving directory
`/mnt/data/mark/LinuxSampler/linuxsampler/src'
make[1]: *** [all-recursive] Error 1
make[1]: Leaving directory `/mnt/data/mark/LinuxSampler/linuxsampler'
make: *** [all] Error 2
bash-2.05b$ 

Re: [Linuxsampler-devel] giga velocity setting algorithm

[Christian S. <chr...@ep...>]

Es geschah am Dienstag, 2. Dezember 2003 14:41 als Mark Knecht schrieb:
>
>    I downloaded from CVS and started building, but it failed:
>
> g++ -DHAVE_CONFIG_H -I. -I. -I..    -pedantic -g -O2 -c -o
> linuxsampler.o `test -f 'linuxsampler.cpp' || echo './'`linuxsampler.cpp
> source='audioio.cpp' object='audioio.o' libtool=no \
> depfile='.deps/audioio.Po' tmpdepfile='.deps/audioio.TPo' \
> depmode=gcc3 /bin/sh ../depcomp \
> g++ -DHAVE_CONFIG_H -I. -I. -I..    -pedantic -g -O2 -c -o audioio.o
> `test -f 'audioio.cpp' || echo './'`audioio.cpp
> audioio.cpp: In member function `int AudioIO::Initialize(unsigned int,
> unsigned
>    int, unsigned int, unsigned int)':
> audioio.cpp:89: error: invalid conversion from `uint' to `unsigned int*'

Nothing changed in audioio.cpp since I implemented automatic plughw fallback 
(2003-11-21) and that version worked AFAIK for you. Already tried a clean 
checkout?

CU
Christian

Re: [Linuxsampler-devel] giga velocity setting algorithm

[Mark K. <mar...@co...>]

On Wed, 2003-12-03 at 02:05, Christian Schoenebeck wrote:
> > g++ -DHAVE_CONFIG_H -I. -I. -I..    -pedantic -g -O2 -c -o audioio.o
> > `test -f 'audioio.cpp' || echo './'`audioio.cpp
> > audioio.cpp: In member function `int AudioIO::Initialize(unsigned int,
> > unsigned
> >    int, unsigned int, unsigned int)':
> > audioio.cpp:89: error: invalid conversion from `uint' to `unsigned int*'
> 
> Nothing changed in audioio.cpp since I implemented automatic plughw fallback 
> (2003-11-21) and that version worked AFAIK for you. Already tried a clean 
> checkout?
> 
> CU
> Christian
> 

Yes, I tried a clean checkout yesterday, and I tried a clean checkout
this morning. In both cases I deleted the existing directory and checked
out 100% new code. Today it fails the same way.

FYI - life here is Gentoo, which means upgrades to gcc far more often
than non-Gentoo platforms. In all likelihood I've had a gcc update since
I last built the code.

Have you looked at the offending line? I don't know a 'uint' from an
unsigned int. They sound similar, but possibly not identical? 

Aah!!!! I bet I get it! I've updated Alsa to 1.0.0rc1. The offending
line looks like it's calling some Alsa function to talk to my card? Have
the Alsa folks pulled a switcheroo on us here???????

Anyway, I still cannot build as of 11:30AM California time.

Cheers,
Mark

Re: [Linuxsampler-devel] giga velocity setting algorithm

[Mark K. <mar...@co...>]

Any headway being made with the code and this uint problem? I cannot
build LS anymore.

- Mark

On Wed, 2003-12-03 at 11:32, Mark Knecht wrote:
> On Wed, 2003-12-03 at 02:05, Christian Schoenebeck wrote:
> > > g++ -DHAVE_CONFIG_H -I. -I. -I..    -pedantic -g -O2 -c -o audioio.o
> > > `test -f 'audioio.cpp' || echo './'`audioio.cpp
> > > audioio.cpp: In member function `int AudioIO::Initialize(unsigned int,
> > > unsigned
> > >    int, unsigned int, unsigned int)':
> > > audioio.cpp:89: error: invalid conversion from `uint' to `unsigned int*'
> > 
> > Nothing changed in audioio.cpp since I implemented automatic plughw fallback 
> > (2003-11-21) and that version worked AFAIK for you. Already tried a clean 
> > checkout?
> > 
> > CU
> > Christian
> > 
> 
> Yes, I tried a clean checkout yesterday, and I tried a clean checkout
> this morning. In both cases I deleted the existing directory and checked
> out 100% new code. Today it fails the same way.
> 
> FYI - life here is Gentoo, which means upgrades to gcc far more often
> than non-Gentoo platforms. In all likelihood I've had a gcc update since
> I last built the code.
> 
> Have you looked at the offending line? I don't know a 'uint' from an
> unsigned int. They sound similar, but possibly not identical? 
> 
> Aah!!!! I bet I get it! I've updated Alsa to 1.0.0rc1. The offending
> line looks like it's calling some Alsa function to talk to my card? Have
> the Alsa folks pulled a switcheroo on us here???????
> 
> Anyway, I still cannot build as of 11:30AM California time.
> 
> Cheers,
> Mark
> 
> 
> 
> -------------------------------------------------------
> This SF.net email is sponsored by OSDN's Audience Survey.
> Help shape OSDN's sites and tell us what you think. Take this
> five minute survey and you could win a $250 Gift Certificate.
> http://www.wrgsurveys.com/2003/osdntech03.php?site=8
> _______________________________________________
> Linuxsampler-devel mailing list
> Lin...@li...
> https://lists.sourceforge.net/lists/listinfo/linuxsampler-devel

[Linuxsampler-devel] Alsa 1.0

[Christian S. <chr...@ep...>]

Es geschah am Freitag, 5. Dezember 2003 16:06 als Mark Knecht schrieb:
> Any headway being made with the code and this uint problem? I cannot
> build LS anymore.

I will look for it hopefully today, if not at least tomorrow. Takes only 
little changes for Alsa 1.0 I think. I will send you then a patch per mail, 
so you can test it (I dont have Alsa 1.0 yet) and if that works for you, I'll 
commit it.

CU
Christian

Re: [Linuxsampler-devel] Alsa 1.0

[Mark K. <mar...@co...>]

On Sat, 2003-12-06 at 06:08, Christian Schoenebeck wrote:
> Hi Mark!
> 
> As promised I tried to make a quick fix for Alsa 1.0. Instead of creating a 
> patch I just attached you the complete tarball, I guess you prefer that 
> anyway, don't you?
> 
> I've not written an automake check for the alsa version, so this tarball only 
> works for Alsa 1.0 by default. If you want to use it for another Alsa version 
> you have to set the ALSA_1_0 define in audioio.h to 0.
> 
> Please report your result to the list.
> 
> CU
> Christian

Christian,
   How's this for a fast response?! ;-) The tarball compiled fine and is
working well with Alsa-1.0.0.rc2,  gcc-3.3.2, glibc-2.3.2-r8,

Thanks!

Now, I'll renew my request for a boring but very helpful command line
option - MIDI connections. How about it?! :-)  :-)

   And I'm fine with CVS once you commit stuff. 

Cheers,
Mark

Re: [Linuxsampler-devel] Alsa 1.0 and bugfixes commited

[Christian S. <chr...@ep...>]

Es geschah am Samstag, 6. Dezember 2003 16:01 als Mark Knecht schrieb:
> On Sat, 2003-12-06 at 06:08, Christian Schoenebeck wrote:
> > Hi Mark!
> >
> > As promised I tried to make a quick fix for Alsa 1.0. Instead of creating
> > a patch I just attached you the complete tarball, I guess you prefer that
> > anyway, don't you?
> >
>
> Christian,
>    How's this for a fast response?! ;-) The tarball compiled fine and is
> working well with Alsa-1.0.0.rc2,  gcc-3.3.2, glibc-2.3.2-r8,

Fine!

> Now, I'll renew my request for a boring but very helpful command line
> option - MIDI connections. How about it?! :-)  :-)

If time allows it ...   one day ...  :P

But that's not a hard task anyway, perhaps you do it and we'll appreciate if 
you'll send us your patch!

>    And I'm fine with CVS once you commit stuff.

I've commited it a couple of minutes ago with a bunch of other fixes.

CU
Christian