yoshimi records

[Cal-8]

... maybe. I've gotten Record to the point where it hasn't done anything violent
to me for a while, but it still has plenty of potential for unexpected
shortcomings. The story so far ...

- new Audio/Midi settings - default recordings directory and a checkbox for 32 bit
float wavs. The default is 16 bit shorts. The wavs are courtesy of libsndfile, so
that's a new dependency.
- Select your wav file via the 'Wav file' button. Opening an existing wav starts
in append mode.
- Orange rewind button does a rewind/overwrite.
- Green starts recording.
- Blue pauses recording.
Yes, that's a weird choice of colours. Send me some hex colour numbers and I'll
gladly change them.

To be honest, I'm struggling to hold focus well enough and long enough to give it
the sort of testing it needs. There's more than a few different usage paths and
despite best intentions, serious disciplined testing just isn't happening. So I'm
asking for some help with testing.

Background ... if record is running, at the point where audio is generated it
gets fed off to a fifo. The buffer on that is currently set to 256k. For no
particular reason I can think of I felt more comfortable with it at 512k. I've no
idea how low that can go before the mechanism chokes and splutters. Yet another
thread reads from the fifo and writes it to wav file. Does it properly manage the
integrity of wav files in all possible use cases? I think so, but I'm not at all
sure. How busy can the system get before audio and/or recording quality fails?
Don't know.

When you find it broken, the best I can promise is that I will try to fix it :-).
<http://www.graggrag.com/yoshimi/yoshimi-0.037.tar.bz2>

cheers folks.
Cal
_______________________________________________
Linux-audio-user mailing list
Linux-audio-user@...
http://lists.linuxaudio.org/mailman/listinfo/linux-audio-user

[Peter Geirnaert]

Hi Cal

On AVLinux I had to install cmake-curses-gui , libmxml-dev  and
libmxml1 , then it compiled and installed fine.

[100%] Built target yoshimi
peter@FreqHost:/mnt/Data/src/yoshimi-0.037/src$ sudo make install
[100%] Built target yoshimi
Install the project...
-- Install configuration: "Release"
-- Installing: /usr/local/bin/yoshimi
peter@FreqHost:/mnt/Data/src/yoshimi-0.037/src$ yoshimi
gzopen failed, errno 2, No such file or directory
Error, xml file /home/peter/.zynaddsubfxXML.cfg could not be loaded or uncompressed
BUG!: XMLwrapper::peek() - empty parentstack
BUG!: XMLwrapper::peek() - empty parentstack
Error, failed to open bank directory ./testbnk
Processors online 3, fftw3 to use 3 threads
No such directory, root bank entry: /usr/local/share/zynaddsubfx/banks
No such directory, root bank entry: /usr/share/zynaddsubfx/banks
No such directory, root bank entry: ~/banks
ADsynth Oscilsize: 2048
Sample Rate: 44100
Sound Buffer Size: 32
Internal latency: 0.725624 ms

Zyn is installed, but zynaddsubfx isn't, so I install zynaddsubfx and try again ...
peter@FreqHost:/mnt/Data/src/yoshimi-0.037/src$ yoshimi
Error, failed to open bank directory ./testbnk
Processors online 3, fftw3 to use 3 threads
No such directory, root bank entry: /usr/local/share/zynaddsubfx/banks
No such directory, root bank entry: ~/banks
ADsynth Oscilsize: 2048
Sample Rate: 44100
Sound Buffer Size: 32
Internal latency: 0.725624 ms
Jack reports error: jack_client_thread zombified - exiting from JACK

The last line comes after sending one midi note, that I can't hear, the output meters show green, and the audio connections in qjackctl are made, but no sound. Also when trying to exit, yoshimi becomes unresponsive.

So I run make clean and then ccmake . again to enable debugging,:
peter@FreqHost:/mnt/Data/src/yoshimi-0.037/src$ cmake .
-- Building Yoshimi version 0.037 for Linux with gdb debug information
-- Found pkg-config /usr/bin/pkg-config
-- Found libz
-- Will build gui
-- Found fontconfig 2.6.0
-- Found fltk
-- Found fftw3 3.1.2
-- Found mxml 2.5
-- Found Alsa 1.0.20
-- Found jack 0.116.2
-- Found libsndfile 1.0.18
-- Default audio driver is Jack
-- Default midi driver is Jack
-- Configuring done
-- Generating done
-- Build files have been written to: /mnt/Data/src/yoshimi-0.037/src

When running make, this time it doesn't build fine:
<snip>
[ 81%] Building CXX object CMakeFiles/yoshimi.dir/MusicIO/WavRecord.cpp.o
[ 82%] Building CXX object CMakeFiles/yoshimi.dir/main.cpp.o
In file included from /mnt/Data/src/yoshimi-0.037/src/main.cpp:315:
/usr/lib/gcc/i486-linux-gnu/4.3.3/include/xmmintrin.h:35:3: error: #error "SSE instruction set not enabled"
/mnt/Data/src/yoshimi-0.037/src/main.cpp: In function ‘int set_DAZ_and_FTZ(int)’:
/mnt/Data/src/yoshimi-0.037/src/main.cpp:336: error: ‘_mm_getcsr’ was not declared in this scope
/mnt/Data/src/yoshimi-0.037/src/main.cpp:336: error: ‘_mm_setcsr’ was not declared in this scope
/mnt/Data/src/yoshimi-0.037/src/main.cpp:357: error: ‘_mm_getcsr’ was not declared in this scope
/mnt/Data/src/yoshimi-0.037/src/main.cpp:357: error: ‘_mm_setcsr’ was not declared in this scope
make[2]: *** [CMakeFiles/yoshimi.dir/main.cpp.o] Error 1
make[1]: *** [CMakeFiles/yoshimi.dir/all] Error 2
make: *** [all] Error 2


PS: Maybe interesting, I checked again in ccmake . > toggle 't' for advanced mode, and there are some X11 related (and one OpenGL_xmesa) PATH's NOT FOUND .

PPS: I like the colors.

On Wed, Nov 11, 2009 at 11:50 AM, cal <cal@...> wrote:

... maybe. I've gotten Record to the point where it hasn't done anything violent
to me for a while, but it still has plenty of potential for unexpected
shortcomings. The story so far ...

- new Audio/Midi settings - default recordings directory and a checkbox for 32 bit
float wavs. The default is 16 bit shorts. The wavs are courtesy of libsndfile, so
that's a new dependency.
- Select your wav file via the 'Wav file' button. Opening an existing wav starts
in append mode.
- Orange rewind button does a rewind/overwrite.
- Green starts recording.
- Blue pauses recording.
Yes, that's a weird choice of colours. Send me some hex colour numbers and I'll
gladly change them.

To be honest, I'm struggling to hold focus well enough and long enough to give it
the sort of testing it needs. There's more than a few different usage paths and
despite best intentions, serious disciplined testing just isn't happening. So I'm
asking for some help with testing.

Background ... if record is running, at the point where audio is generated it
gets fed off to a fifo. The buffer on that is currently set to 256k. For no
particular reason I can think of I felt more comfortable with it at 512k. I've no
idea how low that can go before the mechanism chokes and splutters. Yet another
thread reads from the fifo and writes it to wav file. Does it properly manage the
integrity of wav files in all possible use cases? I think so, but I'm not at all
sure. How busy can the system get before audio and/or recording quality fails?
Don't know.

When you find it broken, the best I can promise is that I will try to fix it :-).
<http://www.graggrag.com/yoshimi/yoshimi-0.037.tar.bz2>

cheers folks.
Cal
_______________________________________________
Linux-audio-user mailing list
Linux-audio-user@...
http://lists.linuxaudio.org/mailman/listinfo/linux-audio-user

_______________________________________________
Linux-audio-user mailing list
Linux-audio-user@...
http://lists.linuxaudio.org/mailman/listinfo/linux-audio-user

[Folderol]

On Wed, 11 Nov 2009 21:50:13 +1100
cal <cal@...> wrote:

> ... maybe. I've gotten Record to the point where it hasn't done anything violent
> to me for a while, but it still has plenty of potential for unexpected
> shortcomings. The story so far ...

<snip>

Just had a quick look at this.
Recording 'In Search Of The Lost Tribes' works perfectly, and shows
virtually no increase in processor usage to just running it.

Oddity
Sometimes when making second or third recordings there is a short snatch
of sound at the start. Is this perhaps a buffer not flushed?

Problem
We seem to have the 'can't load parameters' back again, but this time I
can't work out what has caused it - same symptoms.

--
Will J Godfrey
http://www.musically.me.uk
Say you have a poem and I have a tune.
Exchange them and we can both have a poem, a tune, and a song.
_______________________________________________
Linux-audio-user mailing list
Linux-audio-user@...
http://lists.linuxaudio.org/mailman/listinfo/linux-audio-user

[Cal-8]

Folderol wrote:

> On Wed, 11 Nov 2009 21:50:13 +1100
> cal <cal@...> wrote:
>
>> ... maybe. I've gotten Record to the point where it hasn't done anything violent
>> to me for a while, but it still has plenty of potential for unexpected
>> shortcomings. The story so far ...
>
> <snip>
>
> Just had a quick look at this.
> Recording 'In Search Of The Lost Tribes' works perfectly, and shows
> virtually no increase in processor usage to just running it.
>
> Oddity
> Sometimes when making second or third recordings there is a short snatch
> of sound at the start. Is this perhaps a buffer not flushed?

Possibly, but also possibly related to the the threaded fftw3 usage (which
I'm about to eliminate).

> Problem
> We seem to have the 'can't load parameters' back again, but this time I
> can't work out what has caused it - same symptoms.
Most strange. I thought I did actually test that when I added your fix to
the 'Record' branch.

Sorry for the inconvenience, leave it with me for a bit.

cheers.
_______________________________________________
Linux-audio-user mailing list
Linux-audio-user@...
http://lists.linuxaudio.org/mailman/listinfo/linux-audio-user

[Cal-8]

Folderol wrote:
> [ ... ]
> Oddity
> Sometimes when making second or third recordings there is a short snatch
> of sound at the start. Is this perhaps a buffer not flushed?
I've learned a little more about plumbing, and hopefully got it flushing
appropriately.

> Problem
> We seem to have the 'can't load parameters' back again, but this time I
> can't work out what has caused it - same symptoms.
That was ugly, and quite different to the previous anomaly. Hopefully better
now.

<http://www.graggrag.com/yoshimi/yoshimi-0.038.tar.bz2>
.  
cheers!
_______________________________________________
Linux-audio-user mailing list
Linux-audio-user@...
http://lists.linuxaudio.org/mailman/listinfo/linux-audio-user

[Cal-8]

You've had quite a little tragedy unfolding there, and I felt guilty for ignoring
it. Hopefully things have improved, but ...

Peter Geirnaert wrote:
> [ ... ]
> peter@FreqHost:/mnt/Data/src/yoshimi-0.037/src$ yoshimi
> gzopen failed, errno 2, No such file or directory
> Error, xml file /home/peter/.zynaddsubfxXML.cfg could not be loaded or
> uncompressed
mutter, mutter, that's not nice. One day I should fix that.
[ ... ]

> ADsynth Oscilsize: 2048
> Sample Rate: 44100
> Sound Buffer Size: 32
Are you really running jack with period size 32? Wow!
[ ... ]

> [ 81%] Building CXX object CMakeFiles/yoshimi.dir/MusicIO/WavRecord.cpp.o
> [ 82%] Building CXX object CMakeFiles/yoshimi.dir/main.cpp.o
> In file included from /mnt/Data/src/yoshimi-0.037/src/main.cpp:315:
> /usr/lib/gcc/i486-linux-gnu/4.3.3/include/xmmintrin.h:35:3: error:
> #error "SSE instruction set not enabled"
What cpu is this?
[ ...]

> PPS: I like the colors.
Really? Maybe ... I'm starting to not notice them.

cheers.

_______________________________________________
Linux-audio-user mailing list
Linux-audio-user@...
http://lists.linuxaudio.org/mailman/listinfo/linux-audio-user

[Peter Geirnaert]

Some parts of this message have been removed. Learn more about Nabble's security policy.

On Thu, Nov 12, 2009 at 1:24 PM, cal <cal@...> wrote:

You've had quite a little tragedy unfolding there, and I felt guilty for ignoring
it. Hopefully things have improved, but ...

No problem, I just compiled your stuff to test it, hoping to learn something. 

> ADsynth Oscilsize: 2048
> Sample Rate: 44100
> Sound Buffer Size: 32

Are you really running jack with period size 32? Wow!

Yes, with 3 periods, but I didn't use it for recording a large piece yet. It's a setting I changed one day, but forgot about.
I'll try again after increasing the period size ...
Here's a screenshot of qjackctl's settings now.
 

> [ 81%] Building CXX object CMakeFiles/yoshimi.dir/MusicIO/WavRecord.cpp.o
> [ 82%] Building CXX object CMakeFiles/yoshimi.dir/main.cpp.o
> In file included from /mnt/Data/src/yoshimi-0.037/src/main.cpp:315:
> /usr/lib/gcc/i486-linux-gnu/4.3.3/include/xmmintrin.h:35:3: error:
> #error "SSE instruction set not enabled"

What cpu is this?
[ ...]

AMD Phenom X3:
 

peter@FreqHost:~$ lscpu
Architecture:          i686
CPU(s):                3
Thread(s) per core:    1
Core(s) per socket:    3
CPU socket(s):         1
Vendor ID:             AuthenticAMD
CPU family:            16
Model:                 2
Stepping:              3
CPU MHz:               1050.000
Virtualization:        AMD-V
L1d cache:             64K
L1i cache:             64K
L2 cache:              512K
L3 cache:              2048K


peter@FreqHost:~$ cpufreq-info
cpufrequtils 005: cpufreq-info (C) Dominik Brodowski 2004-2006
Report errors and bugs to cpufreq@..., please.
analyzing CPU 0:
  driver: powernow-k8
  CPUs which need to switch frequency at the same time: 0
  hardware limits: 1.05 GHz - 2.10 GHz
  available frequency steps: 2.10 GHz, 1.05 GHz
  available cpufreq governors: userspace, conservative, powersave, ondemand, performance
  current policy: frequency should be within 1.05 GHz and 2.10 GHz.
                  The governor "powersave" may decide which speed to use
                  within this range.
  current CPU frequency is 1.05 GHz.
  cpufreq stats: 2.10 GHz:13.89%, 1.05 GHz:86.11%  (251)
analyzing CPU 1:
  driver: powernow-k8
  CPUs which need to switch frequency at the same time: 1
  hardware limits: 1.05 GHz - 2.10 GHz
  available frequency steps: 2.10 GHz, 1.05 GHz
  available cpufreq governors: userspace, conservative, powersave, ondemand, performance
  current policy: frequency should be within 1.05 GHz and 2.10 GHz.
                  The governor "ondemand" may decide which speed to use
                  within this range.
  current CPU frequency is 1.05 GHz.
  cpufreq stats: 2.10 GHz:1.67%, 1.05 GHz:98.33%  (6665)

At startup, QjackCtl also doesn't report finding the SSE feature (as it does with UbuntuStudio on this box) and it reports unlimited  memory locking:

11:01:56.676 Patchbay deactivated.

11:01:56.682 Statistics reset.

11:01:56.808 Startup script...

11:01:56.809 artsshell -q terminate

11:01:56.816 ALSA connection graph change.

11:01:57.966 Startup script terminated with exit status=256.

11:01:57.966 JACK is starting...

11:01:57.967 /usr/bin/jackd -R -p128 -u -dalsa -r44100 -p32 -n3 -D -Chw:M1010 -Phw:M1010 -Xseq -H -M

11:01:57.968 JACK was started with PID=3721.

Memory locking is unlimited - this is dangerous. Please alter the line @audio - memlock unlimitedin your /etc/limits.conf to @audio - memlock 2335929

no message buffer overruns

jackd 0.116.2

Copyright 2001-2005 Paul Davis and others.

jackd comes with ABSOLUTELY NO WARRANTY

This is free software, and you are welcome to redistribute it

under certain conditions; see the file COPYING for details

JACK compiled with System V SHM support.

loading driver ..

apparent rate = 44100

creating alsa driver ... hw:M1010|hw:M1010|32|3|44100|0|0|hwmon|hwmeter|-|32bit

configuring for 44100Hz, period = 32 frames (0.7 ms), buffer = 3 periods

ALSA: final selected sample format for capture: 32bit integer little-endian

ALSA: use 3 periods for capture

ALSA: final selected sample format for playback: 32bit integer little-endian

ALSA: use 3 periods for playback

There actually is no /etc/limits.conf here.

Maybe I'll have to make one, or is AVLinux using a different setup?

_______________________________________________
Linux-audio-user mailing list
Linux-audio-user@...
http://lists.linuxaudio.org/mailman/listinfo/linux-audio-user

[Cal-8]

Peter Geirnaert wrote:
> [ ... ]
>     it. Hopefully things have improved, but ...
>
> No problem, I just compiled your stuff to test it, hoping to learn
> something.
That's cool. We can all learn from it.

>      > ADsynth Oscilsize: 2048
>      > Sample Rate: 44100
>      > Sound Buffer Size: 32
>     Are you really running jack with period size 32? Wow!
> Yes, with 3 periods, but I didn't use it for recording a large piece
> yet. It's a setting I changed one day, but forgot about.
> I'll try again after increasing the period size ...
Mostly I'm just jealous I think. If the hardware's up to it then it should just fly,
it certainly shouldn't just quietly implode/explode.

[ ... ]

> At startup, QjackCtl also doesn't report finding the SSE feature (as it
> does with UbuntuStudio on this box) and it reports unlimited  memory
> locking:
Most curious. For yoshi, the compile problem comes from the use of this <http://lists.linuxaudio.org/pipermail/linux-audio-dev/2009-August/024707.html>.
So far, that's served well, but if it can't cope with your Phenom X3 then
there's some research & homework required :-(.

[ ... ]

> Memory locking is unlimited - this is dangerous. Please alter the line
> @audio - memlock unlimitedin your /etc/limits.conf to @audio - memlock
> 2335929
[ ... ]
> There actually is no /etc/limits.conf here.
> Maybe I'll have to make one, or is AVLinux using a different setup?

/etc/security/limits.conf?

cheers.
_______________________________________________
Linux-audio-user mailing list
Linux-audio-user@...
http://lists.linuxaudio.org/mailman/listinfo/linux-audio-user

[Folderol]

On Thu, 12 Nov 2009 19:19:51 +1100
cal <cal@...> wrote:

> Folderol wrote:
> > [ ... ]
> > Oddity
> > Sometimes when making second or third recordings there is a short snatch
> > of sound at the start. Is this perhaps a buffer not flushed?
> I've learned a little more about plumbing, and hopefully got it flushing
> appropriately.
>
> > Problem
> > We seem to have the 'can't load parameters' back again, but this time I
> > can't work out what has caused it - same symptoms.
> That was ugly, and quite different to the previous anomaly. Hopefully better
> now.
>
> <http://www.graggrag.com/yoshimi/yoshimi-0.038.tar.bz2>
> .  

More good work. Thank you very much.

Records with *very* little processor overhead.

I've also noticed that I can now change patches while playing!

On the slow attack, simpler voices the changeover is effectively
seamless, with no xruns. More complex and 'harder' voices sometimes
produce a faint click. Only the most intensive ones actually produce an
xrun, and even then it is just 1 xrun. Changing voices when not playing
results in no xruns at all.

Also, Yoshi starts and stops recording without any xruns. No phantom
sound at the start of a recording anymore either. How difficult would
it be to make it only start recording when the sound starts (like the
old Zyn. one did)? With a lot of my work this would completely eliminate
the need for any post editing.

This is all on my dual core Athlon. Haven't tried on my eeePC yet. I'll
prolly give that ago tomorrow.

--
Will J Godfrey
http://www.musically.me.uk
Say you have a poem and I have a tune.
Exchange them and we can both have a poem, a tune, and a song.
_______________________________________________
Linux-audio-user mailing list
Linux-audio-user@...
http://lists.linuxaudio.org/mailman/listinfo/linux-audio-user

[Peter Geirnaert]

cal wrote:

Peter Geirnaert wrote:

[ ... ]

> At startup, QjackCtl also doesn't report finding the SSE feature (as it
> does with UbuntuStudio on this box) and it reports unlimited  memory
> locking:

Most curious. For yoshi, the compile problem comes from the use of this <http://lists.linuxaudio.org/pipermail/linux-audio-dev/2009-August/024707.html>.
So far, that's served well, but if it can't cope with your Phenom X3 then
there's some research & homework required :-(.

If it helps, I've just started a little page for this homework, about the hardware of my computer, here.
I've also put the output of dmesg there for download.
 

> Memory locking is unlimited - this is dangerous. Please alter the line
> @audio - memlock unlimitedin your /etc/limits.conf to @audio - memlock
> 2335929

[ ... ]

> There actually is no /etc/limits.conf here.
> Maybe I'll have to make one, or is AVLinux using a different setup?

/etc/security/limits.conf?

LOL, yes indeed, there it is.
But I checked the groups for "audio" (using remastersys control panel) and there's no such group.

First breakfast, then more homework ...

Oh and before I forget, building yoshi 0.038 still gave this output at the end:

[ 82%] Building CXX object CMakeFiles/yoshimi.dir/main.cpp.o
In file included from /mnt/Data/src/yoshimi-0.038/src/main.cpp:315:
/usr/lib/gcc/i486-linux-gnu/4.3.3/include/xmmintrin.h:35:3: error: #error "SSE instruction set not enabled"
/mnt/Data/src/yoshimi-0.038/src/main.cpp: In function ‘int set_DAZ_and_FTZ(int)’:
/mnt/Data/src/yoshimi-0.038/src/main.cpp:336: error: ‘_mm_getcsr’ was not declared in this scope
/mnt/Data/src/yoshimi-0.038/src/main.cpp:336: error: ‘_mm_setcsr’ was not declared in this scope
/mnt/Data/src/yoshimi-0.038/src/main.cpp:357: error: ‘_mm_getcsr’ was not declared in this scope
/mnt/Data/src/yoshimi-0.038/src/main.cpp:357: error: ‘_mm_setcsr’ was not declared in this scope
make[2]: *** [CMakeFiles/yoshimi.dir/main.cpp.o] Error 1
make[1]: *** [CMakeFiles/yoshimi.dir/all] Error 2
make: *** [all] Error 2

Have a nice day ...
Peter

_______________________________________________
Linux-audio-user mailing list
Linux-audio-user@...
http://lists.linuxaudio.org/mailman/listinfo/linux-audio-user

[Cal-8]

Peter Geirnaert wrote:
> [ ... ]
> If it helps, I've just started a little page for this homework, about
> the hardware of my computer, here
> <http://sites.google.com/site/experimentalfrush/pc-hardware>.
> I've also put the output of dmesg there for download
> <http://sites.google.com/site/experimentalfrush/pc-hardware/dmesg.txt?attredirects=0&d=1>.
Cool.
[ ... ]
> Oh and before I forget, building yoshi 0.038 still gave this output at
> the end:
> [ 82%] Building CXX object CMakeFiles/yoshimi.dir/main.cpp.o
> In file included from /mnt/Data/src/yoshimi-0.038/src/main.cpp:315:
> /usr/lib/gcc/i486-linux-gnu/4.3.3/include/xmmintrin.h:35:3: error:
> #error "SSE instruction set not enabled"
I don't have an answer for that one yet. If you're comfortable with commenting out a couple
of lines of code, I could show you what to do so that it should at least build. All you'd
be losing is a level of anti-denormals treatment.

cheers.
 
_______________________________________________
Linux-audio-user mailing list
Linux-audio-user@...
http://lists.linuxaudio.org/mailman/listinfo/linux-audio-user

[Peter Geirnaert]

cal wrote:

> #error "SSE instruction set not enabled"

I don't have an answer for that one yet. If you're comfortable with commenting out a couple
of lines of code, I could show you what to do so that it should at least build. All you'd
be losing is a level of anti-denormals treatment.

Sounds interesting, I'm looking forward to your instructions .
I have a basic C++ knowledge so it shouldn't be too hard I guess.

Plus, that way see what can go wrong without the extra level of anti-denormals treatment,
for my audio-synthesis understanding...

Thanks in advance ;-)



_______________________________________________
Linux-audio-user mailing list
Linux-audio-user@...
http://lists.linuxaudio.org/mailman/listinfo/linux-audio-user

[Cal-8]

Peter Geirnaert wrote:
>[ ... ]
> /usr/lib/gcc/i486-linux-gnu/4.3.3/include/xmmintrin.h:35:3: error:
> #error "SSE instruction set not enabled"

A google ... <http://polywogsys.livejournal.com/2007/04/01/>

Try adding -msse to the compiler flags. I can't remember whether you're running
32 or 64 bit. If you're 64 bit, set X86_64_BUILD ON and add -msse to X86_64_OPTS.
If 32 bit, set X86_64_BUILD OFF and add -msse to X86_OPTS. And good luck.

_______________________________________________
Linux-audio-user mailing list
Linux-audio-user@...
http://lists.linuxaudio.org/mailman/listinfo/linux-audio-user

[Peter Geirnaert]

On Fri, Nov 13, 2009 at 11:08 AM, cal <cal@...> wrote:

Peter Geirnaert wrote:
>[ ... ]

> /usr/lib/gcc/i486-linux-gnu/4.3.3/include/xmmintrin.h:35:3: error:
> #error "SSE instruction set not enabled"

A google ... <http://polywogsys.livejournal.com/2007/04/01/>

Try adding -msse to the compiler flags. I can't remember whether you're running
32 or 64 bit. If you're 64 bit, set X86_64_BUILD ON and add -msse to X86_64_OPTS.
If 32 bit, set X86_64_BUILD OFF and add -msse to X86_OPTS. And good luck.

 

I'm using AVLinux, and that's 32 bit only, or I am wrong, anyway, here it's i686 (32 bit).
(maybe I have the wrong gcc installed, for i486 instead of for i686, or is it the same?)

X86_OPTS already has -msse. ???

This is what it shows after ccmake . :
[...]
 X86_64_BUILD                     OFF                                                                                                     
 X86_64_OPTS                      -O3 -march=athlon64 -m64 -Wall -ffast-math -fno-finite-math-only -fomit-frame-pointer                   
 X86_OPTS                         -O3 -march=native -msse -msse2 -mfpmath=sse -ffast-math -fomit-frame-pointer

Anything else I could try ?

_______________________________________________
Linux-audio-user mailing list
Linux-audio-user@...
http://lists.linuxaudio.org/mailman/listinfo/linux-audio-user

[Cal-8]

Folderol wrote:
> [ ... ]
> How difficult would
> it be to make it only start recording when the sound starts (like the
> old Zyn. one did)? With a lot of my work this would completely eliminate
> the need for any post editing.
My initial reaction was "Yeah, I'll have one of them too!", but the noises
from the netbook crowd became a significant distraction. A productive one
though, I do believe. It led to some exploration of rwlocks.

Lightly tested, but promising. Expect anything.

0.039 record autostart added, xruns disciplined
<http://www.graggrag.com/yoshimi/yoshimi-0.039.tar.bz2>

cheers, Cal
_______________________________________________
Linux-audio-user mailing list
Linux-audio-user@...
http://lists.linuxaudio.org/mailman/listinfo/linux-audio-user

[david-602]

cal wrote:

> Folderol wrote:
>> [ ... ]
>> How difficult would
>> it be to make it only start recording when the sound starts (like the
>> old Zyn. one did)? With a lot of my work this would completely eliminate
>> the need for any post editing.
> My initial reaction was "Yeah, I'll have one of them too!", but the noises
> from the netbook crowd became a significant distraction. A productive one
> though, I do believe. It led to some exploration of rwlocks.
>
> Lightly tested, but promising. Expect anything.
>
> 0.039 record autostart added, xruns disciplined
> <http://www.graggrag.com/yoshimi/yoshimi-0.039.tar.bz2>

Sounds like fun!

Totally OT, but the subject line sounds like a good name for a record label.

--
David
gnome@...
authenticity, honesty, community
_______________________________________________
Linux-audio-user mailing list
Linux-audio-user@...
http://lists.linuxaudio.org/mailman/listinfo/linux-audio-user