Paul Davis wrote:
in general, you should forget about the h/w capabilities of an audio
interface. for every user that has a device with some interesting
qualities, there will be 10 who do not.
welcome to winmodem for audio ...
Well, in this particular case, let's just see what the generic
Jens M Andreasen wrote:
Now that mingo's (et al) RT patches are coming into mainstream, what is
the corporate rationale behind it and the running order of urgency?
I am fishing for some information on; if it is the disk-drives, the
network drivers, the usb stack or something else that I am
Fons Adriaensen wrote:
Core Sound http://www.core-sound.com/default.php willsoon
be offering a tetrahedral (Ambisonic) microphone at a very
reasonable price. They are also working on a combined preamp
+ AD converter unit for this mic. This will be able to multiplex
Sounds nice :)
So here's
Dave Robillard wrote:
Sorry, I /was/ referring to sem_post/sem_wait/etc. I mean havn't tried
OK :)
Benchmarking message queues against pipes would be interesting, maybe
Jack could benefit if they're faster?
I did some benchmarking between unix (local) sockets and message queues
and found
Dave Robillard wrote:
realtime thread is pretty sketchy...). Pipes let you communicate
between processes though - I havn't tried the fancier POSIX interprocess
stuff yet.
What do you mean by semaphore then, if not sem_*()?
sem_post(3)/sem_wait(3) are defined in POSIX realtime extensions
Lee Revell wrote:
But, from the original post it seems that pthread_cond_signal is not
realtime safe as it locks a mutex:
Just about any syscall nowadays potentially acquires some sort of lock
inside kernel.
- Jussi
Paul Davis wrote:
writing to a pipe is not 100% RT safe, but if the pipe is created in a
shm filesystem, its as close to it as you will get without ...
Nowadays, there's also available a very good interface from POSIX RT
extensions; posix message queues. See mq_send(3). You can use either
Paul Davis wrote:
nice to hear that they are faster. on the other hand, once again POSIX
screws us all over by not integrating everything into a single blocking
wait call. i've said it before, i'll say it again - this is one of the
few things that the win32 API gets right - you can block in
Esben Stien wrote:
If 192kHz is to high we can always slow it down. 96 or 48 would
allow more processing power to be used for other things.
That's in not the concern. You see, it is mathematically not necessary
to sample at 192kHz. Only reason for doing 96kHz is because of the
Depends on
Stefan Westerfeld wrote:
process, it should be possible to be quite fast (well, of course it
depends on what fast is) while maintaining good quality. From my recent
work on SSE based resampling code for BEAST:
At least you can make SSE process four float streams in parallel without
. And it also works
for pro cards like my Delta1010. First it was argued that there wasn't
enough control and ALSA was better. Now ALSA has taken this to the other
extreme and now we are arguing if it's too complex.
Truth is probably somewhere between, as usual...
--
Jussi Laako [EMAIL PROTECTED]
On Sat, 2006-02-25 at 22:35 -0500, Dave Robillard wrote:
Interesting fact: there is nothing good about GnomeCanvas.
Nothing.
For any free form rendering I use Cairo or OpenGL nowadays. Better
performance and you get antialiased lines etc practically for free. And
what's most important, you
On Sun, 2006-02-26 at 20:38 +0100, Albert Graef wrote:
Canvases give you much more than just rendering. They also manage the
graphical objects that you created and, if anything changes, rerendering
the changed parts happens automatically.
That's usually bad and undesirable for any real time
On Sat, 2006-02-25 at 12:39 +0100, Carlo Capocasa wrote:
Personally, I'd love to see more GTK2 apps around.
I think it's the oldest *full featured* toolkit around and I think even
the bloat is bearable (just keep your hands off anything that says 'Gnome')
Motif is pretty old and full
is not there so i have added it as :
Depending on how the TSC audio chip is connected, it might not be so
straightforward to have ALSA...
--
Jussi Laako [EMAIL PROTECTED]
On Sun, 2005-11-20 at 12:54 +1100, Dave Robillard wrote:
Out of curiosity, how expensive is this runtime architechture check?
It's done only once at initialization time and even there it's matter of
100 machine instructions.
At runtime the cost is doing integer comparison.
--
Jussi Laako
complex functions. All
allocations are aligned to required boundary. There are C, C++ and C++
template APIs. Btw. Intel's compiler can vectorize most of the remaining
functions and for some even parallelize.
Currently the lib is missing autotools stuff, so it's
makefile-configured...
--
Jussi
()
pthread_mutexattr_setpshared()
Yes, that's nice feature, I'm using that in my own apps. You can place
the objects in shared memory. It works with Linux 2.6 kernels and NPTL.
LinuxThreads doesn't have support for those (functions may exist, but
always return error).
--
Jussi Laako [EMAIL PROTECTED]
.
Filter 1: F = 50 Hz, A = 9
Filter 2: F = 2120 Hz, A = 1
and add the two outputs.
From quality point of view, at least I would recommend using IIR filters
for this...
Unless digital'ish sound is preferred... ;)
--
Jussi Laako [EMAIL PROTECTED]
of capabilities of vinyl hardware...
--
Jussi Laako [EMAIL PROTECTED]
involved.
--
Jussi Laako [EMAIL PROTECTED]
in milliseconds,
etc.
--
Jussi Laako [EMAIL PROTECTED]
really know how this is solved in these cases.
In case of OSS backend, this can be handled by the driver. For multiple
samplerate case this is more complex.
--
Jussi Laako [EMAIL PROTECTED]
On Thu, 2005-09-29 at 23:14 +0400, Dmitry S. Baikov wrote:
Jackd needs buffers to be power of 2, and usb-audio - multiples of 1ms.
I don't think there is any hard limit in jack for buffer size to be 2^x.
(?)
--
Jussi Laako [EMAIL PROTECTED]
On Thu, 2005-09-29 at 23:14 +0400, Dmitry S. Baikov wrote:
Jackd needs buffers to be power of 2, and usb-audio - multiples of 1ms.
I just verified jack to work with any buffer size using OSS backend.
--
Jussi Laako [EMAIL PROTECTED]
(speed and latency -wise)
compared to ATI.
Only thing where ATI is better is when you need to use TV-output. At
least nvidia's PAL S-Video output has been broken for ages.
--
Jussi Laako [EMAIL PROTECTED]
. Some reference clocks can be
programmed to output some specific frequency.
--
Jussi Laako [EMAIL PROTECTED]
done in each driver.
This is something rather easy to do while developing the driver. And
from application point of view, there should be way to know how the gain
relates to actual output level. ie. know what the 0 dB level is.
--
Jussi Laako [EMAIL PROTECTED]
when it's sensitivity vs.
frequency is known. Or to know how to get exactly 1Vpp output level.
With my M-Audio Delta1010 this is rather easy, as there are no
adjustable analog gain stages and it can be calibrated. But trying to do
this generally with any sound card is pretty hard.
--
Jussi Laako
adjustments should be in dB (or dBu or some
other commonly used unit). Thus you would know what the adjustment
really is, no matter what the current hardware supports.
--
Jussi Laako [EMAIL PROTECTED]
On Sat, 2005-09-03 at 19:19 +0100, Dan Mills wrote:
annoying. IMO, all volume adjustments should be in dB (or dBu or some
db, as this is a gain change (which is unitless).
It could also be calibrated, thus with unit.
--
Jussi Laako [EMAIL PROTECTED]
01742
Name servers
Name
IP-address
Status
karhu.koillismaa.fi
213.255.163.2
OK
poro.koillismaa.fi
213.255.163.102
OK
--
Jussi Laako [EMAIL PROTECTED]
than a year
(since 1.0.0) is non-realistic. First of all it has to prove it's API
stability first. And there will be partially/completely binary-only OSS
drivers as ALSA is GPL and as such, doesn't allow writing partially or
completely binary only drivers. This is the current reality.
--
Jussi
variant) and
ALSA 1.x. Those classes have basically the same interface.
I still haven't seen proper tool for doing asound.conf for example
M-Audio Delta1010. And I don't want to support all the n+1 newbies on
how to edit their ALSA configuration to do something meaningful.
--
Jussi Laako [EMAIL
to S/PDIF (the last 2
channels). Took me some time to figure this out as this was not
documented, not at least at that time (2-3 years ago). I still don't
know how to map that S/PDIF output to ALSA's OSS emulation /dev/dsp9
where I have it on native OSS.
--
Jussi Laako [EMAIL PROTECTED]
gaming to music producing.
--
Jussi Laako [EMAIL PROTECTED]
On Thu, 2005-06-09 at 18:14 +0200, stefan kersten wrote:
int access(std::vectorint v, int i)
{
return v[i];
}
At least you are making copy here, should be
int access(std::vectorint v, int i)
--
Jussi Laako [EMAIL PROTECTED]
can easily change it to use
doubles instead of floats just by changing the typedef.
--
Jussi Laako [EMAIL PROTECTED]
it's tasks and reliability. Execution speed is
secondary. You can buy more CPU power if required. Most large array
beamformers are heavy SMP systems anyway.
In C/C++ case the performance is more about how you write the code, not
the language you use.
--
Jussi Laako [EMAIL PROTECTED]
On Wed, 2005-06-08 at 22:09 +0300, Jussi Laako wrote:
And you can still access the individual samples by using vData[n]
without significant performance penalty compared to a simple float
array. And I say significant here just because it also performs bounds
checking. It could be made even
are
architecture specifics anyway and always have a pure C++ implementations
also.
--
Jussi Laako [EMAIL PROTECTED]
of the actual code in 'real' classes.
I'm usually limiting inheritance in my implementations to maximum of two
or three levels and absolutely no multiple inheritance.
--
Jussi Laako [EMAIL PROTECTED]
like GPU for audio.
At least I have good experiences on using FPGAs for signal processing.
Cheaper than DSP and more flexible and powerful. However, much more
difficult to program.
--
Jussi Laako [EMAIL PROTECTED]
can read data at rather high
rates).
--
Jussi Laako [EMAIL PROTECTED]
with delta-sigma conversion is worse DIM distortion
performance, this is not something SACD-specific.
--
Jussi Laako [EMAIL PROTECTED]
with
older equipment. That is yet-to-be-seen on DVD-A (well, I have actually
heard something about some dual layer DVD-A).
I'm rather happy with my CD/SACD/DVD-A player anyway. No matter what
kind of discs I feed.
--
Jussi Laako [EMAIL PROTECTED]
/images/product_diagrams/4398blkdiag_mag.gif
and Direct DSD path. Which one of the paths, the Direct DSD or the
PCM you think is easier and cheaper to make good quality?
--
Jussi Laako [EMAIL PROTECTED]
.
That is stupid (IMHO). I thought that they were recommending use of
192/24 PCM for processing. And DSD only at final stage for producing the
actual bitstream for disc.
--
Jussi Laako [EMAIL PROTECTED]
quality 5.1 or 2.0 audio from SACD discs. And those
also play on standard CD players. I personally would like to see SACD
winning over DVD-A. Because of different coding, it also simplifies D/A
converters and thus improves sound quality.
--
Jussi Laako [EMAIL PROTECTED]
added.
- Some makefile cleanups and fixes.
Homepage:
http://libdsp.sf.net
http://www.sonarnerd.net/projects/libdsp/
Download:
http://sourceforge.net/project/showfiles.php?
group_id=25287package_id=17119release_id=305295
http://www.sonarnerd.net/projects/dlbins/
--
Jussi Laako (MiskaX || SonarNerd
is different thing, you are mostly spending just time. I've
still put a few thousand euros of pure money to my software projects.
It's a different thing is someone is willing to sponsor such a project.
--
Jussi Laako [EMAIL PROTECTED]
the goal...
Are we talking in the same context here?
* mil = 1/1000 of inch
--
Jussi Laako [EMAIL PROTECTED]
On Sun, 2004-11-07 at 03:11, Jussi Laako wrote:
Current implementation is available from:
http://www.sonarnerd.net/linux/libjackmm-071104.tar.gz
Uh oh, forgot to include Exception.hh from libDSP sources. I just
updated the package...
--
Jussi Laako [EMAIL PROTECTED]
://www.sonarnerd.net/projects/libdsp/
--
Jussi Laako [EMAIL PROTECTED]
. A basic JACK
application made using this library works OK anyway.
Current implementation is available from:
http://www.sonarnerd.net/linux/libjackmm-071104.tar.gz
--
Jussi Laako [EMAIL PROTECTED]
). To get best possible sound quality out of
these, you'll need high quality samplerate conversion when playing 44.1
kHz files/streams.
All comments on this are welcome.
Patch is available from:
http://www.sonarnerd.net/linux/xmms-rabbit.patch
--
Jussi Laako [EMAIL PROTECTED]
. At
least for P4 and AMD64. I usually do some profiling on code generated by
the compiler and then handcode the SSE2 parts for compiler bottlenecks.
IIR filter was one good example where compilers sucked badly.
--
Jussi Laako [EMAIL PROTECTED]
/websgram.jpg
So basically it could be OK with 3D display?
--
Jussi Laako [EMAIL PROTECTED]
(Fedora Core 1, gcc-3.3.3):
[EMAIL PROTECTED] tests]# ./bench --verify irf1024
Segmentation fault
[EMAIL PROTECTED] tests]# ./bench irf1024
Segmentation fault
--
Jussi Laako [EMAIL PROTECTED]
publish the test results when I get SSE2 optimization for Ooura
code done.
--
Jussi Laako [EMAIL PROTECTED]
On Sat, 2004-02-28 at 15:43, Tim Goetze wrote:
is the benchmark code you use available somewhere?
Older version with some parts removed is in libDSP CVS.
--
Jussi Laako [EMAIL PROTECTED]
=i686 platforms and includes 3DNow
/ SSE stuff. It just crashed on my Thunderbird-core Athlon.
I use *_execute_dft_r2c().
Also SIMD optimizations doesn't do any good on non-x86 architectures
like SPARC, Alpha or HP-PA.
Anyway, I'm going to add some SIMD stuff to Ooura also.
--
Jussi Laako
implementation of
floatdouble version of Ooura radix-4, radix-8 and split-radix
transforms.
Using Ooura (built-in), FFTW (external) and Intel IPP (external) FFTs
with libDSP is just matter of configuration #define.
--
Jussi Laako [EMAIL PROTECTED]
MS/s
30420 us / 65536 point real FFT (single), 33 FFTs/s, 2.154 MS/s
--
Jussi Laako [EMAIL PROTECTED]
On Sun, 2004-02-01 at 12:02, Daniel Risacher wrote:
aRts, etc. Wouldn't it be nice if all the legacy apps just worked?
Without blocking each other?
This has already been implemented in commercial version of OSS. With
samplerate conversion and mixing.
--
Jussi Laako [EMAIL PROTECTED]
On Mon, 2003-12-22 at 11:55, Uwe Koloska wrote:
but this will build only the float variant. And the fftw3 build
procedure is not able to build both variants ...
I've done RPMs of FFTW 3.x which include float, double and long double
versions. See http://www.sonarnerd.net/linux/
--
Jussi
On Fri, 2003-12-19 at 22:12, Taybin Rutkin wrote:
JACK 0.92.0 has been released.
As usual, modified version with OSS driver is available as RPMs from
http://www.sonarnerd.net/linux/
--
Jussi Laako [EMAIL PROTECTED]
time ./resamp_fixp
1.037u 0.000s 0:01.03 100.0%0+0k 0+0io 76pf+0w
jussi/own time ./resamp_float
1.970u 0.000s 0:01.96 100.5%0+0k 0+0io 76pf+0w
jussi/own time ./resamp_float_fistl
4.034u 0.001s 0:04.03 100.0%0+0k 0+0io 90pf+0w
--
Jussi Laako [EMAIL PROTECTED]
.
--
Jussi Laako [EMAIL PROTECTED]
actual projects.
I don't have have any control on jack, so I'm not going to tie myself up
on something which needs unspecified amount of work.
--
Jussi Laako [EMAIL PROTECTED]
users out there... :)
--
Jussi Laako [EMAIL PROTECTED]
= paUInt8, paInt16,
Maybe this is a feature of alsa oss emulation?
--
Jussi Laako [EMAIL PROTECTED]
/automake could see the
drivers/oss/Makefile.am for some better way of enabling NPTL/barrier
stuff... ;)
--
Jussi Laako [EMAIL PROTECTED]
is available at
http://www.sonarnerd.net/linux/ and SuSE 9 packages are coming.
--
Jussi Laako [EMAIL PROTECTED]
with the PortAudio driver,
instead?
No, depending on how PortAudio is implemented...
--
Jussi Laako [EMAIL PROTECTED]
Of course I forgot something out of the patch. Here's additional patch
for drivers/oss/Makefile.am...
--
Jussi Laako [EMAIL PROTECTED]
--- jack-audio-connection-kit-0.80.0/drivers/oss/Makefile.am 1970-01-01 02:00:00.0 +0200
+++ jackit/drivers/oss/Makefile.am 2003-11-14 21:04
OK, some new code.. Attached are some of my experiments.
- Code that uses split registers to process two samples per loop.
- Code for processing interleaved stereo (equally fast to mono version)
TODO:
- Implement resample code in libDSP in e3dnow asm
--
Jussi Laako [EMAIL PROTECTED
On Fri, 2003-10-10 at 23:40, Paul Davis wrote:
PortAudio works on most of them, and is much better for many reasons.
PortAudio uses OSS? Or are there many drivers for M-Audio Delta1010 and
the likes?
--
Jussi Laako [EMAIL PROTECTED]
On Fri, 2003-10-17 at 02:09, Juan Linietsky wrote:
Benno Sennoner and I were discussing today on IRC about
the usual fixed point vs floating point (regarding to some resampling code)
Attached is some quick-and-dirty hack written tired and drunk at friday
night...
--
Jussi Laako [EMAIL
lacks the ./install - [OK] - [OK] - [Exit and
save] - soundon -installation.
SuSE is doing OK on the installation area, but their biggest problem is
hacking the ALSA into their kernel, making it pretty difficult to change
the kernel while keeping ALSA stuff in their distro working.
--
Jussi Laako
to be running 24/7/365 no matter how
many xruns there are, without needing to restart the interface every
time. Minimal dataloss required when xrun happens.
--
Jussi Laako [EMAIL PROTECTED]
On Thu, 2003-10-09 at 17:44, Paul Davis wrote:
OSS is dead. You should not be writing apps with OSS. If you want
portability use PortAudio.
It's not dead if you are doing audio work on any other IEEE-1003.1 OS
than Linux.
--
Jussi Laako [EMAIL PROTECTED]
On Fri, 2003-10-10 at 19:28, Paul Davis wrote:
you need to additively mix your two signals. writing them to DSP
orders them sequentially in time. it doesn't mix them.
The commercial OSS driver does mixing and samplerate conversion.
--
Jussi Laako [EMAIL PROTECTED]
continue to support OSS. And I still recommend it for
users because it's easy to install. Just do ./oss-install and then
just OK-OK-ExitSave and that's it.
--
Jussi Laako [EMAIL PROTECTED]
platform API. ALSA is Linux only.
--
Jussi Laako [EMAIL PROTECTED]
some extra overhead and code to audio
chain.
Are there any other drivers to get any decent PCI soundcard to work on
those platforms?
--
Jussi Laako [EMAIL PROTECTED]
and it is
highly recommended to use alsa at all.
ALSA has it's own drivers and it's own native interface. It's able to
emulate OSS at driver level.
--
Jussi Laako [EMAIL PROTECTED]
to include it in test results.
--
Jussi Laako [EMAIL PROTECTED]
on AMD ones. However, using SSE2 for fp math on P4 fixes this.
--
Jussi Laako [EMAIL PROTECTED]
On Thu, 2003-07-31 at 11:45, Alfons Adriaensen wrote:
Do you know of a *very fast* (probably inline assembly) way to force
denormal FP numbers to zero ?
If I remember correctly, SSE does flush-denormals-to-zero by default.
--
Jussi Laako [EMAIL PROTECTED]
should be defined volatile.
2) compiler doesn't know anything about processes or IPC, so same as
above. All inter-process variables should be defined volatile.
3) on SMP systems each CPU still has it's own dcache
--
Jussi Laako [EMAIL PROTECTED]
) bytes
--
Jussi Laako [EMAIL PROTECTED]
and second for the data, thus the header stuff
is protected from the data.
See also pthread PSHARED flag, which seems to be supported in NPTL
(included in RH9) but not supported under old LinuxThreads. Works also
under Solaris 8/9.
--
Jussi Laako [EMAIL PROTECTED]
explanation.
I haven't got any problems with shared memory, but afterall, I don't use
the SystemV method (shm* functions) because it sucks. I think the
SystemV IPC stuff sucks as whole. I like to use the real POSIX way..
:)
--
Jussi Laako [EMAIL PROTECTED]
.
No problems here with 2.4.19-jl series... Steady at 0.9 ms when using 128
byte buffers with ENS1371.
- Jussi Laako
--
PGP key fingerprint: 161D 6FED 6A92 39E2 EB5B 39DD A4DE 63EB C216 1E4B
Available at PGP keyservers
make the graphs available.
I think software raid driver needs some work to lower it's latency. Current
-ll patch doesn't include anything for it, if I remember correctly.
- Jussi Laako
--
PGP key fingerprint: 161D 6FED 6A92 39E2 EB5B 39DD A4DE 63EB C216 1E4B
Available at PGP keyservers
[EMAIL PROTECTED] wrote:
Does anyone knows a compact and portable DSP library (preferably open
source). I need it to run on ARM or MIPS target boards withe embedded
linux for voice compressions and pther telecom functions.
http://libdsp.sf.net ?
- Jussi Laako
--
PGP key
of
the sound (and ever more with quicker melodies). But maybe it's just because
I'm too used to the real thing.
- Jussi Laako
--
PGP key fingerprint: 161D 6FED 6A92 39E2 EB5B 39DD A4DE 63EB C216 1E4B
Available at PGP keyservers
.
Original motivation for the patch set was to create kernel that is able to
run my signal analysis software properly under full load. It does heavy DSP
operations with 8-32 channels of audio and also heavy network and graphics
load.
- Jussi Laako
--
PGP key fingerprint: 161D 6FED 6A92
.
- Jussi Laako
--
PGP key fingerprint: 161D 6FED 6A92 39E2 EB5B 39DD A4DE 63EB C216 1E4B
Available at PGP keyservers
1 - 100 of 153 matches
Mail list logo