[ANN] Ecasound 2.7.0 released

[Kai Vehmanen]

0. Special note: Ten years of Ecasound!
---------------------------------------

This is the 10th year anniversary release for Ecasound! The first
public version, 1.1.4r1, of Ecasound, licensed under GPL, was released
on 30th June 1999. See the following mailing list post for a quick
rundown of what has happened so far:

- http://eca.cx/ecasound-list/2009/07/0004.html

1. Summary of changes in this release
-------------------------------------

Initial Open Sound Control (OSC) interface for parameter control has
been added. New '-chorder' and '-eadb' options, and 'cop-get'
interactive mode command, have been added. Optional build time support
has been added for using liboil to optimize inner loops, giving a
small performance boost to many common use-scenarios. Various bugs
fixed in JACK support, mp3 output and option parsing. Fixes to build
problems in Mac OS X.

2. What is Ecasound?
--------------------

Ecasound is a software package designed for multitrack audio processing.
It can be used for simple tasks like audio playback, recording and format
conversions, as well as for multitrack effect processing, mixing,
recording and signal recycling. Ecasound supports a wide range of audio
inputs, outputs and effect algorithms. Effects and audio objects can be
combined in various ways, and their parameters can be controlled by
operator objects like oscillators and MIDI-CCs. A versatile console mode
user-interface is included in the package.

Primary platform for running Ecasound is GNU/Linux. Ecasound can also be
run on many UNIX-derived systems such as FreeBSD, Mac OS X and Solaris.
Limited support for Windows is available through Cygwin. Ecasound is
licensed under the GPL. The Ecasound Control Interface (ECI) is licensed
under the LGPL.

3. Changes in 2.7.x series
--------------------------

v2.7.0:

* Open Sound Control (OSC) support. See the initial announcement
   mail sent to ecasound-list:
   http://eca.cx/ecasound-list/2009/04/0036.html
   Current interface is documented at:
   http://eca.cx/ecasound/Documentation/ecasound_osc_interface.txt
   The interface is still limited and subject to change in
   later releases, but it's a start.

* New '-chorder' operator that allows to reorder channels of
   an audio stream. Also duplication and omission of certain
   channels is possible. See ecasound(1) man page for more
   information.

* Added new amplify/gain variant '-eadb' that allows to specify
   the gain in dB. See the related mail thread:
   http://eca.cx/ecasound-list/2009/03/0034.html

* Refactored POSIX signal handling in ecasound. See the following
   mail for some rationale, as well as a list of changes.
   See mail thread:
   http://eca.cx/ecasound-list/2009/02/0027.html

* Various optimizations to Ecasound inner loops using
   the liboil library. See http://liboil.freedesktop.org/wiki/
   To enable the optimizations, liboil-0.3 development files
   need to be installed and '--enable-liboil' must be passed
   to Ecasound's configure script.

* New 'cop-get' interactive mode command. See the updated
   ecasound-iam(1) manual page for further info.

Full list of changes is available at:
  - http://www.eca.cx/ecasound/history.php

4. Interface and configuration file changes in 2.7 series
---------------------------------------------------------

v2.7.0:

Output of '-ev' operator has been renewed.

The name for default chainsetup created from command line is now
"untitled-chainsetup".

Most of the entries in the installed ecasoundrc file
(in ${prefix}/share/ecasound/ecasoundrc), are now commented
out by default.

Major changes to the libecasound library public interface.
This should not really affect anyone anymore, as direct use of
libecasound has been discouraged for a long time and it is
available only as a static library, but just in case someone
is still using it. See libecasound/ChangeLog for a detailed
list of changes.

5. Contributors to 2.7 series
-----------------------------

Patches - Accepted code, documentation and build system changes
~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~

Extracted with 'git-shortlog -s':

v2.7.0:
- Adam Linson (1)
- Jeremy Hughes (1)
- Junichi Uekawa (1)
- Kai Vehmanen (203)

Bug Hunting - Reports that led to bugfixes (items closed)
~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~

v2.7.0:
   * Oliver Oli (2)
       various bugs in new OSC support
   * RProgrammer @ sf.net (1)
       uninstall target broken on OS X (sfbug:1283448)
   * Jason Galyon (1)
       frontend parser bug for '-E' option

Feature suggestions - Ideas that led to new features (items)
~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~

v2.7.0:
   * Julien Claassen (1)
       OSC support
   * Klaus Schulz (1)
       -eadb chainop

6. Links and files
------------------

Web site (and mirrors):
  http://eca.cx/ecasound (fi)
  http://ecasound.seul.org (us)
  http://ecasound.sourceforge.net (us)

Source package:
  http://eca.cx/ecasound/download.php
  ecasound-2.7.0.tar.gz, md5sum:
  0311307fa4fb4f085178843b3cec477a

List of distributions with maintained Ecasound support:
  See http://eca.cx/ecasound/download.php

--

------------------------------------------------------------------------------
Come build with us! The BlackBerry(R) Developer Conference in SF, CA
is the only developer event you need to attend this year. Jumpstart your
developing skills, take BlackBerry mobile applications to market and stay
ahead of the curve. Join us from November 9 - 12, 2009. Register now!
http://p.sf.net/sfu/devconference
_______________________________________________
Ecasound-list mailing list
Ecasound-list@...
https://lists.sourceforge.net/lists/listinfo/ecasound-list

[Pierre Lorenzon-2]

Hi,

I installed 2.7.0 release and have following problem I had not
with 2.6.0.

Sound is played to fast and saturated as well ! When issuing

$ ecasound -i:file.wav

output is the same with 2.7.0 and 2.6.0 with same warnings
etc. But with 2.6.0 sound is correct and with 2.7.0 not !

OSC support is not compiled in 2.7.0 : could the problem come
from that ?

All librairies used by ecasound are uptodate !

Best regards

Pierre



------------------------------------------------------------------------------
Come build with us! The BlackBerry(R) Developer Conference in SF, CA
is the only developer event you need to attend this year. Jumpstart your
developing skills, take BlackBerry mobile applications to market and stay
ahead of the curve. Join us from November 9 - 12, 2009. Register now!
http://p.sf.net/sfu/devconference
_______________________________________________
Ecasound-list mailing list
Ecasound-list@...
https://lists.sourceforge.net/lists/listinfo/ecasound-list

[Kai Vehmanen]

Hi,

On Sun, 18 Oct 2009, Pierre Lorenzon wrote:

> I installed 2.7.0 release and have following problem I had not
> with 2.6.0.
>
> Sound is played to fast and saturated as well ! When issuing

I've not been able to reproduce this, but sounds fairly serious.

Has anyone else seen this?

> $ ecasound -i:file.wav

Does the same happen with:

ecasound -i:file.wav -o:alsa,default

?

The source of the problem could be the change in global ecasoundrc file
(defaults set in the binary versus the conf file). In theory 2.7.0 could
be picking a different default on your setup, and that would lead to bad
audio. Sounds a bit farfetched, but for now I couldn't think of any other
explanation.

> OSC support is not compiled in 2.7.0 : could the problem come
> from that ?

How about liboil support, did you enable it ('ldd /usr/bin/ecasound'
should list liboil if support is enabled)? By default it's disabled.

------------------------------------------------------------------------------
Come build with us! The BlackBerry(R) Developer Conference in SF, CA
is the only developer event you need to attend this year. Jumpstart your
developing skills, take BlackBerry mobile applications to market and stay
ahead of the curve. Join us from November 9 - 12, 2009. Register now!
http://p.sf.net/sfu/devconference
_______________________________________________
Ecasound-list mailing list
Ecasound-list@...
https://lists.sourceforge.net/lists/listinfo/ecasound-list

[Pierre Lorenzon-2]

From: Kai Vehmanen <kvehmanen@...>
Subject: Re: [ecasound] Problem with 2.7.0
Date: Sun, 18 Oct 2009 13:20:47 +0300 (EEST)

> Hi,
>
> thanks for the quick reply!

  No problem I have time to test it now !

>
> On Sun, 18 Oct 2009, Pierre Lorenzon wrote:
>
>>> Does the same happen with:
>>>
>>> ecasound -i:file.wav -o:alsa,default
>>
>>  Unfortunately yes !
>
> Hmm, could you also try:
>
>  ecasound -i:file.wav -o:alsa,default -B:rt

   Same problem. Anyway even without this option there's a line
   in the output saying that rt buffuring mode is selected.


>  ecasound -i:file.wav -o:alsa,default -B:nonrt

   Same problem

>  ecasound -i:file.wav -o:alsa,default -b:1024 -z:nodb

   Same problem

   More precisely, the sound is not really saturated but there
   are little "clicks" at a frequency of approximately
   10hz. Moreover I have following warning with each of the
   options above :

>>>  -- Warning

(audioio_alsa) Warning! Period-size differs from current client
buffersize.

>>>  -- End Warning


But I have it as well with 2.6.0 and sound is corrrect.

>
> ... does any of these cure the problem?
>
>>  I installed 2.7.0 in /opt i.e.
>>  $ ./configure --prefix=/opt
> [...]
>> resource-directory = /usr/share/ecasound
>
> Hmm, again this is odd. I've tested the same thing and for
> me resource-directory is correctly set to the towards
> /opt. It's defined as:
>
> resource-directory = @pkgdatadir@
>
> ... which points in Makefile to:
>
> pkgdatadir = $(datadir)/ecasound
> datadir = ${datarootdir}
> datarootdir = ${prefix}/share
>
> ... so it definitely should match prefix.

  Surely it should ! Maybe the problem was that I did not clean
  the source directory before restarting ./configure etc. I'll
  try with untaring archive again and start with a clean
  source. Yes in this case ressource directory is correct but
  sound unfortunately still not !

>
>> Anyway I don't think that the problem comes from that. Indeed
>> when I first installed 2.7.0 in the root i.e.
>> $ ./configure --prefix=/usr --sysconfdir=/etc \
>> --localstatedir=/var
>> I encounter the problem as well. Moreover when I set manualy
>> the output to be alsa,default, (with the -o option,) problem
>> doesn't disappear !
>
> Agreed, I think this is something else. Thanks for checking
> these.
>
>>> How about liboil support, did you enable it ('ldd
>>> /usr/bin/ecasound' should list liboil if support is enabled)?
>>> By default it's disabled.
>>
>>  Yes ! liboil is not present on my system. It was not needed
>>  by 2.6.0. Should I install it ? When I run configure for
>>  2.7.0 I remember that I saw that liboil support was
>>  disable. Anyway here are the outputs for ldd :
>
> No, it's not needed. Just trying to divide the problem
> space. I've now tested with liboil and without liboil, and
> doesn't seem to have any effect at least on my system.

  Regards

  Pierre



------------------------------------------------------------------------------
Come build with us! The BlackBerry(R) Developer Conference in SF, CA
is the only developer event you need to attend this year. Jumpstart your
developing skills, take BlackBerry mobile applications to market and stay
ahead of the curve. Join us from November 9 - 12, 2009. Register now!
http://p.sf.net/sfu/devconference
_______________________________________________
Ecasound-list mailing list
Ecasound-list@...
https://lists.sourceforge.net/lists/listinfo/ecasound-list

[Kai Vehmanen]

Hi,

>   Same problem. Anyway even without this option there's a line
>   in the output saying that rt buffuring mode is selected.

hmm, ok, I have one hypothesis. Here's one snapshot, let me know if it
helps:

http://ecasound.seul.org/download/snapshots/ecasound-2.7.1+dev-20091018.tar.gz

Also, I'd be interested to know which version of alsa-lib (libasound) do
you have, and which kernel version?

------------------------------------------------------------------------------
Come build with us! The BlackBerry(R) Developer Conference in SF, CA
is the only developer event you need to attend this year. Jumpstart your
developing skills, take BlackBerry mobile applications to market and stay
ahead of the curve. Join us from November 9 - 12, 2009. Register now!
http://p.sf.net/sfu/devconference
_______________________________________________
Ecasound-list mailing list
Ecasound-list@...
https://lists.sourceforge.net/lists/listinfo/ecasound-list

[Pierre Lorenzon-2]

Hi,
From: Kai Vehmanen <kvehmanen@...>
Subject: Re: [ecasound] Problem with 2.7.0
Date: Sun, 18 Oct 2009 20:27:35 +0300 (EEST)

> Hi,
>
>>   Same problem. Anyway even without this option there's a line
>>   in the output saying that rt buffuring mode is selected.
>
> hmm, ok, I have one hypothesis. Here's one snapshot, let me
> know if it helps:


  Yes it does ! The problem no longer exists with this version.
>
> http://ecasound.seul.org/download/snapshots/ecasound-2.7.1+dev-20091018.tar.gz
>
> Also, I'd be interested to know which version of alsa-lib


  1.0.13
> (libasound) do you have, and which kernel version?

  2.6.22.5

  Regards

Pierre



------------------------------------------------------------------------------
Come build with us! The BlackBerry(R) Developer Conference in SF, CA
is the only developer event you need to attend this year. Jumpstart your
developing skills, take BlackBerry mobile applications to market and stay
ahead of the curve. Join us from November 9 - 12, 2009. Register now!
http://p.sf.net/sfu/devconference
_______________________________________________
Ecasound-list mailing list
Ecasound-list@...
https://lists.sourceforge.net/lists/listinfo/ecasound-list

[Kai Vehmanen]

Hi,

On Mon, 19 Oct 2009, Pierre Lorenzon wrote:

>> hmm, ok, I have one hypothesis. Here's one snapshot, let me
>> know if it helps:
>
>  Yes it does ! The problem no longer exists with this version.

ok, great, that is really good news. I now wrote a longer mail about this
to ALSA list:
http://permalink.gmane.org/gmane.linux.alsa.devel/67752

The good news is that this bug should not hit too many people. Most ALSA
PCMs support power-of-two period-sizes (-b:1024, -b:128, ...), and in
these cases no glitches should not occur.

In your specific case, I think the PCM doesn't support -b:1024 but
something close, as the block/period size. Now due to the bug this results
in discontinuity at every block, leading to glitches at a rate of ~20Hz
when playing 44.1kHz audio. This matches pretty closely what you
described.

>> Also, I'd be interested to know which version of alsa-lib
>  1.0.13
>> (libasound) do you have, and which kernel version?
>  2.6.22.5

These match as well. These still have the old implementation of
snd_pcm_sw_params_set_xfer_align both in alsa-lib and in kernel.

Thanks for helping to track this down! I'll release 2.7.1 soonish with
this fix (and possibly a few others as well) included.

------------------------------------------------------------------------------
Come build with us! The BlackBerry(R) Developer Conference in SF, CA
is the only developer event you need to attend this year. Jumpstart your
developing skills, take BlackBerry mobile applications to market and stay
ahead of the curve. Join us from November 9 - 12, 2009. Register now!
http://p.sf.net/sfu/devconference
_______________________________________________
Ecasound-list mailing list
Ecasound-list@...
https://lists.sourceforge.net/lists/listinfo/ecasound-list

[Pierre Lorenzon-2]

Hi,

Thanks first for having fixed this problem and for this precise
and convincing explanation !

Regards

Pierre

From: Kai Vehmanen <kvehmanen@...>
Subject: Re: [ecasound] Problem with 2.7.0
Date: Tue, 20 Oct 2009 00:47:11 +0300 (EEST)

> Hi,
>
> On Mon, 19 Oct 2009, Pierre Lorenzon wrote:
>
>>> hmm, ok, I have one hypothesis. Here's one snapshot, let me
>>> know if it helps:
>>
>>  Yes it does ! The problem no longer exists with this version.
>
> ok, great, that is really good news. I now wrote a longer mail
> about this to ALSA list:
> http://permalink.gmane.org/gmane.linux.alsa.devel/67752
>
> The good news is that this bug should not hit too many
> people. Most ALSA PCMs support power-of-two period-sizes
> (-b:1024, -b:128, ...), and in these cases no glitches should
> not occur.
>
> In your specific case, I think the PCM doesn't support -b:1024
> but something close, as the block/period size. Now due to the
> bug this results in discontinuity at every block, leading to
> glitches at a rate of ~20Hz when playing 44.1kHz audio. This
> matches pretty closely what you described.
>
>>> Also, I'd be interested to know which version of alsa-lib
>>  1.0.13
>>> (libasound) do you have, and which kernel version?
>>  2.6.22.5
>
> These match as well. These still have the old implementation of
> snd_pcm_sw_params_set_xfer_align both in alsa-lib and in
> kernel.
>
> Thanks for helping to track this down! I'll release 2.7.1
> soonish with this fix (and possibly a few others as well)
> included.

------------------------------------------------------------------------------
Come build with us! The BlackBerry(R) Developer Conference in SF, CA
is the only developer event you need to attend this year. Jumpstart your
developing skills, take BlackBerry mobile applications to market and stay
ahead of the curve. Join us from November 9 - 12, 2009. Register now!
http://p.sf.net/sfu/devconference
_______________________________________________
Ecasound-list mailing list
Ecasound-list@...
https://lists.sourceforge.net/lists/listinfo/ecasound-list