SpectMorph 1.0.0-beta1 has been released.
The main changes are:
- WavSource: added formant preserving algorithms for repitching vocals
- KeyTrack: new operator to combine instruments by note range
- Envelope: new operator to generate arbitrary modulation envelopes
- LFO: support arbitrary user defined curves
There is a tutorial on YouTube for the new features:
- https://www.youtube.com/embed/mwVUsuOTcN0
Feedback for any issues you might experience with the beta version is
appreciated.
What is SpectMorph?
-------------------
SpectMorph is a free software project which allows to analyze samples of
musical instruments, and to combine them (morphing). It can be used to
construct hybrid sounds, for instance a sound between a trumpet and a
flute; or smooth transitions, for instance a sound that starts as a
trumpet and then gradually changes to a flute.
SpectMorph ships with many ready-to-use instruments which can be
combined using morphing.
SpectMorph is implemented in C++ and licensed under the GNU LGPL version
2.1 or later
Integrating SpectMorph into your Work
-------------------------------------
SpectMorph is currently available for Linux, Windows and macOS (Intel
and Apple Silicon), with CLAP/LV2/VST plugins. Under Linux, there is
also JACK Support.
Links:
------
Website: https://www.spectmorph.org
Download: https://www.spectmorph.org/downloads
There are many audio demos on the website, which demonstrate morphing
between instruments.
List of Changes in SpectMorph 1.0.0-beta1:
------------------------------------------
#### New Features
* WavSource: support two formant preserving algorithms for repitching
vocals.
* New operators:
- KeyTrack - can be used to combine instruments by note range.
- Envelope - to generate arbitary modulation envelopes.
* User defined curve editor for KeyTrack, Envelope and LFO.
* Add presets using the new operators:
- KeyTrack Choir Ah/Oh
- StringBrass Fusion
* JACK CC values are now interpreted per channel (allows polyphonic
modulation).
* Support double click to reset properties to default value.
#### Fixes
* Apply midi channel bend to new voices.
* Make portamento affect filter cutoff key tracking.
* Fix problems with WavSource paths stored by LV2 plugin (make Ardour
archives work).
* Fix getting the initial click events / focus plugin UI on macOS under
Bitwig.
* Fix invalid access to past last vector element (#24).
* Make PandaResampler work without problems in ASAN builds.
* Refactor leak debugger to fix possible crashes caused by global
destructors.
* Do not ship sample data in "1 Instrument WavSource" preset.
#### Internals: Improvements
* Generate and store spectral envelopes in each AudioBlock for formant
correction.
* Improve the strategy for portamento synthesis (more accurate and less
CPU usage).
* Propagate portamento frequency to sources.
* Randomize start phase per default to improve "Harmonic Resynthesis"
quality.
* JACK: use two audio outputs to be able to implement stereo support
later on.
* Add fast vectorizable log2 approximation (based on Tims version from
Anklang).
* Phases in LiveDecoder, IFFTSynth: change from float to uint for
performance reasons.
* Various minor performance optimizations.
* Minor LV2 meta data updates (port groups, avoid port resizing extension).
* Use newer compilers: g++-14 for windows, g++-13 for static linux plugins.
* Support automatically downloading instruments from configure.
* Avoid using std::mutex in hard RT code.
#### Internals: Minor Changes
* Refactoring:
- Deduplicate morphing code from MorphGrid and MorphLinear.
- Use shared pointers for GenericIn/GenericOut classes (RAII).
- Get rid of `sm_clamp` / `sm_bound` (-> std::clamp).
* Improve CI support (use debug-cxx, asan, ubsan, test static build,
avoid fftw planning).
* Add post install test which tests audio output for all presets after
installation.
* Validate LV2 feature arrays.
* Support user defined scripts in testmidisynth test.
* Improve error messages for preset loader.
* Build cleanups: remove libbse code, simplify Qt build (remove
smsampleedit).
* Update CLAP version to 1.2.1.
--
Stefan Westerfeld, http://space.twc.de/~stefan
The FFADO project (https://ffado.org) announces the availability of FFADO
version 2.4.9. This is a maintenance release which contains minor
improvements.
This is a source-only release that can be downloaded from
https://ffado.org/files/libffado-2.4.9.tgz
Changes since FFADO 2.4.8:
* Support compilation against musl libc. Thanks to Alyssa Ross.
* Remove build-time checks for utilities that are not used (pyuic).
* Install the ffado-mixer appdata file only if ffado-mixer has been built.
Thanks to Jan Tojnar for the suggestion.
* Add a DATADIR option to scons and use it in line with GNU conventions.
Thanks to Jan Tojnar for the suggestion and the patch.
* Update the AppStream document with more complete information and in line
with current best practice (Jan Tojnar).
* Rename XDG files to match modern standards (Jan Tojnar).
* Install XDG files manually to simplify the build script and avoid
depending on xdg-utils (Jan Tojnar).
Thanks to those who have helped with this release, including Alyssa Ross
and Jan Tojnar.
Notes for packagers:
* A consequence of the DATADIR support is that the default MANDIR has
shifted from `${PREFIX}/man/` to `${PREFIX}/share/man/`. The latter is
arguably more common, but if the original path must remain in use it is
still possible to manually set MANDIR at build time.
* The presence of the DATADIR option may allow simplification of build
commands or require minor build option adjustments depending on how
FFADO is built.
* Aside from the manpage location, the introduction of DATADIR is not
expected to change the path of any other installed files by default.
Jonathan Woithe
(on behalf of ffado.org)
Hi everyone,
Just a friendly reminder that the deadline for submitting a paper to the 2024 International Faust Conference (IFC-24) is July 12th, 2024 ;), see https://ifc24.soundmit.com/en
Best,
Stéphane
On Sun, Jun 16, 2024 at 11:05:44AM +0200, Jörn Nettingsmeier wrote:
> Some experiences I've had with Belgian beers were more akin to memory
> barriers... stupid me was enjoying Duvel without parsing the metadata (the
> taste is so light and summer-y), matching my friends' cadence (who were
> drinking pilsener). One of the few evenings where I don't remember how I got
> home.
Duvel, at 8.5%, isn't even the strongest, some go up to 12%.
It played a role in my life as well... My first ever job (long ago)
was as a sound engineer for Belgian radio and TV. One evening we were
recording Schubert's piano trio op.99 at the famous Flagey studio 4.
<https://en.wikipedia.org/wiki/Flagey_Building#Recording_studios>
The producer, who also was a music teacher, had engaged one his
students to turn the pages for the piano player. After the recording
he asked me if I could bring her home, as that was on my way back to
Antwerp. So we ended up in the small village of Breendonk which has
no claims to fame except for being where the Duvel is brewed.
We had some Duvels at the small cafe opposite the brewery and ended
up waggling to her parents' home. A few days later we saw each other
again, and became 'an item' that lasted for almost ten years.
Cheers,
--
FA
Hi everyone,
This is just to let you know that after a couple of years of absence, Linux Audio Conference will take place in Lyon in 2025 on June 26-28. For those of you who speak French, it will be preceded by the Journée de l'Informatique Musicale (JIM) on June 23-25. The call for contributions will be posted on the conference website by the beginning of September: https://jimlac25.inria.fr/
In the meantime, save the date ;).
Cheers,
Stéphane (for the jimlac25 committee)
Robin Gareus:
> Subject: [LAD] Re: PandaResampler 0.2.0
> To: linux-audio-dev(a)lists.linuxaudio.org
> Message-ID: <04da06b8-18f1-4216-a878-c5c8530d75f9(a)gareus.org>
> Content-Type: multipart/signed; micalg=pgp-sha256;
> protocol="application/pgp-signature";
> boundary="------------aDq7SRXL2hBrSytcfa1CO5Lw"
>
> On 2024-06-10 03:16, Yuri wrote:
> > On 6/9/24 15:47, Marc Lavallée wrote:
> >>
> > Why do people still insist on GNU Tools?
> >
>
> mostly to aid cross compilation. It is still the option that sucks least
> for libraries, notably to specify which symbols to expose in a cross
> platform library. Though meson is catching up.
>
Yes, auto tools give the least amount of problems for cross
compilation. (cmake is usually a nightmare for cross compilation.) I
would also add that auto tools give the least amount of problems in
general, it's extremely mature, but I haven't used meson very much.
PandaResampler 0.2.0 has been released.
This is a header only library for C++ which implements fast factor 2, 4
or 8 upsampling and downsampling based on SSE instructions. I've
developed the code for my DSP code in Anklang and SpectMorph.
It might be useful for you if you have some DSP loop which needs to be
oversampled to avoid aliasing.
https://github.com/swesterfeld/pandaresampler
--
Stefan Westerfeld, http://space.twc.de/~stefan
