zamaudio / zam-plugins Goto Github PK
View Code? Open in Web Editor NEWCollection of LADSPA/LV2/VST/JACK audio plugins for high-quality processing
Home Page: http://www.zamaudio.com/?p=976
License: GNU General Public License v2.0
Collection of LADSPA/LV2/VST/JACK audio plugins for high-quality processing
Home Page: http://www.zamaudio.com/?p=976
License: GNU General Public License v2.0
All the LV2 plugins shows up as category : "Plugin" where it' be useful if the each one could appear as its real category.
When changing the tonestack in ladspa, or with the lv2 generic controls, in both A3 and 4, I get silence at certain values. (7 and higher).
Then at 17 it works again.
Once the val has been 7, nr 1-6 also stop working. Once I pass 0 again, 1-6 work again.
The lv2 generic gui shows a probable cause: rounding errors in the values. (for example 8.999999).
It could be nice to have a stereo version of the Dynamic Eq with an Average/Maximum switch to use it on the Master Bus for example... I now use a workaround to use the sum of Left+Right channel as Sidechain for the plugin but it is tricky :-D
Thank you, it is the only nice sounding free de-esser / dynamic eq out there!!
Lorenzo
I am in the midst of revamping the look of some of my favorite zam plugins to my liking (nothing against the original artwork but it suffered from low contrast and a cramped feeling), see for example here. Now I noticed some issues, when I saw how the artwork is actually included in the plugins:
So my question is, are there plans to
If you want, I could help you out with one or more of these points.
Hi, the plugin scanner for Radium reports that this plugin crashes:
kjetil@localhost radium]$ bin/radium_plugin_scanner L2hvbWUva2pldGlsL3phbS1wbHVnaW5zL2Jpbi9aYW1EZWxheS12c3Quc28= L3RtcC9kZXNjci50eHQ=
JUCE v5.4.7
Launched: -/home/kjetil/zam-plugins/bin/ZamDelay-vst.so- -/tmp/descr.txt-
JUCE Assertion failure in juce_MessageListener.cpp:38
Attempting to load VST: /home/kjetil/zam-plugins/bin/ZamDelay-vst.so
Creating VST instance: ZamDelay-vst
Initialising VST: ZamDelay-vst (3.12.0.0)
assertion failure: "obj->plugin == nullptr" in file ../../dpf/distrho/src/DistrhoPluginVST.cpp, line 1251
assertion failure: "obj->plugin == nullptr" in file ../../dpf/distrho/src/DistrhoPluginVST.cpp, line 1251
Segmentation fault (core dumped)
To compile i just wrote "make", but "make" crashed too (!):
Generate ttl data for './ZamDelay_dsp.so', basename: 'ZamDelay_dsp'
Writing manifest.ttl... done!
Writing ZamDelay_dsp.ttl.../home/kjetil/zam-plugins/dpf/utils/generate-ttl.sh: line 28: 28913 Segmentation fault (core dumped) "$GEN" "./$FILE"
make: *** [Makefile:23: gen] Error 139
[kjetil@localhost zam-plugins]$
Don't know if this crash is related to the VST plugin crashing...
It would be very handy to have a display of the gain reduction inline in Ardour.
I heard the Ardour compressor is yours as well.
For that one, I've requested the same.
This line must be like this in order to work.
rm -f
In order to make the gate more useful, I'd love to see a "close gain" parameter. This gain factor defines maxium gain reduction in closed state.
The FreeBSD port fails:
for plugin in ZamComp ZamCompX2 ZaMultiComp ZamTube ZamEQ2 ZamAutoSat ZamGEQ31 ZaMultiCompX2 ZamGate ZamGateX2 ZamHeadX2 ZaMaximX2 ZamDelay ZamDynamicEQ ZamPhono ZamVerb ZamGrains; do \
install bin/"$plugin" /usr/ports/audio/zam-plugins-lv2/work/stage/usr/local/bin ; \
done;
install bin/*-ladspa.so /usr/ports/audio/zam-plugins-lv2/work/stage/usr/local/lib/ladspa
install: bin/*-ladspa.so: No such file or directory
gmake[1]: *** [Makefile:38: install] Error 71
(EDITED)
Was: ZamCompX2 LADSPA version from current git won't load in non-mixer. analyseplugin gives the following error messages:
Now is: ZamCompX2 LADSPA version from current git produces an error message from analyseplugin:
"Stereo Detection" input, control, 0 to 1, ERROR: TOGGLED INCOMPATIBLE WITH OTHER HINT, default 0
"Sidechain" input, control, 0 to 1, ERROR: TOGGLED INCOMPATIBLE WITH OTHER HINT, default 0
LV2 (hosted in jalv.gtk) and JACK versions work fine.
EDIT: ZamCompX2 will load in non-mixer if strip has 3 channels. See my comment below.
Preliminary background:
I build all music stuff cross with openembedded/yocto. Since the DPF turtle generation with lv2-ttl-generator does not work in cross builds - plugins are build for ARM - I implemented two ways to get around:
Symptom:
Running ttl-genaration on target machine segfaults for ZamVerb:
root@raspberrypi3:ZamVerb.lv2# lv2-ttl-generator /usr/lib/lv2/ZamVerb.lv2/ZamVerb_dsp.so
Generate ttl data for '/usr/lib/lv2/ZamVerb.lv2/ZamVerb_dsp.so', basename: 'ZamVerb_dsp'
convolution: samplerate mismatch preset:48000 host:44100
Segmentation fault
Debugging the symptom:
Here is a screenshot of the point of crash:
This is all I have currently - if you cannot reproduce and want me to set breakpoints / take further screenshots - let me know
Hi there - attempting to boot the audio server (Server.local) in SuperCollider 3.10.3 built from source on KDE Neon produces aprox. ~8500 errors of the like posted below, all while scanning the ZamTube-ladpsa.so plugin. The result is a segmentation fault. I got the plugin from KXstudio who packaged it as part of the meta-audio-plugins-ladspa collection. When I remove that plugin the server boots correctly. I have posted part of the output below.
...
==26886== 45000 errors in context 237 of 239:
==26886== Use of uninitialised value of size 8
==26886== at 0x65612BB: __exp1_fma (e_exp.c:273)
==26886== by 0x6562718: __ieee754_pow_fma (e_pow.c:115)
==26886== by 0x20FAAAD3: ??? (in /usr/lib/ladspa/ZamTube-ladspa.so)
==26886== by 0x20FAA655: ??? (in /usr/lib/ladspa/ZamTube-ladspa.so)
==26886== by 0x20FAA9BA: ??? (in /usr/lib/ladspa/ZamTube-ladspa.so)
==26886== by 0x20FAC5EB: ??? (in /usr/lib/ladspa/ZamTube-ladspa.so)
==26886== by 0x20FACF63: ??? (in /usr/lib/ladspa/ZamTube-ladspa.so)
==26886== by 0x20FA872F: ??? (in /usr/lib/ladspa/ZamTube-ladspa.so)
==26886== by 0x4010732: call_init (dl-init.c:72)
==26886== by 0x4010732: _dl_init (dl-init.c:119)
==26886== by 0x40151FE: dl_open_worker (dl-open.c:522)
==26886== by 0x6BF72DE: _dl_catch_exception (dl-error-skeleton.c:196)
==26886== by 0x40147C9: _dl_open (dl-open.c:605)
==26886==
==26886==
==26886== 45000 errors in context 238 of 239:
==26886== Use of uninitialised value of size 8
==26886== at 0x65612A0: __exp1_fma (e_exp.c:274)
==26886== by 0x6562718: __ieee754_pow_fma (e_pow.c:115)
==26886== by 0x20FAAAD3: ??? (in /usr/lib/ladspa/ZamTube-ladspa.so)
==26886== by 0x20FAA655: ??? (in /usr/lib/ladspa/ZamTube-ladspa.so)
==26886== by 0x20FAA9BA: ??? (in /usr/lib/ladspa/ZamTube-ladspa.so)
==26886== by 0x20FAC5EB: ??? (in /usr/lib/ladspa/ZamTube-ladspa.so)
==26886== by 0x20FACF63: ??? (in /usr/lib/ladspa/ZamTube-ladspa.so)
==26886== by 0x20FA872F: ??? (in /usr/lib/ladspa/ZamTube-ladspa.so)
==26886== by 0x4010732: call_init (dl-init.c:72)
==26886== by 0x4010732: _dl_init (dl-init.c:119)
==26886== by 0x40151FE: dl_open_worker (dl-open.c:522)
==26886== by 0x6BF72DE: _dl_catch_exception (dl-error-skeleton.c:196)
==26886== by 0x40147C9: _dl_open (dl-open.c:605)
==26886==
==26886==
==26886== 45000 errors in context 239 of 239:
==26886== Conditional jump or move depends on uninitialised value(s)
==26886== at 0x65611F4: __exp1_fma (e_exp.c:252)
==26886== by 0x6562718: __ieee754_pow_fma (e_pow.c:115)
==26886== by 0x20FAAAD3: ??? (in /usr/lib/ladspa/ZamTube-ladspa.so)
==26886== by 0x20FAA655: ??? (in /usr/lib/ladspa/ZamTube-ladspa.so)
==26886== by 0x20FAA9BA: ??? (in /usr/lib/ladspa/ZamTube-ladspa.so)
==26886== by 0x20FAC5EB: ??? (in /usr/lib/ladspa/ZamTube-ladspa.so)
==26886== by 0x20FACF63: ??? (in /usr/lib/ladspa/ZamTube-ladspa.so)
==26886== by 0x20FA872F: ??? (in /usr/lib/ladspa/ZamTube-ladspa.so)
==26886== by 0x4010732: call_init (dl-init.c:72)
==26886== by 0x4010732: _dl_init (dl-init.c:119)
==26886== by 0x40151FE: dl_open_worker (dl-open.c:522)
==26886== by 0x6BF72DE: _dl_catch_exception (dl-error-skeleton.c:196)
==26886== by 0x40147C9: _dl_open (dl-open.c:605)
==26886==
==26886== ERROR SUMMARY: 2062836 errors from 239 contexts (suppressed: 0 from 0)
When adjusting the knobs, sometimes their position jumps suddenly. I don't have a way to reproduce this consistently but it happens regularly and moving the knobs a few times is enough for the issue to show up.
Please Note that you use the tone stacks, not the cabinets.
http://pickroar.com/1003/the-tone-stack-explained-in-english-for-humans/
regards
Hi all,
I've recently compiled the zam audio plugins as they are and all seem to be working fine except the ZaMaximX2, it just doesn't seem to do anything regardless or threshold and ceiling settings; there's no audible or visual gain reduction. Can't see this in the current or closed issues so just wondering if anyone has any ideas?
I'm using the 32bit windows vst build.
Thanks
When testing #45 , I made the following scenario:
...bang, almost freezed.
Edit: note that it doesn't need to be the same SFZ bank loaded for the almost-freezed to happen
Zam plugins build from GIT master 7d2226a
Ardour 5.12
The standalone works fine, the LV2 as well, but the LADSPA version doesn't sound to be working when playing with the 3 controls in Ardour.
The storytelling:
When using the built-in browser to select a SFZ, one can navigate in the folders and choose a SFZ bank. All good here. One can play with this chosen SFZ.
Now, if I want to select another SFZ, I reopen the built-in browser which didn't record where I was formerly and I've got to re-navigate to the folder I was in previously.
What does I expect as a user?
I expect the browser to reopen at the same place I was, making the work-flow easier for testing different SFZ.
Hope that helps
I think all your beautiful plugins could benefit from a "unified" display for numerical values. The knobs and the corresponding scales are fine for most things but sometime I'd love to see the values I actually dialled in.
Take the EQ as an example: sometimes you need to tune a very specific fequency but the button scale is just not enough. The numerical display would always display the the numeric value of the last touched knob / parameter and tell the exact frequency the knob is set at.
As mentioned at DPF it would be benifical to provide DPF system wide, or alternatively include the desired version in the source release (as part of the assets).
This would allow me to ship zam-plugins for Arch Linux
Users should be able to only build the type of plugins that they need.
Additionally, ladspa plugins are obsolete, superseded by LV2.
I've just ugraded from Debian 8 to 9 and I've noticed the CPU load of the JACK versions of the plugins has greatly increased. ZaMultiCompX2 was around 32% CPU on Debian 8, but over 120% on Debian 9 on the same hardware with the same signal and settings. This is with latest git code on Debian 9, and I also tried the same git snapshot (May 2017) that I used on Debian 8 on the Debian 9 system. The problem was the same.
The LADSPA version of ZaMultiCompX2 hosted in non-mixer uses about 12% CPU with the same settings. The JACK version takes a few seconds to show the GUI on startup on Deb9 but it starts almost instantaneously on Deb8. The other plugins show similar symptoms.
System is Athlon IIx4 2.5GHz with RS780 onboard graphics (Radeon HD3200).
Other GUI-based plugins like the LSP ones actually have a slightly lower load on Debian 9 than on 8. So far the problem appears to be specific to zam-plugins.
Here is a step-by-step how-to crash:
Preparations:
\
by /
(see #43).Ardour 5 + ZamSFZ playing fine together:
Making it crashes:
Now, I relaunch Ardour 5 (from the terminal to get some info) and I choose the session saved just before, and tadam ! Segfault:
Butler drops pool trash
Set buffering params to 262144|131072|10|10
Erreur de segmentation
I relaunch Ardour 5, choosing a new session, it opens. I choose to open the saved before, and tadam ! Crash with the same message.
Information:
Screencast: http://download.tuxfamily.org/librazik/volatile/zam-ardour5-test.mp4
Updated Debian Jessie.
Zam build from 5d9e7c6
Hope that can help,
Olivier
PS: it's been confirmed by another user using Ardour 4.7 and the same Zam-plugins build from 5d9e7c6, on a Debian Jessie too.
Just built from 3.8 tag and this is happening in ardour nightly (couple of days old version). Set the Detect Freq, close the gui, open it again and it is back at 1000hz. It seems that even automating this parameter to be at the desired frequency doesn't make it go where it should.
A "link" button would be nice. When enabled changing the value for a parameter should change the value on the same parameter on the other bands as well. Melda's Multiband Compressor has this and it's quite useful.
Discussed on IRC:
Trying to open a sfz from the SSO with today's git ZamSFZ (a0a21f0) in its standalone JACK version started from command-line, it results in a few errors such as:
`Samples/Percussion/conga-slp-rr2.wav'
Unsupported
Missing samples
System error : No such file or directory.
File: /home/capturixeur/Bureau/Samples/Percussion/conga-mut-rr1.wav
N-48 V-0 Samples/Percussion/conga-mut-rr1.wav
A workaround is to open the sfz file and change all the \
in the pathnames to /
.
Thanks to you @zamaudio for giving me this workaround on IRC, really appreciated.
Hope that helps
Cheers
I'm currently putting together a release news aggregator - https://libreav.org/updates - and I was wondering if I can request that the changelog text could be included in the release message texts so that it comes through via the GH releases.atom feed?
When adjusting the make up gain on the ZamMultiComp, the compressor line curve spills out of the box (bottom left)
USE_SYSTEM_LIBS
disappeared, and HAVE_ZITA_CONVOLVER=true
breaks:
gmake[2]: *** No rule to make target '../../build/ZamHeadX2/zita-convolver.cpp.o', needed by '../../bin/ZamHeadX2'. Stop.
There seem to be some issues with the lv2 version of it in Ardour 3.5.380. Ladspa works fine. It seems to be it's gui when threshold parameters are changed Ardour crash.
As described in the title, I get an error when installing the zam plugins on macOS (Mojave 10.14.2), as it doesn't recognize the -t
option provided to the install command.
In ZaMultiCompX2UI.cpp the below function seems to be missing a switch call for the stereo/Max toggle.
This causes the plugin to fail to load preset value on/off of the toggle on reload.
I added the below case which seemed to fix the problem on my machine.
using latest git clone 7/21/15
// DSP Callbacks
//line 275
void ZaMultiCompX2UI::parameterChanged(uint32_t index, float value)
{
switch (index)
{
case ZaMultiCompX2Plugin::paramStereoDet:
fToggleStereo->setValue(value);
break;
}
Thanks for all the great plugins!!
The current git version of DPF, which gets pulled in when cloning zam, doesn't have dgl/src/Common.hpp anymore and the zam compilation fails when it can't find it:
Compiling ZamEQ2UI.cpp
In file included from ZamEQ2UI.hpp:23,
from ZamEQ2UI.cpp:18:
../../widgets/ZamWidgets.hpp:26:10: fatal error: ../dpf/dgl/src/Common.hpp: Fileor folder not found
26 | #include "../dpf/dgl/src/Common.hpp"
Tried on Ubuntu 20.04 with gcc 9.3.0
Another error seemingly related to the new DPF is
In file included from ZaMultiCompPlugin.cpp:18:
ZaMultiCompPlugin.hpp:140:12: error: ‘DISTRHO::String DISTRHO::ZaMultiCompPlugin::getState(const char*) const’ marked ‘override’, but does not override
140 | String getState(const char* key) const override;
It doesn't occur if I check out an older version of DPF.
(edited for clarity)
I installed current version of zam-plugins from git (f75581f - that is v3.5+)
Zamsfz plugin loaded just fine in carla, but it produces no sound, no matter what sfz file i open.
For example i tried this: SalamanderGrandPianoV3_48khz24bit from http://freepats.zenvoid.org/Piano/
Mentioned above sfz file plays just fine in linuxsampler.
I notice your project includes a libsfz
which incorporates modified versions of substantial portions of code from SFZero, e.g. from the SFZReader
class. I have worked very closely with SFZero, and so the code is unmistakeable to me.
I'm not sure if you did this deliberately, but it is a violation of the SFZero's MIT license to remove that license and re-license the modified code under the GPL.
Can I therefore politely ask that you either license libsfz
under the MIT license including the original copyright notice as required here, or remove libsfz
and replace it with a fully-refactored version that doesn't include substantial portions of code from SFZero
, in which case you can license it however you wish.
ZaMultiComp(X2) seems to be missing two threshold controls. This makes it impossible to set up in what I'd regard as a sensible configuration for a multiband compressor. Most full range musical signals have higher levels at lower frequencies, so this in effect makes it a low frequency compressor. Strictly speaking it should also have separate attack and release controls per band as HF often needs faster settings than LF. If I need a multi-band comp on the mix bus I have to use a crossover plus several ZamCompX2 (an excellent compressor BTW) which defeats the object of having a multi-band one :-(
The LADSPA versions of zam-plugins aren't recognised by non-mixer. At first I thought this might be the Plugin Unique IDs being too large - the official IDs as allocated by Richard Furse, who maintains the list of reserved LADSPA IDs, are still below 10000. Looking into it further I see that the IDs are unsigned long int, so I guess that isn't the problem.
ZamComp sidechain control shows as a fader instead of a toggle switch in LV2 (jalv.gtk -g) and LADSPA (in non-mixer) versions in the host's generic GUI. The LV2 GUI version shows the switch correctly in jalv.gtk.
I know these are experimental plugins, and you're not happy with them, but if it's not to much trouble it would be great if you could make at least ZamChild670 compile again.
Tried to compile on a platform with no x11 support with no luck. Look for arguments like --no-lv2-gui --lv2-only but didn't find them...
When loading the lib,
Critical: signal 8 received (OVERFLOW)
(_Z11onSignalFPEi+0x6e)[0x562360070598]
/lib/x86_64-linux-gnu/libc.so.6(+0x3ef20)[0x7f3a5bb2af20]
/usr/lib/ladspa/ZamGEQ31-ladspa.so(+0x3714)[0x7f39c6db3714]
/usr/lib/ladspa/ZamGEQ31-ladspa.so(+0x3c74)[0x7f39c6db3c74]
/usr/lib/ladspa/ZamGEQ31-ladspa.so(+0x3cec)[0x7f39c6db3cec]
/usr/lib/ladspa/ZamGEQ31-ladspa.so(+0x408b)[0x7f39c6db408b]
/usr/lib/ladspa/ZamGEQ31-ladspa.so(+0x54a2)[0x7f39c6db54a2]
/usr/lib/ladspa/ZamGEQ31-ladspa.so(+0x5d50)[0x7f39c6db5d50]
/usr/lib/ladspa/ZamGEQ31-ladspa.so(+0x147d)[0x7f39c6db147d]
/lib64/ld-linux-x86-64.so.2(+0x10733)[0x7f3a60436733]
zam-plugins/bionic,now 3.9~repack3-1 amd64
Collection of LV2, LADSPA, LINUX-VST and JACK plugins
I'm wondering if there is any chance for some older plugins like noise gate or synth to come back in the future?
ZamNoise does not seem to build as ImageToggle.hpp
it uses is missing.
Same for ZamSynth.
Hi,
I've cross-compiled this plugin for a Raspberry Pi 2 and the xruns are quite high. I cannot use this plugin
at the moment. What is the reference CPU consumption on a Intel i7 for this plugin?
As a NOTE: here you can find my attempts in the past to implement a plugin using official rt-wdf library. Unfortunately I fall in the same issue: CPU consumption using this modeling approach is so high that the plugin is not usable on my platform...
I think the gate would be more useful if it had a source filter. The filtered should then be used as analysis signal for the gate. A highpass + a lowpass fillter would be enough.
A button to monitor the filtered signal would be nice as well.
It makes it easier to regex and most projects configure their changelog this way.
forwarded report from http://mixbus.harrisonconsoles.com/forum/thread-3211-post-20992.html
I have an issue using ZamGEQ31 with Jack's sample rate at 44100 htz, if I move up or down
the fader of 20k's band. The sound of the host (Ardour 5.12, Qtractor 0.9.2, Reaper 5.96) is down
until I reset the preset. No issues with sample rate at 48k Htz.
zamplugins 3.10
Distro : LibraZiK2 (audio oriented distro based on Debian Stretch)
A declarative, efficient, and flexible JavaScript library for building user interfaces.
🖖 Vue.js is a progressive, incrementally-adoptable JavaScript framework for building UI on the web.
TypeScript is a superset of JavaScript that compiles to clean JavaScript output.
An Open Source Machine Learning Framework for Everyone
The Web framework for perfectionists with deadlines.
A PHP framework for web artisans
Bring data to life with SVG, Canvas and HTML. 📊📈🎉
JavaScript (JS) is a lightweight interpreted programming language with first-class functions.
Some thing interesting about web. New door for the world.
A server is a program made to process requests and deliver data to clients.
Machine learning is a way of modeling and interpreting data that allows a piece of software to respond intelligently.
Some thing interesting about visualization, use data art
Some thing interesting about game, make everyone happy.
We are working to build community through open source technology. NB: members must have two-factor auth.
Open source projects and samples from Microsoft.
Google ❤️ Open Source for everyone.
Alibaba Open Source for everyone
Data-Driven Documents codes.
China tencent open source team.