first of all, thanks for the great software you provide. I already made a small donation and am planning to make some more, but there’s a showstopper to me in Ardour, which is the to total unreliability of its plugins. I mean, a DAW is supposed to make use of plugins, and Ardour’s “official” plugins are LADSPA. To me, no plugins, no DAW. But then my experience with LADSPA has been one of love and hate. I learned to cope with the poor interface, but I had too many other issues.
TAP plugins are unstable and make some tracks go crazy
Multiband EQ is unstable
Calf Compressor sometimes go crazy and clips for a second
. Triple Band Parametric EQ sometimes makes some tracks go crazy
and so many other issues… After my experience with the TAP plugins and the Multiband EQ, I decided to use ONLY the recommended plugins on the plugins page on this site, absolutely nothing else. But still, sometimes the Triple Band Parametric EQ make a track clip at 500db for a couple of minutes, the TAP limiter can’t be summoned more than three times or it will blow my speakers away, and overall I get the feeling I’m working with something I can’t totally rely on. I know, you are not developing plugins, but, to be honest, I think you should. Or make a list of the ones that never fail. Because Ardour is great and never crashes, but when I add or adjust a plugin, I always have to pray, and I don’t like it.
I wonder whether it would be worthwhile to provide along with Ardour a small set of stable plugins – just the basics, like EQ, compression, reverb – that are rock-solid and reliable. I don’t see how, conceptually, this is much different from how we provide our own stable versions of several libs, like libsndfile.
Two drawbacks (apart from closed source) of the linuxdsp effects are:
they aren't plugins, they're standalone so once I've got CH-EQ1s on each of my 24 tracks, how the heck do I know which one is the rythm guitar one? With LADSPA/LV2 they're "attached" to the track so I can immediately open and change the one I want.
Cumbersome interface. Say what you will about the ugliness of the LADSPA GUI but it's much easier and more intuive to drag a slider than grasping how you should move the mouse to adjust a knob. Also, I don't know if I can manually type in a value in any linuxdsp effect the way I can in LADSPA.
Well, I’ m using the linuxdsp plugins since a couple of weeks, and I agree with what peder says. Anyway they sound amazing (the reverb is the best available in linux for vocals), and I always have 3-4 instances of them in all of my projects.
But, my point is: Ardour needs at least a bunch of very stable and reliable plugins, just the essentials, and this is just not happening, and I wonder whether it’s better to invest time and resources into adding new features or solving this issue instead, cause with a showstopper like this you are not going anywhere in the pro world, I think. Sometimes I use Ardour and I feel like installing Windows XP and Cubase would give me more stability. And this is, well, bad.
I agree that linuxdsp ist not as easy to integrate as LADSPA plugins. But for the “apart from closed source”: I wouldn’t expect great open source plugins. From all what I’ve seen, good plugins are a hell of a work - they won’t be made without some financing, just as you see that Ardour isn’t made without financing, except that it’s a lot harder to get money for a single open source plugin…
So I don’t have a problem with closed source plugins in an open source DAW. I mean, no one so far has re-programmed the triple parametric EQ to not start with 0 Hz values or has found a solution for the +insanevalue db peaks in open source LADSPA plugins. It always comes down to “use this” and “don’t use that”.
I think that Ardour will get some real attention when there are some professional plugins available, and I wouldn’t bother paying for them (just imagine that waves or Sony Oxford would release their plugins for Linux in LV2 format
I was wondering about some kind of LADSPA wrapper for pure data, to have the ability to attach Pd patches to bus, tracks, and reload them with the session. In my opinion it can be a great feature for Ardour since Pd is a nice plateform to create clean DSP effects or synths (for ardour 3) and is very stable.
What do you think about it?
I totally agree with Benjamin. And I also think that Ardour itself would see the benefits of developing his own plugins: more professional feeling and ease of use = more donations. Since it’s so hard to get payed developing open source plugins, this is one more reason why Ardour developers should spend their time on that instead of giving unstable VST support, I think…
I think that a reliable set of open source plugins would benefit Ardour greatly.
The Calf LV2 compressor, FIL four band parametric and an easy to use convolving reverb would cover the essentials of recording and production.
If it’s not possible to use plugins reliably, then it doesn’t matter if they come with Ardour or are added by the user. Either way, people will have problems making ‘in the box’ mixes.
As part of a long, slow migration of my work onto Ardour/open source software, I’ve been getting some outboard effect units. However, basic eq and channel compression is a pain in the ass to do like this, especially when you need recall.
I appreciate that there are already a million things the Ardour developers are working on at the moment, so perhaps an outside project would be needed.
I’d be prepared to donate money, and time spent testing, for a project of this kind. Essentially just to polish a few existing plugins so that reliable operation and recall is guaranteed as far as possible.
I think it’s one problem shared with lots of open-source software: you try to give as many features as possible, without caring too much if they are stable or not, and keep adding features on features without stabilizing them. It’s a lack of professional thinking spread everywhere in the open source world.
While I can agree that it’s probably more fun for open source developers to hack on new features that to fix problems in existing code, I guess a fair amount of the problems has to do with the (single) developers limited resources in testing the software in different circumstances (different compilers, CPUs, optimizing flags, et.al.).
FWIW, Steve Harris has moved his LV2 plugins to github and invites people to hack on them: http://plugin.org.uk
There is also a LV2 version of the FIL 4-band parametric EQ that has a nice GUI in LV2Rack but the GUI extensions don’t work in Ardour without a patch (I can find no explanation of what that means on the website) http://nedko.arnaudov.name/soft/lv2fil/
As far as linuxDSP plugins, they have all just been updated and remember their settings on exit, If you don’t want to juggle too many of them open a new buss insert the plugin and send your various tracks to that buss, not so good for EQ’s but a great use of the Reverb. You have to think of linuxDSP more like patching in external hardware to a mixing console. I really like his Reverb and the new Graphic EQ is a great addition on the master buss IMHO
All this stuff will be in the next AV Linux ISO along with peder’s super groovy gtk Bristol launcher.
Hehe. I guess “super groovy” is surely a stretch, but anyways…
And, btw, I never got the lv2fil to work, even though I patched Ardour. It sorta hung (I even tested running in through lv2launch [is that what it’s called? Anyway, the zynjacku thing] and it hung there as well).
Thanks for passing that along, I may try it but both yourself and linuxDSP have tried it and no go, and you are both advanced Linux users, I’m not going to make it a high priority that’s for sure. Some of these plugin developers have to realize most people just want to use it and don’t get their jollies out of “funhouse mirrored” build dependencies.
I tip my hat to the Invada guy, he’s got Ubuntu packages ready, an easy to use compiling guide to build Debian packages and his source code is pretty much self contained except for the normal LV2 runtime stuff so folks on other systems (like rpm) can build them easily.
If any of you LV2 devs care, that is what USERS want, and it’s the end user that will determine whether LV2 is a viable and accepted plugin alternative or not. Sorry to sound pissy but some of this stuff is just ridiculously obtuse to build, package and distribute.