| *** gianMOD has joined #lv2 | 00:12 | |
| *** gianMOD has quit IRC | 00:19 | |
| *** unclechu has quit IRC | 01:23 | |
| *** Magnus_RM has quit IRC | 01:32 | |
| *** ricardocrudo_ has joined #lv2 | 03:27 | |
| *** ricardocrudo__ has quit IRC | 03:30 | |
| * drobilla is forced by ingen's property editor to manually clean up all the vocabularies we use | 04:29 | |
| drobilla | Such a mess | 04:29 |
|---|---|---|
| drobilla | Or just used schema.org or something | 04:30 |
| * drobilla wonders where he got this completely botched regexp for xsd:date | 05:53 | |
| drobilla | Fun facts discovered in unpleasant ways: no LV2 thing has been released on a day where the second digit is 0 (e.g. the 20th) | 05:54 |
| *** edogawa has joined #lv2 | 07:20 | |
| *** falktx_ has joined #lv2 | 07:24 | |
| *** edogawa_ has joined #lv2 | 07:24 | |
| *** edogawa has quit IRC | 07:27 | |
| *** damo22 has joined #lv2 | 08:38 | |
| damo22 | is there a way to mark one of the input audio channels as a sidechain input? | 08:39 |
| falktx_ | I don't think so | 08:42 |
| drobilla | damo22: Indirectly. You can specify a port group as the main input/output | 08:42 |
| drobilla | "sidechain" = not that | 08:42 |
| drobilla | I suppose we could add a sidechain PortProperty | 08:43 |
| damo22 | is there a plan to handle sidechaining in ardour better? would it include doing something like ^ | 08:50 |
| drobilla | That would depend on how you define "plan" :) | 08:57 |
| *** ricardocrudo_ has quit IRC | 08:58 | |
| damo22 | drobilla: i mean have others requested it enough for someone to think about how to implement it | 08:58 |
| drobilla | Sure. Sidechain talk is practically as old as Ardour itself | 08:59 |
| drobilla | As are numerous other issues with the current relatively shitty I/O routing that just doesn't let you insert plugins at all far too often | 08:59 |
| damo22 | i mean, if i write an lv2 with 2 inputs, (one sidechain and a mono in), how can i tell ardour not to treat it as a stereo plugin | 09:00 |
| damo22 | is that an LV2 problem or an Ardour problem | 09:01 |
| drobilla | Well, theoretically, as I said, you can designate a main group, which the sidechain wouldn't be in | 09:01 |
| drobilla | In practice, you can't. | 09:01 |
| damo22 | why not, in practice | 09:02 |
| drobilla | Because it isn't implemented? Obviously? :P | 09:02 |
| drobilla | You might be able to add a send and get what you want anyway, I'm not sure. | 09:03 |
| damo22 | ok so if i write some lv2s with a main port group on the i/o i want as main, i can theoretically patch ardour to recognise that any lv2 with a main port group that also has inputs not part of the group are sidechains? What if there are other kinds of inputs | 09:04 |
| drobilla | Well, good question. What is a "sidechain" anyway? | 09:05 |
| drobilla | It pretty much means "not main input" | 09:05 |
| drobilla | A PortProperty would be easier to graft on to existing plugins, though | 09:05 |
| damo22 | a sidechain is something that affects the plugin's sound but isnt passed through as main input | 09:05 |
| damo22 | and does not add to the 'main' channel count | 09:06 |
| drobilla | I don't know how hard it would be to have Ardour automatically make returns for ports on adding a plugin, which is presumably the thing to do. | 09:07 |
| drobilla | Wasn't on the 4.0 shortlist | 09:07 |
| damo22 | returns for ports? | 09:09 |
| drobilla | Return as in the opposite of a send. | 09:10 |
| drobilla | What else would you have it do? | 09:10 |
| damo22 | drobilla: so you mean, insert the plugin on the track where the sidechain is supposed to connect and then add returns on the main outs to another track? | 09:13 |
| drobilla | main outs? | 09:14 |
| drobilla | Why is this confusing? | 09:14 |
| damo22 | main outs of the plugin | 09:14 |
| drobilla | You want to insert a plugin into a track with, say, a stereo main pair and one sidechain | 09:14 |
| drobilla | That stereo track has 2 audio channels | 09:14 |
| drobilla | Where is the channel for the sidechain going to come from? | 09:14 |
| drobilla | There is no such channel at that point in the strip | 09:14 |
| drobilla | => you need a return before the plugin to get one | 09:15 |
| damo22 | nah | 09:15 |
| damo22 | that will cause all sorts of ordering problems | 09:15 |
| drobilla | Uh... okay | 09:15 |
| drobilla | As opposed to what? | 09:15 |
| damo22 | you need a separate menu with sidechain inputs so you can route channels to them | 09:15 |
| damo22 | it wont be part of the strip | 09:16 |
| drobilla | So you want to build a completely separate all-to-all routing mechanism in Ardour specifically for sidechains that does the same thing | 09:16 |
| drobilla | and this *avoids* problems? ;) | 09:16 |
| damo22 | not all to all | 09:16 |
| damo22 | all to sidechain | 09:17 |
| damo22 | it can be a "connect sidechain..." menu in the right click for a plugin | 09:17 |
| drobilla | Still the same problem. | 09:17 |
| damo22 | all it has to do is identify which port is for sidechaining and add it to the menu for that plugin | 09:19 |
| damo22 | and if a plugin has no sidechain, the menu item is missing | 09:19 |
| drobilla | Making a menu is trivial | 09:19 |
| damo22 | yeah | 09:20 |
| drobilla | Irrelevant veneer to how it actually would work, which you can add or not | 09:20 |
| drobilla | or a dialog to do it on insert or whatever | 09:20 |
| drobilla | Point is completely reengineering the ardour audio engine just for this instead of doing it the way all other I/O send/return/insert type things are already done is insane | 09:20 |
| damo22 | no one wants to see a return at the top of the strip just for a sidechain | 09:21 |
| drobilla | Tell you what, you actually get this working and we can ardour in #ardour about UI trivialities ;) | 09:22 |
| drobilla | argue* | 09:22 |
| damo22 | well in order to make this work, i think it needs a port property added | 09:22 |
| damo22 | or i'll have to detect it with the port groups | 09:23 |
| drobilla | (anyone who wants a level control for it would, though) | 09:24 |
| drobilla | Either way. A property for this seems reasonable | 09:27 |
| drobilla | The main utility of groups is actually doing multi-channel correctly, but Ardour and pretty much everythign else doesn't do that either | 09:27 |
| * drobilla likes to joke that Jack doesn't support stereo | 09:27 | |
| damo22 | haha | 09:27 |
| drobilla | It'd be nice to support discrete surround, ambisonics, etc, with automagic insertion of mixing and panning and so on, but bunch of work for stuff that almost nobody cares about | 09:29 |
| drobilla | Though I'd think with the video stuff this would become more of an issue | 09:29 |
| drobilla | damo22: I'll toss in lv2:isSideChain tomorrow provided I remember etc | 09:30 |
| damo22 | :D | 09:30 |
| drobilla | For now.... it's 5:30 am and that is extremely not good. | 09:30 |
| * drobilla => zzzzzzzz | 09:31 | |
| damo22 | sleep well | 09:31 |
| drobilla | damo22: I lied http://lv2plug.in/git/cgit.cgi/lv2.git/commit/?id=fc14fc9a14532f2bf1efa39f670322b60415473b | 09:37 |
| damo22 | thanks! | 09:38 |
| falktx_ | drobilla: is that restricted to inputs? | 09:40 |
| damo22 | it doesnt make much sense to have a sidechain output does it? | 09:42 |
| falktx_ | everything can make sense in a modular host | 09:43 |
| falktx_ | but modular hosts don't need those special properties anyway | 09:44 |
| drobilla | Not in any formal sense. I don't see much benefit in over-specifying it. | 09:44 |
| drobilla | Though now that I think of it, maybe sidechain is a reasonable designation | 09:44 |
| drobilla | Which would mean e.g. if there was a sidechain output and a bunch of other stuff, the host would connect like to like | 09:45 |
| drobilla | That's probably better, port properties are kind of annoying | 09:45 |
| drobilla | Might change it tomorrow | 09:45 |
| drobilla | g'night | 09:45 |
| falktx_ | night | 09:45 |
| falktx_ | Harry might be interested to hear about this sidechain port | 09:52 |
| *** HarryHaaren has joined #lv2 | 10:06 | |
| *** ricardocrudo_ has joined #lv2 | 10:22 | |
| *** NickSB2 has quit IRC | 10:54 | |
| *** ricardocrudo_ has quit IRC | 11:49 | |
| *** gianMOD has joined #lv2 | 12:25 | |
| *** gianMOD has quit IRC | 12:31 | |
| *** ricardocrudo_ has joined #lv2 | 12:34 | |
| falktx_ | HarryHaaren: did you see the latest change in lv2-git ? | 12:37 |
| HarryHaaren | falktx_, nope | 12:38 |
| HarryHaaren | sidechain stuff? | 12:39 |
| falktx_ | I think they might interest you | 12:39 |
| falktx_ | yes | 12:39 |
| * HarryHaaren is looking at changlog | 12:39 | |
| falktx_ | it's now a port property | 12:39 |
| falktx_ | you can set an audio port as sidechain (meta-data only) | 12:39 |
| HarryHaaren | cool | 12:39 |
| HarryHaaren | lets talk about things at LAC - i'm working on ArtyFX 1.3 now (1 bug remaining.. clicks on preset-switching) | 12:40 |
| HarryHaaren | and then i'm going for ice-cream in the city - its a beautiful day here! | 12:40 |
| HarryHaaren | tomorrow I'll have time to code some things - and look at the sidechain | 12:40 |
| HarryHaaren | that said - only Ducka currently has a sidechain, if its as simple as adding a lv2:portProperty, then np | 12:41 |
| *** HarryHaaren has quit IRC | 13:10 | |
| *** gianMOD has joined #lv2 | 13:23 | |
| *** gianMOD has quit IRC | 13:27 | |
| *** gianMOD has joined #lv2 | 13:27 | |
| *** gianMOD has quit IRC | 13:32 | |
| *** gianMOD has joined #lv2 | 14:02 | |
| *** gianMOD has quit IRC | 14:05 | |
| *** gianMOD has joined #lv2 | 14:33 | |
| *** rncbc has joined #lv2 | 14:41 | |
| *** gianMOD has quit IRC | 14:51 | |
| *** edogawa_ is now known as edogawa | 15:48 | |
| *** gianMOD has joined #lv2 | 15:50 | |
| *** falktx_ has quit IRC | 15:52 | |
| *** gianMOD has quit IRC | 16:12 | |
| *** gianMOD has joined #lv2 | 16:43 | |
| *** falktx_ has joined #lv2 | 17:00 | |
| *** NickSB2 has joined #lv2 | 17:08 | |
| *** gianMOD has quit IRC | 17:13 | |
| *** gianMOD has joined #lv2 | 18:15 | |
| rgareus | damo22, drobilla: dedicated side-chains in ardour won't happen. the graph (what feeds what) is linear, with the only exception of internal sends (which do have latency compensation built into) | 19:30 |
| rgareus | what needs to become easier howerver is to move N -> M channels plugins around the processor box in Ardour. | 19:32 |
| rgareus | similar to what we discussed a while ago: http://robin.linuxaudio.org/tmp/fancy_wires.png (just an illustration) | 19:35 |
| *** magnetophon has quit IRC | 20:40 | |
| *** gianMOD has quit IRC | 20:48 | |
| *** falktx__ has joined #lv2 | 20:49 | |
| *** falktx_ has quit IRC | 20:53 | |
| *** unclechu has joined #lv2 | 21:13 | |
| *** falktx__ has quit IRC | 21:50 | |
| drobilla | rgareus: Reasons why all of ardour should be a graph ++ | 21:58 |
| drobilla | Then, what the hell, throw a patcher in there, etc | 21:58 |
| drobilla | rgareus: There's always "use a Jack return and don't care about latency"? | 21:58 |
| rgareus | except if there's no JACK | 22:00 |
| rgareus | besides, we need better latency compensation not more 'meh'. | 22:00 |
| rgareus | the ardour-port-engine is perfectly capable of doing non-linear graphs. the route itself is not. | 22:02 |
| rgareus | you can already model the whole thing with send/returns (both internal/external). that's IMHO the way to go. not slapping on dedicated stuff only for side-chains. | 22:04 |
| *** rncbc has quit IRC | 22:09 | |
| drobilla | Well, yes, clearly using normal sends/returns is the way it must be done. | 22:14 |
| drobilla | The details of internal/external/latency comp/etc needn't have anything to do with plugin sidechains directly | 22:15 |
| drobilla | We just need a return(s) above it clever enough to put the new channel in the right place and so on | 22:15 |
| *** ricardocrudo_ has quit IRC | 22:33 | |
| *** gianMOD has joined #lv2 | 22:50 | |
| *** gianMOD has quit IRC | 23:07 | |
| * drobilla wonders how to deal with plugins that, er, change "plugins" on save | 23:12 | |
| drobilla | e.g. if you load an ingen template in ardour and build something in it, it's forever "Ingen Stereo Instrument Template" (or whatever) in Ardour | 23:13 |
| *** edogawa has quit IRC | 23:18 | |
Generated by irclog2html.py 2.13.0 by Marius Gedminas - find it at mg.pov.lt!
