+- Harrison Mixbus Forum (https://forum.harrisonconsoles.com)
+-- Forum: Mixbus User Forum (https://forum.harrisonconsoles.com/forum-1.html)
+--- Forum: Mixbus General Discussion (https://forum.harrisonconsoles.com/forum-2.html)
+--- Thread: Supplemental Summing (/thread-11865.html)
Supplemental Summing - hodger - 04-28-2024
I've tried a few obvious methods for this but all end up making the mix engine crackle pop and distort horribly.
I"m trying to summ many channels to create a signal that can be used to externally key compressors assigned to those faders. In this test, 43 channels with their own compressor, externally keyed to the same routed feed. The effect should be similar to "overall" 2-mix compression, but at the individual channel point. The compressors work ok, and there is an overall level to feed the side-chains, but in all my efforts, the mix engine rolls over.
I tried feeding a Group to the side-chains.
I tried routing a MixBus to a track-IN, then using THAT track as the side-chain source.
I tried creating a separate additional summing bus by assigning all the channels to a common track-IN and sending that to the side-chains.
Seems that even routing a MixBus to a track-IN makes the "feedback" indicator light up even though I've created no loops.
Any limits on the quantity of side-chains available? Maybe I exceeded it?
Any rules about using the router as a summing amp onto a track-IN?
Maybe side-chains are limited to individual track I/O?
All the things I'm trying would work on an analog desk, and theoretically should here, but something's getting in the way.
Thanks for any suggestions.
h
RE: Supplemental Summing - madmaxmiller - 04-28-2024
(04-28-2024, 05:33 AM)hodger Wrote: Seems that even routing a MixBus to a track-IN makes the "feedback" indicator light up even though I've created no loops.
I can explain at least this:
The feedback*warning* lights up because every track or aux bus can potentially be sent to that mixbus (with one mouse click) - hence the warning. In earlier versions routing "back" was not possible at all for that exact reason.
Other than that: what do the meters at your "aggregated sidechain" say? Have you tried to build it one by one? Like, send 1 channel to a bus, then the bus output to the compressor key? If that works add another one and check, then another 3 or so and so on?
Cheers and please keep us posted on the results,
MMM
RE: Supplemental Summing - hodger - 04-29-2024
Some updates...
Summing through an Audio Buss also has the same problem.
Found that recording the output of a mixbus onto an unassigned track also puts horrible stuff into the master mix, but the track recorded is clean.
Using this newly created mix/bounce as the side-chain source works flawlessly, except that it is no longer real-time. So, at least there doesn't seem to be an issue related to the overall quantity of Side-Chains being built, just the nature of the signal source being used.
Seems MB has issues with cycling signals through the mixer, even when properly assigning in the router.
One other odd note, saving and recalling these sessions shows acceptable DSP load upon save, but 99-100% overload upon recall, even when the MacOS activity monitor shows small to moderate loads.
h
RE: Supplemental Summing - madmaxmiller - 04-29-2024
(04-29-2024, 01:23 AM)hodger Wrote: Found that recording the output of a mixbus onto an unassigned track also puts horrible stuff into the master mix, but the track recorded is clean.
Have you checked the master sends? The master send of the track you record into should be off.
MMM
RE: Supplemental Summing - hodger - 04-29-2024
(04-29-2024, 05:46 AM)madmaxmiller Wrote:(04-29-2024, 01:23 AM)hodger Wrote: Found that recording the output of a mixbus onto an unassigned track also puts horrible stuff into the master mix, but the track recorded is clean.
Have you checked the master sends? The master send of the track you record into should be off.
MMM
In my case, the recording track assign to the master was off. I decided the choice was either the record track unassigned or the mixes unassigned. Even both assigned should work in theory, but the feed into the master bus would be doubled up and messy.
h
RE: Supplemental Summing - hodger - 04-30-2024
Some success finally, with an uncomfortable amount of trouble discovered.
Skipping all the compressor keying concerns (which seem NOT to be the problem) I made 11 sessions with just the tracks. All 43 stereo tracks routed to MixBus1, MixBus1 routed to an unrouted audio track on input. MixBus1 not assigned to MSTR. Always either trash from the mix engine in the master and/or several mixgroups.
Tested 11 apps, MixBus 5 and MixBus 5 32C all the way to current MixBus 10. 32C v7 v8 and v9 all "could" be useful as long as you choose not to use the last 3 groups, plus the successful part of the summing was slightly delayed. Not great for a compressor KEY.This is a consistent problem and fully reproducible.
Today I tried using MixBus12 as the group instead of MixBus1, and the routes succeed and I hear no garbage in the master or other groups.
I can call this a successful recipe for my trial, but I'm not really comfortable with all the findings along the way. The router sure indicates many things are possible, but so many of them failed these last few days it's not a confidence-building discovery.
h
OOPS, looks like I spoke too soon. Saved the session, wrote the post, recalled the session and everything is garbage. Session DSP was a restful 34% when I saved, but 100% and glowing and trashy upon recall and another play.
Guess this pursuit may not be possible with MB, at least without some formal advice or troubleshooting.
RE: Supplemental Summing - madmaxmiller - 04-30-2024
(04-30-2024, 02:34 AM)hodger Wrote: Some success finally, with an uncomfortable amount of trouble discovered.
Skipping all the compressor keying concerns (which seem NOT to be the problem) I made 11 sessions with just the tracks. All 43 stereo tracks routed to MixBus1, MixBus1 routed to an unrouted audio track on input. MixBus1 not assigned to MSTR. Always either trash from the mix engine in the master and/or several mixgroups.
Tested 11 apps, MixBus 5 and MixBus 5 32C all the way to current MixBus 10. 32C v7 v8 and v9 all "could" be useful as long as you choose not to use the last 3 groups, plus the successful part of the summing was slightly delayed. Not great for a compressor KEY.This is a consistent problem and fully reproducible.
Today I tried using MixBus12 as the group instead of MixBus1, and the routes succeed and I hear no garbage in the master or other groups.
I can call this a successful recipe for my trial, but I'm not really comfortable with all the findings along the way. The router sure indicates many things are possible, but so many of them failed these last few days it's not a confidence-building discovery.
h
OOPS, looks like I spoke too soon. Saved the session, wrote the post, recalled the session and everything is garbage. Session DSP was a restful 34% when I saved, but 100% and glowing and trashy upon recall and another play.
Guess this pursuit may not be possible with MB, at least without some formal advice or troubleshooting.
Thanks for sharing. I wish I had more time to follow up with my own research.
But maybe you could do the last partially successful move again in a new session and compare the session files after saving and after reopening. Could it be that the routing ends up in a loop at reopen? 100% DSP and garbage in the speakers point at something like this.
MMM
RE: Supplemental Summing - Nathan@Harrison - 04-30-2024
(04-30-2024, 03:12 AM)madmaxmiller Wrote:(04-30-2024, 02:34 AM)hodger Wrote: Some success finally, with an uncomfortable amount of trouble discovered.
Skipping all the compressor keying concerns (which seem NOT to be the problem) I made 11 sessions with just the tracks. All 43 stereo tracks routed to MixBus1, MixBus1 routed to an unrouted audio track on input. MixBus1 not assigned to MSTR. Always either trash from the mix engine in the master and/or several mixgroups.
Tested 11 apps, MixBus 5 and MixBus 5 32C all the way to current MixBus 10. 32C v7 v8 and v9 all "could" be useful as long as you choose not to use the last 3 groups, plus the successful part of the summing was slightly delayed. Not great for a compressor KEY.This is a consistent problem and fully reproducible.
Today I tried using MixBus12 as the group instead of MixBus1, and the routes succeed and I hear no garbage in the master or other groups.
I can call this a successful recipe for my trial, but I'm not really comfortable with all the findings along the way. The router sure indicates many things are possible, but so many of them failed these last few days it's not a confidence-building discovery.
h
OOPS, looks like I spoke too soon. Saved the session, wrote the post, recalled the session and everything is garbage. Session DSP was a restful 34% when I saved, but 100% and glowing and trashy upon recall and another play.
Guess this pursuit may not be possible with MB, at least without some formal advice or troubleshooting.
Thanks for sharing. I wish I had more time to follow up with my own research.
But maybe you could do the last partially successful move again in a new session and compare the session files after saving and after reopening. Could it be that the routing ends up in a loop at reopen? 100% DSP and garbage in the speakers point at something like this.
MMM
I would have to hear the final mix to determine if this process is even merited. Just because you can doesn't mean you should. Keep it simple.
RE: Supplemental Summing - hodger - 05-01-2024
(04-30-2024, 09:50 AM)Nathan@Harrison Wrote: I would have to hear the final mix to determine if this process is even merited. Just because you can doesn't mean you should. Keep it simple.
In other words, it SHOULD be more easy and simple to see if it is worth it
It is obviously NOT simple which leads to your suggestions to possibly avoid it. If stereo master buss compression was unnecessarily difficult to patch up, maybe folks would be hesitant to experiment. History shows how useful master bus compression is, even in the "stereo" portion of Mixbus, simple as intended. It is not at all strange to consider keyed compression opportunities to tame object-based entities while watching out for loudness targets. Lot of engineers do, even if just mix and match on groups.
I'd like to see the roadblocks removed. I've pointed out several methods to test this, with complete expectations that they WOULD work, except they failed. This is an application failure, not a user failure, and an unnecessary set of hurdles. I've pointed out that at least the past ELEVEN versions of the apps have this failure, and your comment seems more of a diversion instead of admission and concern. Perhaps a better response is for troubleshooting and app improvements. Strange.
FYI - I have completed my recipe after jumping over yet another app failure.
A - As explained yesterday, this summing exercise does not work the same across all MixBusses. Group 12 did work for me yesterday in MB10, while group 1 failed in ALL apps since v5 and 32Cv5. Maybe one day I'll look into all possible groups across all possible apps, but I hope there are some fine programmers taking a look into what is wrong, how nobody noticed for years. and how quickly a patch can be rolled out.
B - The failure to recall the same thing that was saved is worrisome. For my recipe, I found that I can disable the Mixbus12-to-Track-IN route BEFORE saving, and then re-enable it AFTER recalling, and all is ok. Just another thing that should be "SIMPLE" but it is not. Again, it would be reassuring to hear that some fine programmers are going to look into the failure-to-restore problem and work on a quick patch.
I've publicly praised the team already for how well Atmos has been integrated into v10, with the caveat of noting how many items were avoided to accommodate the elegance and haste. Compared to MB10, all other Atoms implementations are cumbersome and overly complex.
I want to use MB10 for Atmos, but want just a little more than is currently provided, and yes it makes me willing to do some of these non-SIMPLE things to pull it off. Note that I'm still here, while perhaps others took a long look and kept on with their other apps.
h
RE: Supplemental Summing - Lars - 05-01-2024
Wouldn't be the old Sidechain-Bus in MB V3/4 something which could make a really nice comeback here to add a simple, fast and very much Harrison-style solution?