I started teaching myself a new workflow on Mixbus

[Domino]

Group colouring is a bit more visible than channel ones

NICE!!! that's a great way to do it.....thanks! Actually looking more into groups, it seems like a core of Mixbus workflow......

does collect group work for you? not for me.....I though that would be a GREAT was to quickly organize all my imported tracks in order, and then colour them of course......but it doesn't work?

Collect Group is broken for me too. I've not reported the issue to support if you want to do it - I'm already annoying them enough over other things..

[haraldthi]

I have similar problems with lack of buses.
For me, it started out with trying out a type of reverb that was recommended to me. Room, plate, a slightly delayed hall, plus slapback. That's four channels, and feeding those channels from many buses by setting up sends to aux channels is a workflow killer. This is precisely what those knobs on the mix strip for sending to mixbus channels work well for, sending differing amounts of signal from many tracks to several buses.

While sending the main audio signal goes to one bus only per track, so in theory I could just as well use the output at the bottom at the strip to an aux bus.
The lack of delay compensation is actually a minor problem, as I mostly use this for grouping, fader automation and perhaps some EQ. Even if it isn't all that difficult, in theory. You need an algorithm to detect loops, but the basics of this is well known computer theory. (Have two pointers A and B. A goes through every node in the chain, B goes through every other node in the chain. Compare A and B at every step, and if A and B are equal at any time before you come to the end of the chain, you have a loop. You just need to adapt this to branches in the chain, which should be easy enough. (Edit: With a bit of thinking through, perhaps not that easy. But tracking paths and using a stack or list to pile up the different alternative paths and checking them all, it's far from impossible either.)) A loop needs minimum delay, while a non-loop needs delay compensation.
But as I said, I can live with the small delay. I don't use tape saturation much either, a little bit at the main bus is usually enough for me.

But the lack of panning to the aux bus send? Aaaarrrgh! Everything ends up in the middle!
The practice of only sending mono tracks to mono buses and stereo tracks to stereo buses should be amended too, as I mix mono and stereo tracks.
* A mono track/bus to a mono aux bus should be sent without panning.
* Stereo track/bus to a mono aux bus shold be summed to mono.
* A mono track/bus to a stereo aux bus should use the panner to split into stereo.
* A stereo track/bus to a stereo aux bus should use the balance/panner before sending stereo. It's easy enough to leave the balance knob at neutral if you don't want to change it.
Only question I see is how much volume should eventually be reduced when summing stereo to mono. But I guess you have some established practice here.
(And am I the only one who almost don't use "send to master"? Perhaps replace it with the output at the bottom defaulting to the master bus? I don't know. Perhaps the "send to master" button is easier to find for beginners.)

When 3.1 is out and stable, I hope this is a high priority.

[mcadder]

I have similar problems with lack of buses.
For me, it started out with trying out a type of reverb that was recommended to me. Room, plate, a slightly delayed hall, plus slapback. That's four channels, and feeding those channels from many buses by setting up sends to aux channels is a workflow killer. This is precisely what those knobs on the mix strip for sending to mixbus channels work well for, sending differing amounts of signal from many tracks to several buses.

While sending the main audio signal goes to one bus only per track, so in theory I could just as well use the output at the bottom at the strip to an aux bus.
The lack of delay compensation is actually a minor problem, as I mostly use this for grouping, fader automation and perhaps some EQ. Even if it isn't all that difficult, in theory. You need an algorithm to detect loops, but the basics of this is well known computer theory. (Have two pointers A and B. A goes through every node in the chain, B goes through every other node in the chain. Compare A and B at every step, and if A and B are equal at any time before you come to the end of the chain, you have a loop. You just need to adapt this to branches in the chain, which should be easy enough. (Edit: With a bit of thinking through, perhaps not that easy. But tracking paths and using a stack or list to pile up the different alternative paths and checking them all, it's far from impossible either.)) A loop needs minimum delay, while a non-loop needs delay compensation.
But as I said, I can live with the small delay. I don't use tape saturation much either, a little bit at the main bus is usually enough for me.

But the lack of panning to the aux bus send? Aaaarrrgh! Everything ends up in the middle!
The practice of only sending mono tracks to mono buses and stereo tracks to stereo buses should be amended too, as I mix mono and stereo tracks.
* A mono track/bus to a mono aux bus should be sent without panning.
* Stereo track/bus to a mono aux bus shold be summed to mono.
* A mono track/bus to a stereo aux bus should use the panner to split into stereo.
* A stereo track/bus to a stereo aux bus should use the balance/panner before sending stereo. It's easy enough to leave the balance knob at neutral if you don't want to change it.
Only question I see is how much volume should eventually be reduced when summing stereo to mono. But I guess you have some established practice here.
(And am I the only one who almost don't use "send to master"? Perhaps replace it with the output at the bottom defaulting to the master bus? I don't know. Perhaps the "send to master" button is easier to find for beginners.)

When 3.1 is out and stable, I hope this is a high priority.

That is spot on! I've just started using Mixbus and really like the workflow. There are some minor things that needs a fix or an implementation though...

[haraldthi]

One is usually used for a gated reverb on snares (so it can't be used with anything else) and only one left for all the remaining sends. Some of them can be hacked using aux busses, but they're not latency compensated and many of the fx's have a strong latency. So yes, the 8 mixbuses introduce some limitations for certain music genres... ;-)

One way to get around things, if you can live with doing it with single tracks, is duplicating the track. You can even link the playlist instead of copying it, so all changes to the tracks themselves end up in both tracks. I often use this with parallell compression of single instruments, and can be used for reverbs and such too.

Another way to do it is use parallell pathways in the plugin section (using the pin connections) and then a small mixer at the end put it together again. It works with simpler setups where you are only about to do a tweak or two, but you can't involve the Mixbus strip in it.

[jonetsu]

One way to get around things, if you can live with doing it with single tracks, is duplicating the track. You can even link the playlist instead of copying it, so all changes to the tracks themselves end up in both tracks. I often use this with parallell compression of single instruments, and can be used for reverbs and such too.

Thanks for mentioning the linking of playlists. Looks like it can be useful. Haven't used it yet.

[Matt]

I can't say that I've discovered "the workflow" for Mixbus yet. When ever I think I've figured it out I tend to hit a strange limitation. Like using region gain to gain stage all tracks to -18.0 and finding each track needs to be pretty much full screen for editing. Mixbus is missing a separate visual zoom level so details can be seen on normal sized tracks for that way to be correct.

(Snip)

Not sure why you're doing this. Unless the tracks were recorded way too hot. But even if so, why not use the input trim knob to bring them down. Not sure if it has a range of -18 dB, but would bring down the level without changing the waveform.

http://www.santoclemenzi.com/wp-content/...otated.png

Edit: just looked it up. Looks like there's a range of -20 to +20 dB.

[Domino]

I can't say that I've discovered "the workflow" for Mixbus yet. When ever I think I've figured it out I tend to hit a strange limitation. Like using region gain to gain stage all tracks to -18.0 and finding each track needs to be pretty much full screen for editing. Mixbus is missing a separate visual zoom level so details can be seen on normal sized tracks for that way to be correct.

(Snip)

Not sure why you're doing this.

I do some mixing for online collaborations where recorded levels can be all over the place. Selecting a group of tracks and using the region normalise to get them in the ballpark is just a time saver. I use trim for adjustments from there. The normalise can do individual levels or a single adjustment over the selection, so it's a fast approach for changing multiple levels at the same time and can maintain balance between tracks when needed.

Edit: It's my take on this https://www.youtube.com/watch?v=3b3DtQALtuY without the noise

[madmaxmiller]

https://www.youtube.com/watch?v=3b3DtQALtuY

We should post that into the "looking for script ideas" section
Seriously, a script can do that, probably more accurately than a human, without making a noise, in a fraction of a second. We could name it "quick_mix_by_numbers". It would be a big stress relief in the CR when people sit on your back wanting a sneak peak of the latest take. I mean, we can still can throw around (fake)faders while the script got us ready

MMM

Edit: ...and the script would always consider the whole length unless told otherwise...

[Domino]

https://www.youtube.com/watch?v=3b3DtQALtuY

We should post that into the "looking for script ideas" section
Seriously, a script can do that, probably more accurately than a human, without making a noise, in a fraction of a second.

That was my thoughts too. And not long after I posted about it Harrison kindly added an RMS option to my workflow for it

Select the regions you want to set, right click the selection and in the popup menu do "Selected Regions -> Gain -> Normalise". You can then adjust to a peak or rms level with the target per region or across the selection. It's non destructive as it doesn't alter the source data, it just adjusts the region gains.

I can't think of anything a script would add to how I already do it

[madmaxmiller]

Select the regions you want to set, right click the selection and in the popup menu do "Selected Regions -> Gain -> Normalise". You can then adjust to a peak or rms level with the target per region or across the selection. It's non destructive as it doesn't alter the source data, it just adjusts the region gains.

But it pulls up the noise floor before it hits a compressor or anything, too, so the compressor will emphasize the noise even further... and channels sum up... and a hiss might become audible because its now in the "good" 16 bits. After someone here explained that to me I tend to rather pull down louder tracks and add gain/normalisation at export time...

MMM