VST Live song/part transition with Crossfade

I’m looking into using VST Live to do real-time Cubase-type audio mixing for live-streaming. The first attraction is the ability to switch between songs in a manner similar to scene save/recall capability on hardware audio mixing consoles. I’d like to be able to take snapshots of levels (and sends and other parameters) for various music segments before the live performance and then be able to recall those scenes and have VST Live perform a smooth and seamless crossfade transition from the current settings to the next song or part. There should be a way to specify the crossfade time.

It is not necessary to reroute/repatch during a scene/song/part change, but it is necessary to be able to change the value of any parameter such as volume level, pan, effects send levels, etc.

Played with VST Live enough to make a song with a part, a stack and track with mic input set to a level. Then duplicated the part, changed in the input level in the stack and then toggled between the two parts. The mic had an instant level change with a little bit of a click. Definitely not a crossfade.

It seems changing the input level in the track is global to the song parts and did not change when toggling parts. I think it would be more logical for the stack to be the constant and the track level to be what I would want to change between parts…

Perhaps there is another way to achieve what I’m trying to accomplish, but its not jumping out at me.

That is to be expected, but why would you want to change the input of a Stack so much?

You cannot change input levels for tracks, only with Stacks, and even there it is just a helper function, because we cannot change the gain of your audio interface.

For Layers as well as for Stacks, there ist the Sustain settings in Preferences. The assumption here is that all signals, reverbs etc are released within that time frame, otherwise there will be glitches.
There is Song crossfading on the wishlist, but that’s not quite so easy as each Song is like an entire Cubase project. But Layer and Stacks Sustain works across Song boundaries, give it a try.

Thank you @musicullum, I spent more time investigating VST Live (1.1.40) and have a better understanding of how it works. Here’s some things I ran into when trying to set this up in a manner similar to Cubase or a hardware audio console for live stream audio mixing:

I don’t think I need any Layers or Tracks. I seem to be stuck with at least one Layer - can delete all but the last one remaining. This unused Layer occupies a channel in the Mixer.

Stacks is where I’m bringing in, say, 48 mic channels. I set up a Stack for each mic. When I go to the Mixer, I find I want to reorder the Stack channels a little. Looks like that’s not possible, so I go back to Stacks where I can re-order. However a Stack that is moved disappears in the Mixer, and its not in the Visibility list. If I try to remake it in Stacks, it is always added at the bottom, so moving it anywhere makes it disappear in the Mixer. When a stack is moved it no longer works at all… no audio appearing on the Stack’s input meter. Reassigning its audio input doesn’t fix it.

Organizing Stacks/mics into Groups works. But I can’t move/reorder Groups in the Mixer.

Is there a way to link the faders of two Stacks? Its not uncommon to build a stereo signal from two mono sources, but there are some different EQ/volume/FX settings on the different mono channels.

A Stack can send to two different FX channels. Where do FX-1 and FX-2 feed to? I’d expect to find Stack-like channels in the Mixer that I could put effects on and then route them as I please. In the Mixer view of a Stack it’s possible to have 4 Sends. Hum, seems odd to have 2 sends in Stack view and 4 in Mixer view. Personally I’m preferring the Mixer view, and I don’t care so much about the Stack view. I definitely need to be able to send to more than just FX-1 and FX-2. The 4 sends in Mixer view should be enough to send to a range of FX channels. The “e” view of a Stack in Mixer view could have more stuff, like the source input, better pan number visibility, etc.

Back to Crossfade… for my purposes I don’t really need to reconfigure a complete setup, I just need to be able to change some levels and some parameters like EQ settings and FX send levels, etc. I’m going to keep all the same Stacks with their Inserts and Sends. If I really want to do something tricky, I’m more likey to duplicate a Stack and make changes there. I say this because I don’t really care if I’m crossfading between two Parts or two Songs. I could load all my setting changes as Parts for all I care and just have one Song… if that would make Crossfading more practical. But it seems there is not really any internal difference between a Song and a Part - just a logical grouping convenience for the user.

Another thing that is needed is some method to compute latency. Cubase displays the latency for each channel which makes it relatively easy to add delay to channels with lower latency in order to match the delay of long latency channels.

1 Like

Edit/Preferences/Layers/No Default Layer

I guess that what you want to do you don’t need Layers, you could just use the Mixer.

A Stack is an entire submixer, mixing its core channel (with the inserts) to “OUT”, sending to its own Fx1 and/or Fx2 channels via SEND (see pic), and then OUT, FX1, and FX2 are summed to the resulting channel (large volume etc representing the mixer channel). Nothing is sent outside.


in → inserts →
→ to FX 1 and 2 via SEND
→ to OUT
OUT + FX1 + FX2 → mixer channel

So I guess what you really want might be better done in the mixer. Stacks are meant for guitar and vocal processing per Part. That beeing said: Mixer Groups (and OUTs) are project wide, tracks are Song wide, and Stacks are Part wide.

Channel moving will be added soon.

Will check!

True, fixed with the next version.

Way off. Songs not just contain Parts, they contain TRACKs (that’s a lot), Notes, some Modules etc, organize Part triggers and so forth.
If you exclude any transport (timeline) actions you are about right, but then there are still assets belonging to the Song.

Thank you for the responses, they are most helpful and explanatory.

I can’t figure out how to bring a microphone input into the Mixer. The manual indicates a Stack is required. Is there some other way?

You can either create a Stack, or an audio track.

So Stacks seem to be overkill for microphone inputs. I tried Tracks (which is similar to Cubase). At first there were no channel strips for the tracks in the Mixer. It took awhile to figure out that clicking on the “e” symbol of the Song would make channel strips for the tracks unfold to the right. But they unfold along with the rack for the Song, so it takes up some extra room. So if I have 48 mics that will take up some room and the Song rack will consume space as well. It seems odd the track channel strips are not persistent.

Then I tried duplicating the Song, the duplicate had no Tracks. So I would need to completely rebuild the “Mixer” setup for each Song, even if I’m keeping all the same tracks, just changing levels.

I can reorder tracks in the Track view, but it has no effect on channel strip order in the Mixer view. I cannot reorder channel strips in the Mixer view. And there is some odd behavior which causes clicking on the Song “e” in the Mixer, after a move in the Track view, causes there to be empty spaces where the channel strips should be. Adding and deleting songs leaves left over channel strips in the Mixer.

That is the (default) buss for all tracks. Will check if we can avoid its inserts etc to eat space.

What you mean “persistent”?

Not here, and that would be bad. How did you duplicate the Song, and what type of tracks were in there, just audio? Just tried:

  • new project
  • create 24 audio tracks
  • right-click “Song 1”, duplicate
    All good, no?

Reorder Channels is in the making.

Tried hard to reproduce but to no avail. Do you have any sort of receipt?

Maybe the word you want (in english) is ‘recipe’ - i.e. a list of instructions to bake a cake; or here, a set of steps to reproduce a problem/bug, :slightly_smiling_face:

Perfectly right you are, sorry and thanks for poining it out!

1 Like

You’re welcome… no need to apologise.! I thought it might be some ‘Auto-translation’ or ‘Auto-correct’ mechanism getting it slightly wrong… and you’d not noticed, that’s all… :upside_down_face:

For my purposes, I’d prefer all the channel strips to be visible all the time without having to take action to make them appear. It would be ok to have a way to make them disappear. Perhaps the simple solution is to just reverse the default so they are visible unless you click on “e” or some similar button on the Song channel strip to fold/hide the channel strips.

Well, yes duplication appears to work, it creates Song 1 (D) which has the same Tracks as Song 1. However no audio is passing through the Tracks in Song 1 (D). So just to rehearse a little bit, when I made Song 1, Track 1, it was setup to receive audio from the sound card (Dante Virtual Soundcard) and I turned on the orange speaker icon and can see the meter move and hear sound on the output. (Incidentally I think the documentation still shows that as a mic icon).

Looking at Song 1 (D), Track 1 appears to be a carbon copy, with the same DVS audio channel and orange speaker icon turned on. However the meter is not moving and I don’t hear any sound. Ahhh, but now if I toggle the orange icon, then sound re-appears. So this may be some sort of copy-initialization bug.

On a second pass, I saved the project and then reopened it, selected Song 1 and there was no audio, nor with Song 1 (D). Toggling the orange speaker icon in each Song track made audio come alive for each Song… so definitely an initialization bug.

Ok, making progress. Now I can setup a complex audio mix “master” scene or song and then make a bunch of copies and make tweaks to individual songs (of course it would nice to fix the initialization issue to avoid having to toggle N-channels times M-songs speaker icons). With the simple setup described above, I can now change the audio level of the input from Song to Song and it changes as expected. So this is where I would like a crossfade and some way to make sure the audio doesn’t click. The discussion about Layers preference “Sustain Seconds” is relevant, but doesn’t seem to be active in this case as I’m using Tracks, not Layers.

One minor odd behavior when adding Tracks. The first track appears in the top position in the Tracks view. The second track appears in the second from the top position (as expected). The third track appears in the second from the top position which is in between the first track and the third track - odd. I’d expect it to appear in the third from the top position. Number 4 and all additional tracks appear in the second form top and the other tracks are pushed down… so the sequence ends up like 1,4,3,2.

Hum, some reordering seems to now be working. Using the example above of Song 1 and Song 1 (D), if I reorder in the tracks in the Tracks view and then go to the Mixer view and toggle between the two songs, the reorder has taken effect. It might be acceptable to only do reordering in the Track view, but there would still need to be a way to move Groups around… maybe groups should be visible in the Track view as well?

I’ve struggled to make this happen again. It happened once yesterday and I thought I was cross-eyed, but then it happened a second time. Unfortunately I didn’t capture a recipe. So I’m here trying stuff.

One odd things is if I add a Group to Song 1, and then toggle to Song 1 (D), the Group appears there as well, with the same name “Group 1”. If I remove it from Song 1 (D), then its also gone in Song 1. Given from what I understand of VST Live this is probably a bug. From what I’m looking for, this is not necessarily bad. But certainly it violates the sense of local/global partitioning. “Out” has the same behavior as Group in this regard.

After considerable playing, I haven’t been able to recreate the empty channel strip display of yesterday. But will let you know if it happens again.

After reviewing some other posts and thinking about this more, I’d like to propose (or add my voice to) the idea of having “global” tracks whose definition and existence transcend Songs. This appears to be what Groups and Outs are currently doing. Perhaps there could be an option on a Track to be global or there could be a whole different type, such as a (global) Channel.

With this capability, there would be a volume level (and other parameters) that are local to the Song. When transitioning from Song to Song, the crossfader would smooth out the parameter changes. No risk of clicks because the Tracks are not being changed.

I can easily think of cases where I might want to make massive changes to a Track, including re-patching inserts and sends. However, in the hardware world we work around that by using different channels (sharing the same input) and muting all but the one channel version we want in that particular scene. Certainly would be easy to do that in VST-Live as well.

Right, Groups and Outputs are global.
Will think about the Song Group xfade idea - problem is transport, because a fade is not possible when transport is jumping, and there is only one transport. But we’ll think about it.

Groups and Outs are global and visible all the time. There is also a “Visibility” item top left of Mixer. Once we have channel ordering, that should suffice, right?

…despite that Track Monitor is on - got it, that’s supposed to work, will fix.

Another possible attraction to VST Live is the integration of DMX controlled lighting (and possibly other DMX controlled gadgets). I certainly haven’t completely figured out the VST Live DMX setup, but the discussion about global audio tracks shares come commonality with DMX control. I haven’t actually hooked up any DMX fixtures to VST Live, but I did make a DMX Track in my Song 1, Song 1 (D) setup and the DMX Track is local to the Song, just like Audio Tracks. Again this seems counterintuitive. I can’t imagine changing Songs and having a sudden change in DMX lighting. Or for that matter the definition of the DMX Tracks suddenly changing. I would expect a crossfade capability and pre-determined DMX hardware address assignment.

Although when I look at the DMX mixer, nothing changes between Songs, so I’m not quite clear how the DMX mixer and DMX Tracks interact. The mixer seems to act globally, and tracks locally.

Back to the audio case for a minute, I didn’t make it clear that in live performances, it is very common to have someone talking between Songs… there are multiple ways to handle this in audio mixer land using scene memories, but in all cases its not acceptable to have glitches on the the audio channel spanning the song/scene change.

Fundamentally there is the concept of hardware resources, such as microphones, speakers and DMX fixtures and then there are software defined resources such as synthesizers and other MIDI devices. How to manage the intersection of these two worlds is interesting indeed. Ok, sometimes in the hardware world, we’ll re-patch some connections on-the-fly between songs, but that is always done while those channels are muted and usually something else is the focus of audience attention. The ability to software re-patch is really no different, other than it can be faster and easier. But we still need cross-fades and “recall-safe” channels to support smooth transitions and interludes.

@ musicullum it seems my trial license has expired. Is there a way to get it extended? I’d like to participate in the VST Live development, but I don’t think it’s right to pay for something I’m not able to use. I’d be more than happy to buy a license or two once I can productively use the software. Thanks!

@bob_k, drop me a mail. Let’s see what we can do.
m.spork ( at ) steinberg.de