I find it so aggravating that on a multichannel MIDI VSTi I have to add input channel filtering to each routed MIDI track in order to limit input to just one channel at a time. Why can’t I just select a MIDI channel in the inspector and be done with it?
Since this seems so counter intuitive for other than routing directly to a multitimbral VSTi or soundmodule, where presumably you could just as easily select ‘any’ as your MIDI channel in the inspector … I find myself asking: Is there some reason that this is sometimes an advantage?
Anyone else have either ‘illumination’ or a take on this issue?
Have you ever used a multichannel MIDI VSTi like, for an example, RapidComposer?
The channel selection will send to the selected channel from the track properly, i.e., all note data will go to the voice on the selected channel, but while the MIDI VSTi carries specific MIDI channel information, when Cubase receives this information it is not recognized. Instead Cubase will send all the information in the port on each channel to the channel selected in the track inspector.
For example, imagine the MIDI VSTi was sending on channels 2, 4 and 5. If you send the output from the track hosting the MIDI VSTi to another track sending its output on Channel 1 to a synth, the synth will receive and playback all the data from channels 2, 4 and 5 … even though you are sending on Channel 1. The same would be true if you selected any of the other 15 remaining channels.
I would like it to be that when the input comes to the synth track and Channel 1 is selected, no data comes to the synth, since no data came from the MIDI VSTi on Channel 1.
How do we know this is Cubase and not the VSTi? Because when we place an input transformer on the Cubase MIDI track to limit that track to only receiving one channel by using the filter, we now hear properly only the single channel from the VSTi in proper sync: e.g., VSTi Channel 3 is now only playing on the Cubase track that has the input filtered to only permit Channel 3.
I have never seen the need to use the input from a VSTi to control another VSTi, that’s what you want, isn’t it?
If that is about just echoing the VSTi’s input to it’s output, one could simply use a MIDI send on the sending track.
But perhaps some VSTi generate different data then their input, I do not know of any but I probably do not know many…
Anyway, I did some testing using Kontakt.
What is happening is what should happen here. Data from the sending VSTi - multiple channels - is correctly received by the receiving track, if set to Any. If not set to any, all data is redirected to the channel the track is set to. That is the absolutely correct normal behaviour. You could never play any selected track and the device it outputs to from your keyboard by just selecting a track if it were not so.
So if you only want data from one specific channel of the VSTi’s output, you’d have to set the track to Any and apply a filter.
There is just no way around it.
A MIDI VSTi sends/ outputs midi data. A regular VSTi receives midi data and produces audio. Imagine things similar to Kontakt’s script generators or Halion and Halion SE’s flexiphrases … but independent of program tools that can be used anywhere.
MIDI VSTi plugins like SugarBytes TheSys and Consequnce, 7Aliens Cantanya, Hypercylic and others send midi data to other tracks by placing the MIDI VSTi on a MIDI track and then using that MIDI VSTi as the ‘input’ on other tracks which host the audio VSTi or sound module.
The MIDI VSTis work in various way. Many require midi data ‘piped’ in via virtual audio cables from another midi track to trigger arps or sequences from the MIDI VSTi.
A few innovative ones like MusicDevelopments RapidComposer actually stream entire midi tracks processed in the MIDI VSTi. RapidComposer is a very powerful algorithmic progression and melody generator.
The behavior of Cubase I’m complaining about is simply a pain of klunky setup that doesn’t seem to have any practical purpose. There are well known work arounds, i.e., input transfomers, but they seem a time consuming and unneccessary step.
Send 5 cents to my paypal account for the tutorial.
If you fully understood what I am saying you’d see that what I request would create zero changes to how you currently have been using MIDI channel track assignments in Cubase with Kontakt or any other VSTi or sound module.
If the ‘input’ side, what I want, was changed, it would have no effect that I can see on the ‘output’ side that you are already happy with.
That’s it! If I select a specific channel in the inspector, I want it to both filter to that channel on the way in as well as send to that channel in the VSTi or sound module on the way out.
You now see why it would be a great time saver using a MIDI VSTi and have no effect on how you are accustomed to using it.
Actually, it’s not that I want a separate control, it’s that I can’t see why the output setting shouldn’t parallel the input setting.
Does a new MIDI track in Cubase even have channel assignment? I don’t think so. So, if there was no channel assignment info, the data would pass to any out channel you wanted to select. The input filter would only be in effect in the case where the input MIDI stream to the track contained channel assignment.
Anyway, how many cases can anyone think of where this would be an issue? All I can imagine is if you were trying to play a SMF and wanted to double up a track and send it to a different MIDI channel on the same multitimbral instrument … but in those cases, you’d filter out the channel assignment.
What I really think is that ‘back in the day’ there were no equivalents to Multichannel MIDI VSTi, so Steinberg never considered it an issue … even though it is their own standard by which the MIDI VSTi is now possible.
My original question in this post was whether there was some specific reason Steinberg didn’t originally implement it this way. I thought there might be some other advantage of the present design … but so far, no one has been able to suggest any important ones.
I’m still not saying there isn’t … I’d just like to know.
I can’t see why the output setting shouldn’t parallel the input setting
That is where you are missing something IMO.
Standard setup, MIDI keyboard, sending on channel one.
Project containing tracks, each sending to different channels, probably some different outputs (Hardware/VSTi)
Now simply by selecting another track the same keyboard (sending channel 1) is redirected to play the instrument that gets input from that track. This is a classical setup.
Now should your input setting mirror the output setting, our keyboard would not be received by any track not set to channel one.
Hence we’d need two seperate settings for in and out channels. In the setup I’ve sketched the default for input would have to be ‘All’.
Very clear, Jan. You’ve answered my question. As I am more likely/ frequently generating or drawing MIDI than playing it directly, I over looked this.
It still seems to me that there should be either an additional inspector track control, as you suggest, default set to ‘all’ or ‘off’ and/ or a preference option for linking of parallel midi channel input and output.
I disagree with this statement also. I’ve had multi-channel keyboards for 20 years or more. Since SX, I’ve had to use the input filter as well. It’s not a huge problem, but this was one of the big MIDI bitches I had when SX came out. IMO Cubase has taken many HUGE steps backwards in external MIDI support as they tried to advance VSTi.