When connecting an XS8 to Cubase, there are four main settings that matter in the [UTILITY] section of the Motif:
[General] > [IEEE 1394 Driver] = FW
[CONTROL] > [MIDI] > [Interface - MIDI In/Out] = mLAN
The previous two MUST be set to these values.

Okay. But let's know why you are making each of these settings. You can understand why the Yamaha Steinberg FW driver is a combination of two drivers - one for AUDIO and one for MIDI

[General] > [IEEE 1394 Driver] = FW = AUDIO
On the Motif XS - because it bridged the gap between Yamaha's original mLAN protocol and the new YSFW protocol, you must set the Motif XS to AUTO LOAD the IEEE1394 DRIVER = FW. Once this preference is set in your Motif XS's UTILITY mode. Press [STORE] and it will remain like this from this point on.This setting selects the the Yamaha Steinberg FW protocol for AUDIO communication via the FireWire connection

[CONTROL] > [MIDI] > [Interface - MIDI In/Out] = mLAN = MIDI
On the Motif XS - because it still uses the Yamaha protocol for moving MIDI data through FireWire (mLAN). This will be stored with your UTILITY mode settings as your preference, and will remain here until you change it or reset your instrument. This is your setting for MIDI communication via the FireWire Connection.

[CONTROL] > [MIDI] > [Switch - Local Control] = On/Off
[I/O] > [mLAN Monitor Setup] = stand alone / with pc / with pc (direct monitor)
These last two can be set to any of their values, but some combinations may produce a 'doubling' effect in the sound.

These last two deal with MIDI and AUDIO respectively. You set LOCAL CONTROL = OFF when you are recording MIDI data to Cubase. This disconnects the Motif XS keybed from the XS tone generator... so that MIDI messages generated by the keys, MW, PB, CS, RB, FC, knobs, etc., etc. travel OUT via MIDI (FW) to your DAW's MIDI track first. There they are rerouted to a MIDI OUT destination (which can be back in the XS or to any other MIDI device connected to your computer).
Doubling can occur because if you leave LOCAL CONTROL = ON while recording MIDI you have established TWO pathways for the signal. One is the normal press key > trigger tone engine > audio out, and the other is press key > OUT via MIDI > THRU the DAW > back to trigger the tone engine > audio out.
The former (direct) arrives first at the speakers, the latter (goes "thru" the DAW) arrives just a little late (latency) and causes the doubling. You prevent the doubling by setting the parameters appropriately. You need to know how to select what you are going to want to monitor in each situation... (mostly you opt to monitor yourself "direct" - there is no reason to be on the downside of the latent signal - unless you are expressly monitoring a VST EFFECT you have added in the DAW and you need to hear it while you add it.

The setting for "mLAN MONITOR SETUP" - are pretty clear - they deal with the audio routing and how you are listening.
_ STANDALONE when you are using the XS without a computer at all. This is normal operation

_ WITH PC when you are using VSTi routing, this prevents the XS from sending audio directly to the audio outputs.This exists because the XS is also your audio interface. It RETURNS the audio to the DAW first so that you can monitor the signal through the VSTi Folder > Motif XS VST SubFolder > Audio Lane. WHY? Okay - instead of MIDI data triggering the XS tone engine and that audio going directly to the output... when "WITH PC" is selected you are opting disconnect this direct output to the speakers, and are instead sending the audio back through FW to the Motif XS VST ... the FILE > VSTi SETUP screen setting is where you have opted to RETURN the audio to the DAW via the 8 possible stereo RETURNS. Once it arrives and is monitored through the Cubase VST Audio Lane you can process it further in the DAW before it is output (back to the XS again) this time to the audio outputs.

_ WITH PC (DIRECT MONITOR) when you are using VSTi routing and you need to hear some tracks back while adding new information (like when overdubbing).

The way doubling occurs is when you are monitoring both the Direct output and the audio via the VST Audio Lane routings together. It is highly recommended that you monitor your audio signal DIRECT. You can temporarily MUTE the audio in the VSTi AUDIO LANE... simply click on the "M" Mute button on the AUDIO LANE to Mute this audio while monitoring yourself direct.

The diagram shows a very good understanding of what is going on. You have it completely nailed for MIDI data. Particularly the placement of the LOCAL CONTROL ON/OFF switch... You've got it.

Now if you place a switch very much that will play the same roll for audio... at The mLAN MONITOR SETUP can be understood in the same way as the Local Control ON/OFF switch - albeit it is for audio routing while Local Control is for MIDI routing. In that circle you use to denote the mLAN MONITOR SETUP... understand it as a switch like you do for LOCAL CONTROL... when...

STANDALONE is selected the line goes Directly from the MOTIF SOUND ENGINE to the ANALOG AUDIO OUTPUT

WITH PC is selected the Switch would be shown OPEN... "with PC" breaks the connection between the 'dot' and the circle labeled "mLAN MONITOR SETUP"... so audio signal must go via "AUDIO OUT TO COMPUTER' path via mLAN and must return via " AUDIO IN FROM COMPUTER via mLAN" to reach the ANALOG AUDIO OUTPUT. (When "with PC" is selected there is no (direct) connection between the "dot" and "mLAN Monitor Setup" circle.)

WITH PC (DIRECT MONITOR) both audio RETURN pathways are allowed because you will be monitoring some tracks from the DAW (all that are set to playback) and you will be monitoring yourself (sending new audio data) direct. This is the option when you are overdubbing... naturally: you need to hear previously recorded audio in Cubase, while you add your new Track live. Only your live playing will be routed to the computer via mLAN, all returning audio from the DAW will go to the analog audio outputs.

it's: one, the other and both
it's: "standalone", "with PC", and "with PC (Direct Monitor)".

Excellent!!

I really could not have expected a reply before Monday - this is above and beyond the call of duty - thank you.

Ok - that's all very clear. So let me go a step further and add a standard Audio track (NOT a VST instrument). Please see the attached image, where in addition to adding a stereo Audio track I have also added:

- some detail to the Motif for accuracy, like the sequencer and the A-D converter
- illustrations of the three states of the mLAN Monitor Setup to clarify how my pictogram corresponds to reality
- detail to Cubase: I/O Ports and Busses

I hope this can be useful to someone else in the future.

As before - confirmation (or otherwise) that I've captured the essence of how things work will be greatly appreciated.

Especially since experimenting directly on the Motif in the [UTILITY] menus, as I change the settings of the various parameters and switches, everything works as you'd expect from the diagram - Except for one thing:

I set up with the Audio and MIDI tracks connected to their respective In/Out ports and busses, and the FW driver auto-loaded & MIDI In/Out set to mLAN.

Then I set Local Control = On and mLAN Monitor = stand alone, and disconnect both MIDI and Audio tracks outputs, and play the keyboard. I get a nice direct, single, sound, as expected. MIDI directly from the keybed, the MIDI from the MIDI track is cut off from the track output disconnection, and any audio from the Audio track is cut off both by the track output disconnection and the mLAN Monitor setting = stand alone.

Hit record in this setup, play the keyboard, and Cubase records the MIDI. Perfect.

Now connect the MIDI track output to "MOTIF XS8 Main Out", hit Play in Cubase, and I get a nice single sound playback - with the MIDI data being supplied by the Cubase MIDI track, generating sound through the Motif Sound Engine, and monitoring it directly. OK.

Next - Still playing MIDI track in Cubase - switch mLAN monitor to "with pc", and connect the Audio track output to the Stereo Out bus. Still playing MIDI track in Cubase. As expected, I get a single audio rendition of the MIDI data. The MIDI data is being supplied by the Cubase MIDI track, to the Motif sound engine, back to Cubase on the mLAN to the Audio track, then back out to the Motif on the mLAN, and out to headphones. The sound engine is receiving MIDI data only from Cubase because even though Local control = On, the keybed is not being played, so there's only one MIDI stream. And I'm hearing only a single sound because the direct sound from the sound engine is being cut of by the mLAN Monitor Setup = 'wtih pc", which passdes only the audio signal from Cubase. Excellent.

If I now switch the mLAN Monitor to "with PC (Direct Monitor)", as you'd expect I get a doubled sound. I have purposely set the YSFW Driver to a high buffer size so that if there is any sound doubling, it will be clear.

BUT - now I switch the mLAN monitor to "stand alone". (Still playing MIDI track in Cubase. ) And lo and behold, I still hear a doubled sound with the expected significant latency on one. But I shouldn't, I don't think. With mLAN Monitor set to "stand alone", I should only hear the sound coming directly from the sound engine. To double check, I mute the Audio track in Cubase - yep, suddenly just a single pure sound.

Hmmm - so I stop playback, disconnect the MIDI track ouptut just to be sure, un-mute the Audio track, and play on the keybed - still the doubled sound.

So - everything works as expected, EXCEPT that mLAN Monitor = 'stand alone' produces a doubled sound corresponding to the two audio signals - the direct one and the one from Cubase.

Looks like either I've misunderstood something, or my Motif has an issue?

Thank you again.

Where to begin...?

First would be: No, your Motif XS does not have an issue... It is simply a misunderstanding that selecting one thing does not prevent another. (hopefully that statement will make sense on the other side of what follows):

In the diagram you made - here are few corrections.

Correction 1: In the box that represents the Motif XS8 hardware (left) the "Sequencer" should be placed between the Keybed/Controller and the Motif XS Sound Engine (as it is) but is should be post the Local Control ON/OFF switch. Think about it - if you turn Local Control = OFF, you will not be able to record to the internal sequencer. That is your first clue that it is after the Local Control Switch. Also, placing it where you did you would not be able to record to the XS's sequencer from an external MIDI device... your second clue that it is post that external MIDI IN...

It is always a good idea to place the sequencer between the Controller and the Tone Engine - because the *active* MIDI Track determines which of the Tone Engines PARTS will respond.

Correction 2: the "Interface MIDI In/Out" selection - when you make a selection here (either USB, MIDI, or mLAN)... if you select USB or mLAN you have selected a type of MIDI connection that can be multiple ports simultaneously. This retasks the 5-pin MIDI jacks to be "Motif XS8-3" for USB or "Motif XS8 MIDI IN" and "Motif XS8 MIDI OUT"

Correction 3: the Motif XS8 Remote In/Out are not reserved for the Motif Editor VST but are reserved for the DAW REMOTE mode.

Digital Audio Out to Computer via mLAN: the digital audio Outputs going to the computer via mLAN consist of 16 buses (configured as Main L/R plus 14 assignable mLAN outputs. The assignable digital outputs can be configured in 7 odd/even stereo pairs or as individual mono buses, as you may require, m1-m14.

Digital Audio In from Computer via mLAN: the digital audio Inputs coming from the computer via mLAN consist of 6 buses (configured in three stereo pairs):
Motif XS8 Main L and Motif XS8 Main R
Motif XS8 Asgn L and Motif XS8 Asgn R
Motif XS8 mLAN L and Motif XS8 mLAN R

As to your reasoning - which sometimes is spot on but sometimes you are concluding the wrong things based on the settings.

Local Control = ON/OFF is a MIDI setting - and although it has an impact on whether or not you hear audio. It is a setting used during setup for external *MIDI* Recording. During any MIDI recording to any external sequencer you would shut LOCAL CONTROL = OFF - thus disconnecting the Controller from the internal Tone Engine so that MIDI data traverses the *active MIDI Track* in the sequencer. That *ACTIVE* MIDI Track is responsible for which Tone Engine (port), and which specific Part (channel) will respond (generate audio).

"with PC" is an audio setting to determine where that AUDIO goes, and whether you are monitoring the normal path of audio from the XS Tone Engine to the DA converter and main Outputs. If you select "with PC" in a similar fashion to how Local Control 'breaks' the normal connection for MIDI signal flow, you are breaking the normal connection for audio signal flow from the Tone Engine to the main analog Outputs. Instead you are routing the signal BACK to the DAW via the digital mLAN connection where it arrives in Cubase at the VSTi Audio Lane (virtual audio). The Audio Lane in Cubase will send the audio to the audio interface (yes, that is the Motif XS8).

And that is the confusion - if the XS8 was not also the *audio interface* this would not be confusing. So to sort it out you have to be clear about when the XS8 is a synthesizer and when it is acting as an audio interface. A separate audio interface doesn't generate audio itself, so you don't expect anything. But if you think about the audio interface as a separate it becomes some what clearer.

You are, as best I can tell, a bit confused because you think making that switching back to "Stand alone" should undo something about the digital audio routing and it doesn't. And it should not... you will need to accept that it does not. Then you will be able to deal with it.

There are three scenarios - best understood simply as:

Using the Motif XS8 all by itself - no computer connected at all = "Stand alone" audio travels to the main outputs as expected
Using the Motif XS8 to record MIDI data and monitor thru the computer = "with PC" audio does not go directly to the main outputs, it goes to DAW first.
Using the Motif XS8 to record MIDI data and monitor playback from the computer = "with PC (Direct Monitor)" audio is routed through both audio paths.

However, what you are playing is selected by that "active" MIDI Track. That is why I kept accenting *active*. the active (currently selected) MIDI Track's data will be echoed back to the XS tone engine and its audio will NOT go to the main OUTPUTS... the audio of data originating as MIDI on the active track RETURNS to the VSTi Audio Lane. But also since Direct Monitor is available - the sound goes also directly to the output. It is a situation where YES, you are going to be able to hear BOTH... you need to choose.

You say "With mLAN Monitor set to "stand alone", I should only hear the sound coming directly from the sound engine" - that is not true. What would be a true statement is: and appreciate the difference

"with mLAN Monitor set to "stand alone", I can hear my key presses traveling direct to the Motif XS8 Tone Engine and that Tone Engine will Output data directly to the main L/R Outputs". You cannot conclude that you should not hear the signal that travels to the computer and is returned digitally from the computer. The "Stand alone" setting simply assures the normal routing for AUDIO.

It does not, however, prevent the Motif XS8 from broadcasting audio via the mLAN FW bus, and that is your (incorrect) assumption. Not sure why you think that. It ensures the routing of the MIDI data to the synth and the synth's audio to the main output (as normal - think: nothing else connected = stand alone). But if you have a FW cable connected and a computer set to receive audio via the connection, simply setting "stand alone" does not defeat the XS from sending the signal. "Stand alone" as your setting does not DEFEAT anything from traveling out the digital outputs. Audio always goes out the digital outputs. If you are using the XS "stand alone" it assumes you are serious about it standing alone... but if you connect something to the digital outputs - that signal is still there - to prevent signal from going out the digital output, you simply have to not "listen" to it.

Just like the headphone jack is outputting signal at all times - even if you do not plug in a set of phones... you can imagine that signal is still available at that output source. If you don't want to hear it, don't listen 🙂

Audio is not defeated from the mLAN/1394 bus when you select STAND ALONE. And the way to simply Not Listen - is to click the "M" on the VST INSTRUMENT Audio Lane, which you are recording your data.

I most often select to work with the VST Routing. Which means LOCAL CONTROL = OFF, but I'm set "with PC". But I opt not to monitor the audio coming via the VST Folder's audio lane (I elect to MUTE that lane by clicking the "m" button to MUTE the returning latent audio).

I monitor myself using "with PC (Direct Monitor)" - I mute the latent audio in the VSTi Audio Lane so that I can monitor what I play "direct". The goal is to never stand downwind of the latent signal. Latency should never be a problem for a person with an external VST. Because you can choose to hear your hardware direct during the record process. The only time you might ever need to "stand downwind" of the latent signal (monitor the Audio Lane during Record) is if you decide to process the signal with a computer-based VST EFFECT (while you are recording). Normally, you know you will be at the mercy of processing the signal post the computer initially receiving it, then processing it, then sending to back out. This is why most software based Effect processing is done on the already-recorded signals. (When you buy into computers as your studio, you buy into this "fix it in the mix" concept as a work ideology.

With hardware, and this will be an advantage for external hardware, is you can choose to monitor yourself "direct" - if you know where to "monitor" and where to "mute". 🙂

Hope that helps. Good job though - it is complex, yes, we know. Audio engineering is not necessarily the easiest thing to grok but once you understand SIGNAL FLOW... you are well on your way to mastery.

Once again, thank you for taking time and sticking with me on this.

Determined to get this crystal clear for all of humanity now - testing on the Motif/Cubase along with reading your reply, I've made the changes to the Local Control that appear to best reflect the operation I've observed. In essence, Local Control appears to have no effect whatsoever on playback directly from the sequencer, or on the ability to record in Cubase from the sequencer, or on the ability to pass MIDI from the sequencer via Cubase back to the sound engine. It appears to simply be an on/off switch for the MIDI stream directly from the keybed to the sound engine and TO the sequencer- as I've illustrated, attached.

Thus - Local Control = Off does not affect keybed MIDI output to the DAW, does not affect playback from teh sequencer to the sound engine, and does not affect sequencer output to the DAW. It only affects the ability to play the keybed directly to the sound engine, and to record the keybed to the sequencer.

I think that's got that - subject to your confirmation .

The main other sticking point is the mLan Monitor.

I do so hate to seem thick - but, in short - I can find absolutely no functional difference between "stand alone" mode and "with pc (direct monitor)" mode. I mean that both in terms of what I read about it, and the behaviour I observe on my system.

OK, we have Cubase running and the Motif connected to it by mLAN/FW. in Cubase I have a regular Audio track, and it's connected to the input and output stereo busses. Record is armed and Monitor is enabled on this track. It is the 'active' audio track. The Motif is set to Local Control = On and I'm playing the keybed. I can see the audio input meter jumping on the audio track, confirming that it's receiving an audio signal from the Motif when I play the keybed.

With mLAN Monitor = "standalone" - I hear the doubled sound. Clearly audio from the Sound Engine is going directly to the phones jack, but also to the computer and being returned from the computer and also to the phones jack.

With mLAN Monitor = "with pc (direct monitor)" - I get the exact same result. Clearly audio from the Sound Engine is going directly to the phones jack, but also to the computer and being returned from the computer and also to the phones jack.

In either case, I can mute the sent/returned signal by muting the audio track in cubase, or disconnecting the track from its input or output bus. But its the same in either case.

The only setting that seems to make any difference whatsoever is "with pc" - which cuts the direct signal. I've included this in the new diagram as a simple on/off switch on the direct audio signal.

From a functional perspective - it's as if "stand alone" and "with pc (direct monitor)" were the same setting just with two different names. As far as I can tell, the two settings route signal exactly the same way. The only way to make a difference is within Cubase.

Just in case the significant difference was using the VST instead of a regular Audio track, I tried that too - same result.

Is there any functional difference that a user should see that you can indicate to me? Or even an internal difference that matters but won't be seen functionally by a user? Or something that differentiates these two settings in any practical sense?

And thank you for your indulgence.

I do so hate to seem thick - but, in short - I can find absolutely no functional difference between "stand alone" mode and "with pc (direct monitor)" mode. I mean that both in terms of what I read about it, and the behaviour I observe on my system.

I know, this can be confusing, at first.

I think you understand that "Stand Alone" allows the XS audio to go directly to the output. Everybody's XS does that. It is meant for when you use the instrument by itself, sans computer. Audio simply goes via the main stereo outputs to your sound system. We don't care about what else it does because it is standing alone.

I think you understand that "with PC" is designed to disconnect XS audio from traveling directly to the main LR outputs, it directs audio to the DAW where it arrives as virtual audio (audio is heard via the Virtual Instrument's Audio Lane create for each output bus). An audio version of what Local Control is for MIDI. The audio is redirected initial through the computer. And then sent to the audio interface and out. This signal is always the latent signal.

"With PC(Direct Monitor)" is the third switch. The third AUDIO option. It does what "stand alone" does in that it allows the sound that you are playing to be heard directly to the main (analog) LR outputs... All other audio follows the setting of "with PC" which you already understand. You need this third switch (selection) so that you can OVERDUB and take advantage of the multiple audio (assignable/mLAN) outputs, that unique situation where you are playing some tracks back while you are adding just one new PART, and the one you are playing needs to be routed discreetly.

The object is to never be downwind of the latency when YOU are actively playing/recording. Small that it may be... "With PC" latency... "direct Monitor" zero latency.

Playing along with the track but hearing yourself direct. Don't move on until the need for this is clear. Of course this is what you want, and that's why it exists. Only the currently active Part (the one you are playing) goes Direct to the main audio outputs...of course it also travels to the computer to be recorded, but you do not listen (monitor) the path going to the computer - you must 'mute' that audio. What you are playing goes two places: Direct to the speaker outputs, and to the Cubase audio track destination.

You only monitor what you are playing Direct, the audio that travels digitally to Cubase remains unheard (muted), until you playback. "With PC(Direct Monitor)" gives you an option that will prevent the digital (recorded) signal from returning (to the audio interface) during the process, as you can imagine: it would be late.

What Delay Compensation does in any modern DAW, is place the overdubbed audio precisely where it belongs time wise. It literally compensates for your computer's inability to chew gum and walk. Er, playback audio and record audio simultaneously, precisely.

Final note, and this will help you see it clearly: So far you have been touching the elephant's tail and drawing some conclusions based on a partial picture - and this is an okay methodology, but only takes you so far. And I salute your digging in to grok the whole thing... here is what I think will start to bring it into focus for you:

So far you have only been experimenting and considering only the main Stereo L/R Outputs. Oh, I know you have mentioned the 14 assignable outputs in your diagram - but try assigning one of the PARTS to one of those m1-m14 buses - Assign the PART you are playing, for example, to "m1&2" as a stereo pair. You will want to monitor yourself direct (alone) - this assignment, literally, removes this PART audio from the System Stereo mix, while all the other PARTS playback through the stereo bus. When you assign a PART to an assignable mLAN/FW output you will immediate notice it is removed from the Stereo mix... (until now you have only been monitoring/comparing/analyzing the stereo mix...)

The "withPC(Direct Monitor) audio routing scenario behaves very much like the Stand Alone until you consider the assignable (mLAN) OUTPUTS. This is where the difference will become clear for you. It allows you to monitor (direct) Parts that have been removed from the stereo mix (that is, Parts that have been assigned to an assignable mLAN/FW output. (See the need?)

When you create the Motif XS VST notice that Cubase creates a "VST INSTRUMENT" Folder (if one does not already exist)... it creates a subFolder for the "Motif XS VST" within the VST INSTRUMENT folder... it creates an AUTOMATION LANE and a signal stereo AUDIO LANE... it asks you if you would like to create a MIDI Track for this Plugin. And it assigns the MIDI OUT to "Motif XS VST-Midi In"

So it gives you one stereo Audio Lane, and just one MIDI Track, to start you off.
You probably realize that you can create additional MIDI Tracks, one for each PART you want to record with the XS.
But remember, you have 8 stereo output pairs... of which the first two (Main L and Main R) are the default.

When you decide you want to record a piano overdub and you want to isolate it to its own assignable mLAN/FW bus output - this is when you assign it to its own discreet output, say "m1/m2" And this is when the STAND ALONE setting will differ greatly from the WITH PC(DIRECT MONITOR) setting. Try it.

When you assign a PART via the OUTPUT SELECT parameter in the VST Editor, you will need to activate that OUTPUT from the XS. By default only the Main L/R Stereo output is activated and you see just one AUDIO LANE - a new AUDIO LANE will appear for each OUTPUT you activate.

On the Editor VST - in the extreme upper right corner is a small chevron (circled in red below) - click on that to see the drop down > click on ACTIVATE OUTPUTS > Activate the second STEREO PAIR of XS outputs for use.

Notice an additional AUDIO LANE appears in your MOTIF XS VST folder (second screenshot)... "Motif XS 1" = your main Stereo Out. "Motif XS 2" = m1&2 and notice more importantly the difference between STANDALONE and WITH PC (DIRECT MONITOR)>

Let us know.

"With PC(Direct Monitor)" is the third switch. the third AUDIO option. It does what "stand alone" does in that it allows the sound that you are playing to be heard directly to the main (analog) LR outputs... All other audio follows the setting of "with PC" which you already understand.....

You only monitor what you are playing Direct, the audio that travels digitally to Cubase remains unheard, until you playback.

"With PC(Direct Monitor)" prevents the digital (recorded) signal from returning (to the audio interface) during the process

"

Except it doesn't.

OK - On the Motif, Local Control = On so I can play the keybed. I set mLAN Monitor to "stand alone". I have headphones plugged into the headphone jack of the Motif. I can hear what I play on the keys.

The Motif is connected to the PC via FW. I open Cubase 7 and open a new, empty, project. I add an Audio track - not a VST instrument or a MIDI track - a regular stereo Audio track. It gets created connected by default to the Stereo In and Stereo Out busses. I arm the track for recording and turn the track Monitor On. When I play the keybed, I can hear the direct signal, and the latent sent/returned signal - as expected since mLAN Monitor is still set to "stand alone", which broadcasts the audio signal both directly and to the PC and does not block the signal returned from the pc.

So I switch to "with pc (direct monitor)". I do this directly on the Motif, via the [Utility] dialogs - not with the VST Editor, which I have not loaded or started. I expect that, as you say, "With PC(Direct Monitor)" prevents the digital (recorded) signal from returning". That is, I expect to hear only the direct signal of what I'm playing on the keybed. But I don't. I still hear both the direct and the latent signals.

Sure, I can fix that by turning off the track Monitor in Cubase, or by muting the track, or by disconnecting the track from the Stereo Out bus - but then that's Cubase preventing the digital audio signal from returning - not the Motif.

So maybe it's the fact of making the settings on the Motif - maybe it only works when they're done via the Motif XS VST Editor. So I delete the Audio track, and load the VST Editor into the Instrument Rack, and say "yes" to creating an associated MIDI track. The VST Editor opens defaulted to "with PC", and when I play the keybed I hear only the returned audio. (I know it's the returned signal 'cause I've set a large buffer size causing a significant delay). I click the "PC Drct" button - and now I hear a doubled signal - both the direct and the latent sent/returned - just like doing the settings manually. On the VST editor, under MIDI (bottom left corner), I click "Rec on PC", which turns Local Control Off, just to be sure what I'm hearing is not simply doubled MIDI.

Mr. M - what can I say - I hate to abuse of your patience like this - I'm sure you have more important things to do - but know that your help is immensely appreciated.

In the meantime - on to those assignable channels....

.....The Motif is connected to the PC via FW. I open Cubase 7 and open a new, empty, project. I add an Audio track - not a VST instrument or a MIDI track - a regular stereo Audio track. It gets created connected by default to the Stereo In and Stereo Out busses. I arm the track for recording and turn the track Monitor On. When I play the keybed, I can hear the direct signal, and the latent sent/returned signal - as expected since mLAN Monitor is still set to "stand alone", which broadcasts the audio signal both directly and to the PC and does not block the signal returned from the pc.

So I switch to "with pc (direct monitor)". I do this directly on the Motif, via the [Utility] dialogs - not with the VST Editor, which I have not loaded or started. I expect that, as you say, "With PC(Direct Monitor)" prevents the digital (recorded) signal from returning". That is, I expect to hear only the direct signal of what I'm playing on the keybed. But I don't. I still hear both the direct and the latent signals.

Okay, no problem. The best way to learn what happens when you involve the VSTi Routing is to understand the routing without it. I follow your thinking... now please follow mine... I assume you are playing PART 1 of a SONG Mixing or PATTERN Mixing setup. Using the parameters of your experiment as listed above:

Please press [MIXING]
Press [EDIT]
Press [1] to select PART [1] parameter
Press [F1] VOICE
Press [SF2] OUTPUT
Set the OUTPUT SELECT parameter = m1&2
This removes the PART from the STEREO BUS (please recognize, again, you have been listening to the Stereo bus) by assigning what you are playing to a discreet output, how are you now going to monitor yourself?)

Now check the difference between "Standalone" and "withPC(DirectMonitor)"

In order to monitor yourself, the "withPC(DirectMonitor)" prevents the audio of the signal your are playing from returning to the audio interface after it has traveled through Cubase. You will hear just yourself "direct". The reason to exist is now clearly understood.

Adding MIDI to the equation does complicate it, but only because the MIDI data, typically sent from the keyboard with LOCAL CONTROL = OFF (or from a MIDI Track) to the DAW, is then sent back to the synth, which generates audio... then the routing is as explained above for AUDIO signal.
The "Standalone", the "withPC" and the "withPC(DirectMonitor)" settings all deal with audio routing.

In any recording situation - there are three main functions you will be performing: Record, Playback(mix), Overdub.
Record can be a situation where all channels are coming IN
Playback (mix) is a situation where all channels are going OUT
Overdub combines some of both... you are Recording new signal while Playing back pre-recorded material.

The audio interface being built into your keyboard does make this one step more complex to think about than an external interface. But the designer's thought about all the situations for both MIDI and/or AUDIO recording that you might ever encounter. The QUICK Sequencer Setups in the XS (are those that will be the most often used) - the operating system leaves ways to customize things as well, as you find necessary.

I assume you are playing PART 1 of a SONG Mixing or PATTERN Mixing setup.

No - for the purposes of clarifying this, I've stripped it down to the minimum. I'm just trying to play a voice into a track in Cubase. A single regular empty Audio track in Cubase, a single Voice on the Motif in Voice mode, & the mLAN. There are no other tracks playng back at the same time. There are no other tracks at all.

The Motif is in Voice mode, and I just want to record what I play in a track in Cubase via Firewire, but without hearing the latent audio signal as I play. It seemed "With pc (direct monitor)" was deisgned for this purpose - but it does not block the returning latent signal. The only way I can avoid hearing the latent returned signal is to do something in Cubase.

But - I think I've belaboured this point quite long enough. There are other ways to get the desired result, and much more important issues to deal with before exhausting your kindness. This will just be one of those things that aren't crystal clear, but I can work around, and maybe one day the light bulb will go off. I will, however, look more deeply into th emulti-channel arrangement.

Thanks again. Hope this helps someone else some day.

Here is where your scientific method fell apart:

"I'm just trying to play a voice into a track in Cubase."
and
The fact that you connect to the computer in "stand alone" mode, which is an oxymoron. (and draw incomplete conclusions).

Both are outside the realm of what you need to do - particularly when attempting to understand this routing issue. The first is a fatal flaw (VOICE Mode, since you cannot select assignable outputs in this mode, period - and the difference between the two setting is based on being able to MONITOR the Assignable mLAN/FW outputs when in an OVERDUB situation)... I can understand, for an experimental point of view, why you might try to see what happens in "stand alone" connected to a computer - as long as you know it is an illogical setting. For it does not matter what comes out of the digital outputs when you are not connected. Sort of like headphone jack when no headphones are connected.

You cannot record using any of the Assignable mLAN/FW outputs from the VOICE mode. PERIOD. OUTPUT SELECT is "PART" parameter - not a VOICE mode parameter. So this is moot. Completely!!! A "VOICE" must use the SYSTEM outputs (that is, the main L&R outputs). A "PART" can be routed to any of the available audio ports.

You are looking for a difference between it ("stand alone" and "with PC (Direct Monitor)") and you will never encounter the situation where they are different given these first two premises you are working under. It's like concluding the elephant is skinny based on just touching and analyzing the tail. You can't even begin to see the difference because you are in the Wrong MODE with the Wrong settings to see the difference. It is for a specific USE CASE.

You did not try what I wrote or you would understand this by now... Come back to it at a later time. It will help others (and already has, thank you for the question). Please, try it, when using the Motif XS with a SEQUENCER - it is completely logical to use it in one of the two SEQUENCER modes: SONG or PATTERN - this is true whether you are using an internal or external sequencer.

This is where you can effectively use the MULTI-TIMBRAL and MULTI-MIDI CHANNEL functions of the instrument. This is why it is labeld SEQUENCER - MULTI PART on the front panel. In VOICE mode you cannot take advantage of the multiple audio port outputs. In VOICE mode you cannot take advantage of the multiple audio port outputs! In VOICE you cannot route signal to any assignable output. Only in the modes where Voices are placed in "PARTS" can you route signal to an assignable output.

For the record: Of course "stand alone" mode looks and acts just like "With PC(DirectMonitor) when you are using only the STEREO OUTPUTS. In "stand alone" mode you are using just the Stereo Outputs (you are not using the mLAN/FW outputs at all, you CANNOT). Just like it is when you are only using VOICE mode - you cannot use the mLAN/FW outputs. Remember assigning a PART to an assignable output REMOVES it from the Stereo (system) Outputs!. Assigning a PART to an assignable output REMOVES it completely from the STEREO (system) Outputs. And it travels via FW to the computer. (this is it in a nutshell).

When you are using the Motif XS with a sequencer - in a sequencer mode - and you are using the assignable mLAN/FW output, the "Stand alone" setting is indeed VERY DIFFERENT from the "with PC(Direct Monitor)". You cannot see this difference unless you are in SONG, PATTERN, and are using assignable outputs.

The designers have created a scenario for all the record, playback, mixdown, and overdub tasks you might need to do. Trust me.

A true professional ethos - keeps encouraging success even when the knucklebrain on the other side is ready to give up.

OK - I will read, digest, and give 'er another try.

Thanks.

Mr. M,

Thank you for urging me to heed that critical advice to try this out in Song Mode.

I've spent a couplw of days working with it, and getting familiar with it - also with the help of the Editor VST user's manual. I am able to set it up and route signals through the various channels, and it feels more and more like I'm getting a handle on it.

One function continues to elude me though: mLAN Monitor, "PC Direct".

I have made a series of screen shots, attached, so you can see exactly what I'm doing.

Before the images - the basics: Firewire connected, start computer, start Motif, start Cubase, select YSFW Driver when prompted. I have set the driver to a high sample count to induce large latency so a doubled signal will be easily identifiable.

Then, image by image:

01_rack: load the Motif XS VST Editor into the instrument rack; it asks to create a MIDI track associated with it, and we click "create". A MIDI track, "MOTIF XS VST 01" is created, along with a "VST Instruments" folder which contains the "Motif XS VST" instrument, with an automation lane and an audio lane, "MOTIF XS 1", corresponding to the MOTIF XS 1 channel. The audio lane is created with it's output bus set to Stereo Out. The VST Editor has been previously set to Auto-sync, so it starts "online". Switch the VST Editor to Song Mode - it re-syncs.

02_VST_Editor_and_channels: Activate output channels 1 through 3. Two new audio lanes are automatically added to the MOTIF XS VST instrument - MTOIF XS 2 & 3. They are also created with their output bus set to Stereo Out. In Quick Setup - set MIDI = "Rec on PC", and mLAN Monitor = "PC Drct". Under Output Selection, set Part 1 = L&R, Part 2 = m1&2, Part 3 = m3&4.

03_VSTi_setup: Ensure ports are set correctly. As done, the VSTi Setup ultimately routes Track 1 to DAW channel MOTIF XS 1 (the audio lane), and future Track 2 to MOTIF XS 2, and Track 3 to MOTIF XS 3 audio lanes.

04_setup: Ensure Data Port = mLAN, mLAN Device Motif XS8, and MIDI Out = Motif XS8 Main

05_cubase_midi_1: Create two more MIDI tracks and entitle them "MOTIF XS VST 02" and "MOTIF XS VST 03". This image shows MIDI track 01 connected to MOTIF XS8 Main for MIDI-In, to the VST Editor input for MIDI out via MIDI cahnnel 1.

06_cubase_midi_2: shows the setup parameters for the MIDI 2 track - same, except output on MIDI channel 2.

07_cubase_midi_7: shows the setup parameters for the MIDI 3 track - same, except output on MIDI channel 3.

08_cubase_vst_audio_lane_1: shows the settings for audio lane 1 - output connected to STereo out, as it was created.

09_cubase_vst_audio_lane_2: shows the settings for audio lane 2 - output connected to STereo out, as it was created.

10_cubase_vst_audio_lane_3: shows the settings for audio lane 3 - output connected to STereo out, as it was created.

With all this, I record something simple into MIDI tracks 1 & 2. So now when I go to record a new track - MIDI Track 3 - I have an existing track routed to MAin L&R, Track 2 routed to m1&2 (taken out of the main channel), and the new track 3 routed to its own channel.

OK - here's the thing I don't get:

I select Track 3, which arms it for recording, and I hit Record. Tracks 1 & 2 start playing back - clearly doubled. When I play on the keyboard at the same time to overdub the new track, I hear my performance - also clearly doubled. We're still set to "PC Direct".

I can cure the doubling by switching to "With PC" - but, of course, I'm getting just the latent signal.

The only way I can find to get a non-latent, non-doubled signal for both the played-back tracks and the new track during recording, is to disconnect all the audio lanes from the Stereo Out bus in Cubase via the Output Routing of the Track Inspector, and use "PC Drct" mode. But, in this case, it's simply Cubase that's preventing return of the latent signal by disconnecting its source. This is how I recorded tracks 1 & 2. But that means, for example, that all my audio is cut off from the Stereo Out bus. I can even de-activate all the Output Channels and it makes no difference in this arrangement. And of course it can't be exported because there's no signal on the Stereo Out bus.

What am I doing wrong?

Thanks.

OR - AM I doing anything wrong at all?

In the light of a new day, I tried just going with it the way it works. Maybe that's how it's supposed to work?

First - it seems that in fact all you really need is one non-Main L/R channel - for the new tack being overdubbed. All other existing tracks can be routed to Main L/R.

So starting with yesterday's project - I routed Tracks 1 & 2 to Main L/R (channel 1 = "MOTIF XS 1" audio lane) and disconnected both audio tracks in Cubase from the Stereo Out bus (track output routing = "No Bus"), and set the VST Editor to "PC Drct".

This way, I hear only one version of the playback & of any new performance - in sync.

So, I hit Record, and can record a new track in perfec syncronization with the first two. When I'm done, I set this third track's outpt to Main L/R in the VSt Editor as well. Now I have three well synchronized MIDI tracks, all outputting to the main Stereo out bus of the Motif (the Firewire Main L/R channels).

Then - in Cubase I re-connect the MOTIF XS 1 audio lane to the Stereo Out bus, and solo MIDI track 1. Then, go to file > Export > Audio Mixdown. Select "Stereo Out" as the output channel, and in "Import into Project" select the Pool and Audio Track options, and hit Export. It lets me know that with the Motif export can only be done in real time - ok, fine. At the end, I have the audio of MIDI track 1 in a new Audio Track in Cubase - a faithful, digital, copy of the MIDI performance. Repeat for each track.

Now I have audio tracks transferred in pristine, noise-free, conversion-free manner, ready for standard audio processing in Cubase. Mind you, i fI decide they need any editing in terms of notes - it basically has to be thrown away and re-export the MIDI track. Not a deal breaker.

But is THIS the way it's meant to be used?

Thanks.

It is always difficult to discern - the "okay, what am I doing wrong..." situation.

I can tell you that what you missed the first time around is still what you are missing now. You need to appreciate the difference between the Main Stereo Out and the mLAN/FW assignable outputs. The main Stereo Output functions two ways... as the main analog output of the Motif XS synthesizer and it functions as the main analog output for the Audio Interface.
When is it doing each function?
It is the Motif XS synthesizer output when you are routing signal DIRECT to that output.
It is the Audio Interface output when you are routing signal from Cubase to that output.

02_VST_Editor_and_channels: Activate output channels 1 through 3. Two new audio lanes are automatically added to the MOTIF XS VST instrument - MTOIF XS 2 & 3. They are also created with their output bus set to Stereo Out. In Quick Setup - set MIDI = "Rec on PC", and mLAN Monitor = "PC Drct". Under Output Selection, set Part 1 = L&R, Part 2 = m1&2, Part 3 = m3&4.

By setting PART 1 = L&R you realize that this PART is using the Main Stereo Out and therefore is different from the others. Even if you are only as a casual observer, you can recognize that it is therefore going to behave differently in some situations from the other two which are assigned to assignable mLAN outputs. How it behaves differently is that it is the default for all PARTS and therefore there is no isolation of the data there (initially all PARTS are assigned there). You can anticipate that other PARTS will share this output bus.

03_VSTi_setup: Ensure ports are set correctly. As done, the VSTi Setup ultimately routes Track 1 to DAW channel MOTIF XS 1 (the audio lane), and future Track 2 to MOTIF XS 2, and Track 3 to MOTIF XS 3 audio lanes.

Tracks 1, 2 and 3 are MIDI data, the MIDI data is sent through the Motif XS VST, then to the Motif XS tone engine and it triggers audio.

What you are trying to understand is: what happens to that audio. In all scenarios with recording MIDI data - it is assumed LOCAL CONTROL = OFF.
In all scenarios with recording MIDI data - it is assumed that LOCAL CONTROL = OFF.

In all cases the audio is sent both to the Motif XS main analog L&R outputs, and simultaneously it is sent back to Cubase to the VST Instrument Audio Lane (where you can choose to monitor it or not). If you choose to monitor it there - then you are choosing to monitor through the PC (with PC). If you choose not to monitor it there, then you must MUTE that audio return. The key to all of this is to know, definitively, which path you want to monitor, and where you can prevent the listening to the doubling. It can always exist because the signal is routed on two paths.

Here is your checklist
In the case of audio generated in response to a MIDI Track's data you can anticipate the following reaction:
_ When you select "with PC" you will only be able to hear the Track's data through Cubase and the VST Audio Lane. This will always be true. (Local Control = OFF).
_ When you select "PC Drct" - the audio signal will be sent to both places. Of course, both places. And you can select to hear *just* the DIRECT signal by clicking the "M" (Mute) button on the appropriate AUDIO LANE to prevent the latent signal from arriving at the audio interface. This is how you can choose to monitor *just* the Direct signal for the data generated by the MIDI data on a MIDI Track.

So when you are overdub recording, that is, you are adding a new MIDI Track using the VSTi routing... you will select "PC DRCT" so that you can hear what you are doing direct. You could opt to do it all 'withPC' but if you prefer to monitor direct (PC DRCT) - here's what happens:
_ You will set "M" for the latent audio signal so that your previously recorded MIDI data travels to the XS tone engine and the audio goes direct to the audio outputs. The latent Return (audio lane) is muted.
_ You will set "M" for the latent audio signal so that your currently selected live data (the new record) travels direct to the audio outputs. The latent Return is muted.

If you are listening to an AUDIO LANE you are downwind of the latent signal. Which is inherently not a bad thing (unless you do something silly like set the latent signal unbelievably high - fine for a test, although unnecessary because you should be able to tell the latent signal audibly doubling even if it is just single digit milliseconds). In most cases the latency on an average computer should be single digit milliseconds... in this case, if you were to operate completely downwind of the latency it would be undetectable by most humans. Of course, if you listen to both - anyone can hear that. But if everything you are listening to is a consistent 6ms late, you'll probably never know. (Computer-based soft-synths depend on that being a fact).

When you are overdubbing this is the most likely time when you want to monitor yourself DIRECT. Singers and guitar players (even drummers) are the first to complain about latency. This is why most pro setups use a hardware (external) audio interface with a Direct Monitoring option. Something about the nature of those instruments they are able to pick up on the latency earlier than keyboard players - I say that, and I'm a keyboard player, because I've observed it to be true. When the latency is greater than 20 milliseconds it is actually impossible to perform - laughter ensues - as you simply cannot perform. But with 3-9ms of latency, I'm not talking doubling here, but just listening through the software - you may not even be aware that the signal is late. Just FYI - when a guitar strums a chord there can be as much as 3-6ms between strings. So we are talking about a very small amount of time here. Small.

The only way I can find to get a non-latent, non-doubled signal for both the played-back tracks and the new track during recording, is to disconnect all the audio lanes from the Stereo Out bus in Cubase via the Output Routing of the Track Inspector, and use "PC Drct" mode.

The way to do it, that you did not find, is hiding in PLAIN SIGHT (as they say). It is the MUTE button. That's why it is there. That is it's raison d'etre - reason to exist. So you can decide if you want to listen to the signal thru Cubase or direct. You do not have to disconnect - simply MUTE. (Mute being a purposeful 'disconnection' of the routing. Actually disconnecting the audio lanes for the STEREO OUT bus is a bit severe, but accomplishes what the Mute button does. Use the PC DRCT and use the audio lane MUTES instead.

Final notes:
If you MUTE the MIDI Track - this is different from MUTING the AUDIO LANE... sure, they both accomplish the same result of silence for that data - but you must 'see' the deeper difference:
Muting the MIDI Track automatically mutes everything that follows. So muting the MIDI Track automatically stops all audio from ever being generated.
However, muting the AUDIO Lane still allows the MIDI data to travel to the Motif XS hardware and to trigger audio which is routed directly to the XS main analog audio outputs.