Traditionally, I’ve just set up a midi track, pointed to the keyboard’s channel, and connected the keyboard’s audio outputs. Whenever I’m ready to render audio I create an audio track that takes the keyboards audio outputs and hit record. The results are in real time.
I’ve fooled around with setting up my keyboards as external instruments. With that approach I can create an instrument track. If I render in place, the render is understandably still in real time.
I’m trying to get my head around what benefits I gain by using the external instrument approach vs the midi / audio way of doing things. In both cases I can record midi, play back the midi, and render audio from the midi.
I’m guessing there are other advantages to using external instruments. However, I’m not seeing them so I was hoping others could enlighten me.
I see three benefits in the external instruments approach :
When your instrument is set as a MIDI device in the MIDI Device Manager, it is at your disposal in the Output routing drop down menu of any MIDI track inspector, with all the other VSTis available, after having add your external instrument in the VST instruments window. I have three external modules that I added as rack instruments, each of them being multitimbral.
Opening the MIDI device, you can define a list of banks/presets for your instrument : it will then appear as so, in the Programs list of the MIDI track inspector . This allows to change the presets directly from it : a true workflow help, IMO. The list is saved as an .xml file.
In the same place, you can create a device panel for your instrument, providing that you have the documentation concerning the SysEx commands of your keyboard.
External instrument compensates the delay caused by asio buffer and plug-in inserted to the external inst track.
With MiDI->audio track approach, you have to shift the MIDI track earlier to match asio buffer size. And then inserting buffered plug-in to the return audio track delays the signal, you have to compensate this by yourself as well. If the inst has more that 1 return, it becomes impossible unless tracks has exactly the same amount of plugins.
Thanks, guys. Those are all excellent observations and help me see the value in using this approach.
I tend to fall into patterns of behavior based on how I’ve always done a thing, so I like to take time on occasion to learn new workflows . There’s so much in Cubase that I still feel like I haven’t fully utilized, so it’s like finding new toys.
I’m doing some side by side testing and getting confusing results. I have one track where I’ve set the Kronos up as an external instrument, and the other is just a plain midi track pointing to the input / output of the Kronos midi port.
Everything records and plays back as expected in both cases. However, the playback on the external instrument is significantly louder. Also, it sounds like there’s a reverb on it (this is a test project with zero plugins loaded). The delay and output return level are both at 0.00.
I’m using a Focusrite 2i2 and the Kronos is coming into the two mic inputs, gain in the green. It also has stereo monitoring. This is how I normally play and listen to the keyboard so that I don’t have to be running Cubase.
So, when I play back from the regular midi, it sounds exactly the same as it does when I’m playing without Cubase in the loop. Same volume, no reverb. I don’t understand why the audio coming back from the external instrument has reverb and is louder.
Additionally, I thought I’d watched a video where you could right click on the external instrument track and see a “render in place” option, but I’m not seeing it here.
I think the external instrument approach could be cool, but I would appreciate your thoughts on the results I’m getting. As it’s currently working, it’s unusable.
Could you post screenshots of your project and the related Audio Connections window pane ? Here, I have these, when using one of my three external instruments, a Korg 05R/W module, and it works without hassle (no added effect of any kind) :
Additionaly, I see that the Kronos has different audio outputs, some with effects applied and some not. Something to check, maybe. Check also again all your audio signal path, in search of any kind of ‘audio improvement/experience’ stuff that could be more or less hidden and active, this, from the output of the synth to your headphones/monitors.
About the loud level return, this can be adjusted in the Audio Connections window, so it should be easily solved. But beside this, there is absolutely no reason to get a reverbed sound if only the Cubase external instrument implementation is involved. I have never experienced such a thing and have been using it since Cubase 4…
EDIT - Check also maybe your 2i2 combo inputs : they should be at line level/impedance. Added to this, I’m wondering about the ‘Air’ implementation on the inputs of the 3dr gen of this unit. What is it for ?
EDIT 2 - Here, the Render in place command is available only when right-clicking on an already recorded MIDI part. I have never used it, by the way, relying on a ‘dummy’ bus to record the audio signal coming from my external instruments…
Render in place works on external inst tracks. It basically behaves exactly the same as VSTi tracks except only allows realtime exports and rendering.
The sound is exactly the same as you route it to an audio track. I guess you set up different audio outputs from the synth to a different set of inputs, so just try swapping the input on the audio setup, i.e. if input 1&2 from the audio interface is set to audio input, remove that and assign it to the external instrument.
Thanks for all the details. You got me thinking in some new directions and I was able to isolate the problem.
The 2i2 is a simple creature. It has two inputs (which are set to line), and two outputs. That’s it. The Kronos L/R master output is connected to it.
The problem was a dual signal path because of the direct monitoring. The audio was being routed back to the Focusrite inputs, which is fine. That then played via Cubase as expected. However, the audio was also coming out of the direct monitoring at the same time. The difference in latency between the two caused the delay / reverb sound that I was hearing, and was also responsible for the volume increase.
Takashi, thanks very much for your observations as well.
Nice of you guys to take the time to help me with this.
