Midi conversion at mixdown vs before mix

Pure curiosity

To be able to render a midi track to audio using

Track > Midi > Convert Tracks to Wave File,

your windows audio control panel has to allow you to select the midi output as an audio input, or you have to use a loop-back cable. Yet, when you do a mix-down, the midi tracks are rendered to audio without the need for either of these things. Why is this?

What are you using for MIDI? VSTis? Just curious as a lot of guys don’t realize that there is no reason to convert a VSTi track to audio.

Yeah Bubba, I just mentioned that in tspringer’s other post on the subject.

Yeah, I know that there is usually no reason to convert VSTI output to a wav file. I guess that I’m trying to understand the signal processing a little better.

At mixdown, the VSTI’s audio output is included automatically in the mixdown wave file. So why does n-track have you fiddling with the audio interface to create a virtual or physical loop-back if you want to convert the VSTI output to an audio track prior to mix-down? After all, I could solo the VSTI track, mixdown, and then import the resulting wave file back into the original song. Is the loopback required just to maintain phase-perfect synchronization?

The Wizard is just legacy functionality that was there long before n-Track supported VSTi.

Perhaps it needs to be removed/deprecated.

Quote (Mark A @ April 23 2006,09:31)
The Wizard is just legacy functionality that was there long before n-Track supported VSTi.

No, the wizard functionality is useful and needs to stay as it was designed to be used with OUTBOARD MIDI gear… not VSTis. Remember back in the day before you could fit the whole London symphony orchestra in your PC you actually had to buy rack mount sound modules and route all that crap with real wires and render. The freeze function is the equivalent for VSTis/DXis. Freeze them tracks, don’t mess with all this loop back junk.

EDIT: DAH! We have two threads where we are saying the same things. I just read the other thread and can we say echo? Any who, tspinger should get it now. :D

<!–QuoteBegin>

Quote
No, the wizard functionality is useful and needs to stay as it was designed to be used with OUTBOARD MIDI gear


Not sure I quite agree there Bubba.

The Wizard specifically leads the user through “connecting” (via software) the soundcard’s MIDI engine to the recording input. It therefore is for folks using the builtin MIDI engine in their soundcard. I remember Flavio adding this feature sometime after I started using n-Track - the forum had lots of queries about how to include the MIDI sounds in the mixdown.

Afterall, for external devices it’s easier to connect up the external device to the PC and hit “record”.


Mark

EDIT: Just seen your EDIT Bubba. Yes, 2 threads.
Quote (tspringer @ April 23 2006,08:09)
Yeah, I know that there is usually no reason to convert VSTI output to a wav file.

Others have already answered your main question. But this point is not quite true. There are a number of good reasons to convert VSTi to audio before mixdown. I always do it. I wouldn't say "usually no reason"; I'd say "not absolutely necessary". Probably what you meant, but I'll continue my rant with my reasons:

1) reduce CPU usage

2) to apply volume evolutions. (While it's true you can do volume evolutions on MIDI tracks, they interfere massively with programmed MIDI volume messages, and they're in general clumsier than audio volume automations. And yes, it's also true that you can do volume automations on the VSTi channel, the volume line doesn't lie directly over the MIDI track you're controlling. And finally, I like to look at the wave when I do it.)

3) signal stability, #1. With MIDI, unless you're a system wizard or have an unusual system, you can't get MIDI timing to better than 8 msec resolution. Furthermore, certain things I won't go into here can cause additional MIDI jitter. Render to wave, and each note will always fall at the exact same point.

4) Visual reasons: seeing exactly where the sound falls in the timeline. For reasons I don't quite understand, there's a time gap between the start of a MIDI message and the resulting waveform. If I need to look at timing issues, I'd rather look at the horse's mouth (the rendered waveform) rather than the MIDI data. Also, the waveform display is useful and helpful when doing things like volume automations or surgical rogue peak suppression.

5) Archival reasons: Some day I may want to remix and the VSTi may be long gone or not supported for Windows XYZ.

6) Collaboration reasons: I'm going to ship the track to the song engineer, so I need the instrument wave file anyway. (OK, that's a special case.)

As far as I can figure, there are really 4 cases, and the wizard only the first two:

1) soundcard MIDI synth
2) standalone softsynth (quite rare, we normally use a plugin synth)
3) plugin softsynth
4) external MIDI synth

The Wizard should be updated to ask which of these cases it is (possibly ignoring case 2) and provide correct instructions for all cases.

Furthermore, unless it’s changed, the default behavior was to DELETE THE MIDI TRACKS! This is a really poor choice for default, since we might change our minds about the MIDI part. There’s no need to delete it! By default, it should just get muted.

Quote (Mark A @ April 23 2006,12:18)
The Wizard *specifically* leads the user through "connecting" (via software) the soundcard's MIDI engine to the recording input. .

Correct sir. Can you tell how often I use it? (Never....) Freeze is your friend.