Forums
(youtube link: https://youtu.be/nOSeug-M5-s)
hi, since i got feedback here for my last videos. i thought i might share this here. if i made any mistakes or need corrections.
Thanks.
(backup link:
internet archive: https://archive.org/details/qtractor-tutorial-by-rounak-with-voiceover-this-is-it-final-im-done )
re. tutorial
Wow, it's really comprehensive.
Thanks!!!
You've reminded me of some features I'd forgotten and others I didn't even know about (like modifying the record/mute/solo switching in multi-track selection with Shift+Control).
It also demonstrates how powerful and capable our beloved DAW is.
Just one tip: to control clipping, add a limiter plugin at the end of the master chain and set its output to -1dB. (Calf Limiter is my favorite).
This only works for rendering while recording in real time (like you did, for example, recording the system audio output, which only played the master bus).
When you use export and select multiple buses, the outputs of each bus are exported together, making it impossible to prevent clipping. The only thing you can do in this case is set the audio format to 32-bit and then use the normalize clip function. But the sum of the different buses will NEVER correspond to the mix, since in the mix, parallel and serial transmissions are made at different levels, all converging on the Master, which is the true sum, not a new element to be added.
The Qtractor exporter was never prepared to mix buses... maybe someday. :)
Exporting the whole mix
I tried to set jackd into freewheeling mode to record a song in faster-than-realtime but with no luck. In theory this should work or I misunderstood the concept.
r. exporting
I tried it once with PipeWire and the Cadence recorder, which has a "freewheel?" box for that very purpose, but it didn't work either.
re. Exporting the whole mix...
that's exactly how qtractor exports audio: using the jack-freewheel mode.
I think you just tried to outsmart it, didn't you?
:)
r. export 2
Our naive logic tells us that freewheel mode works the same as normal, only it's not real-time audio, but rather priority processing. That is, rendering will generally be faster, and sometimes, if there's a lot of processing involved, it will be slower.
We don't get XRUMs (since it can take as much time as it needs for the heaviest processing parts), and in general, processing is faster.
So, if to record in real time I select only the output of the Master (which includes all the auxiliary sends that converge on it) to a recorder, to render in freewheel mode I should select only the Master. That's what logic tells us. But it's clear that our logic is wrong.
Seeing that this isn't the case, our logic tells us, well, if Qtractor doesn't want to do it that way, I'll put the audio engine in freewheel mode, which, since it includes Qtractor and the external recorder, might work.
But no...
We're naive (which is good because it's linked to curiosity and a desire to discover, learn and create solutions), but we don't really know how Freewheel actually works :).
And in this case, just experimenting wherever curiosity and logic lead us isn't enough to overcome our ignorance. Audio development is extremely complex.
thanks for the tips and…
thanks for the tips and feedback! I'll try them out.
Add new comment