Hey guys, for the first time I’m trying to record my hardware jams from DTII via usb into bitwig with overbridge. Synth audio also routed to dt audio ins. Writing/arranging on my monitors sounds great with a mix that I like, but I seem to “lose it” on recording.
Lost kicks, hats, punch — various percussive elements seem to fall away or playback notably attenuated. One hi hat was also monstrously loud.
This issue seemed to improve when I re-recorded percussion standalone, without the track’s melodic elements, but dynamics were still notably different from my hardware output.
Unfaithful melodies — I had a sort of staccato lead line with some delay dialed in that sounded great. When I go to record, it was almost like half the triggers didn’t capture. Basslines have also come out sounding a touch different.
Are these the latency issues I hear so much about?
Are there phase cancellation issues that only arise once I start recording?
Are there workarounds besides recording 3-4 tracks at a time and doing multiple passes? Even this produced a sub-optimal result.
Any other advice for getting a faithful recording?
My hunch is that you’re expecting the send effects to be recorded on each individual track in your DAW. However, these are send effects and they exist in a seperate fx track. So you’ll have to make sure you set up a track in your DAW to capture the effect track from DT, and then mix that back in. They will not be captured on each individual tracks recording.
Also, I’m not sure how this works while recording with OB, but I do know there is a compressor on the main out of the DT. I assume this is also not going to be recorded on each track when multitracking, so you may have to do some post processing to get that punchy sound back.
Personally, I don’t usually mess around with Overbridge in the DAW, because of these aspects of the audio architecture.
If I really want to maintain the mix I’m hearing from the machine, I will do multiple passes, muting/unmuting elements as I go.
If I really need stems, I usually just use OB in standalone, multi track record there, and then move to my DAW for mixing/arranging.
Incidentally I just posted a thread moments ago with a similar issue for syntakt. Seems that in overbridge some of the relative track volumes do not match what they were when using syntakt standalone. Yet other tracks do match (this is after correcting the sum of individual tracks outputs to match the overbridge output volume and after setting the audio routing to post fader on the takt box)
Apart from that, even after matching volumes the raw output sounds a bit different. I wonder if there is some DAC magic or hidden master compressor going on.
Individual stems will always be pre-fader, pre-fx. That’s the nature of overbridge.
Check the stereo track in isolation if you want to confirm your hardware mix, it’s the only output OB gives you which reflects track levels or effects.
Everything else assumes that you’re outputting stems because you intend to mix and add effects in your daw so it would not be practical to give you anything else.
Since they’re send fx, they all go to the same fx bus and therefore one fx output is representative of the reality of the combined fx bus.
You might not want that, but that’s physically how it works.
If you record the stereo output of each track individually you can retain the mixed output but you run into issues where the nature of send fx may also impact how you hear the mix since tracks combine in the fx and compressor so the output of individual stereo track recordings when recombined may also not meet expectations, but it’s still something that you can try if you’re patient enough to record all individually and align them.
To touch on your concern about latency, think of latency as start point alignment. Alignment meaning if you start the daw and the recordings start out of sync with the grid, that is alignment and would not impact volume.
I do have delay/verb/chorus being routed into their own channels and definitely getting closer to the expected result.
One follow up — if I do multiple passes and record the effects tracks on each pass, I would end up with a perc effects layer, a synth effects layer, etc. Stacking these seems to sound nice, but would recording each effects layer and stacking them produce a different result than recording a single effects track with all channels playing together?
(Aka is the sum of each tracks’ effects equal to the whole?)
Right now your mix reflects everything stirred together in the effects bus which is what the fx stem gives you. Recombining tracks with individual effects recordings will probably sound different but you can definitely give it a shot, however the individual track will only be applying the rev/delay to that one track and that is not how your desired mix sounds.
idk if you’re using the compressor and overdrive on your DT2, but they both have distinct colour and dynamics that are hard to replicate. the closest sounding compressor I have found is TDR Molotok (free and the best compressor plugin ever made) and for the overdrive, either FL studio waveshaper or Cableguy’s WaveshaperCM which they gave out in an old issue of Computer Magazine which you can find here Computer Music 192 | FileSilo (licensed archive btw). Bitwig might have a waveshaper that can match though.
Most faithful recording you’ll get is if you record each track separately (one by one) directly from master. It will include all fx, compressor and drive baked in.
I usually would like to record the fx track with all elements together. I like blending all elements through the fx at once because then you can edit that stem like one big sample.
Seperately though, you would have more control of each element, which could be good.
When they’re all together, it can be fun to add further processing to that stem, as it sort of glues everything together.
The more stuff I record in general, the less precious I am being about keeping everything seperate. At first that felt reckless, but now I find it’s a just a lot less finicky, and the blended elements can sound very dynamic once you start treating them as new samples, seperate from the source material.