Help for rhythm tracking: sync hardware to a drummer’s beat

Nice, thank you. If that works reliably, I might consider swapping the softpop for the tx-6 and stick with my op-z.

:eyes: just happened to notice bastl no longer carries the Heat Wave edition. Hm. Might hang on to it now.

Interested to see this. It hadn’t occurred to me but my A&H PX5 also has midi clock out based on tempo detection, but it doesn’t seem to handle tempo changes or fluctuations very well for clocking other gear, fine for effects though.

I’d say it’s okish… not sure if it works at an adequate level for you.

CC: @benway

2 Likes

Thank you for that. I like a lot of it, actually. I feel like the aggressive tempo changes wouldn’t be happening in the middle of the song, so that might help. And I would only have the kick and snare using contact microphones.

Were you routing all audio out to the same channel, or just kick/snare, etc? No need to test again, but just curious.

I shall tilt my head up and to the left while I ponder, gazing at the trees towering over the neighborhood

So all the drums where routed to one channel and the synths to another… and the TX-6 used the drum track as clock source.

1 Like

Going to have a chat with my drummer soon. We might end up going both routes, using the softpop input dynamics for odd improvisation, then using Ableton with BeatSeeker for more managed melody. TX-6 will have to wait for another day.

It’s interesting how many times I’ve seen this idea floated over the last couple of decades. Always good insight into what people say they want to do, and pursuing thought experiment and physical experiments about what the technical limitations actually are.

I’ve also seen a couple pieces of gear come and go that claim to make this possible in real-time, and also some claims by existing pieces of gear where the vendor later backs off those claims.

I just want to point out 4 things based on my experience that I think will be relevant.

First, I will say that your description of how you think it should work is spot on for how Tap Tempo is usually implemented in gear. Tap Tempo is normally based on taking X number of inputs for (re)calculating a new, average tempo. If it reacts to each and every input, it’s not good since just one off beat tap will cause a drastic shift. Also, when the tap input stops, the tempo/clock does just continues running. Then, when the next tap happens sometime later, it doesn’t recalculate based on the the elapsed time of the new tap and the previous one. It discards the whole “stack” of sampled taps and starts new. Otherwise, the tempo would actually come to an immediate, grinding halt.

So with that in mind, you can experiment with any piece of gear that currently generates a clock and accepts an external tap tempo. It doesn’t even have to be a dedicated Tap Tempo input, it could be a MIDI Command. Lots of gear with Tap Tempos implement a MIDI Command that triggers the tap, and you can find ways to convert an audio source into that MIDI Command (if you aren’t using a pad that just generates MIDI to begin with). There are limitations, such as most Tap Tempos accept Quarter Note input only and can’t be changed, but this is just for a proof of concept anyway.

Second point: What gets discovered is the inherent difficult of having a human-machine feedback loop. What I mean by that is, your input for tapping is suppose to drive the sequencer tempo in real-time, but the sequencer doesn’t adjust to each tap instantly, there is a delay. While you are “tapping”, you are also listening the the notes of the sequencer in order to perform along with it. The feedback loop is that you want the quarter notes you play to be on your internal “time line”, but you are immediately hearing the quarter notes from the sequencer on a slightly different “time line”. Hearing that offset causes you to naturally want to adjust your performance to compensate. You are trying to drive the tempo, but also can’t help reacting to the tempo that you hear. It’s a very tricky thing to focus only on driving the tempo and not reacting to feedback. (with an all human performance, there’s a natural tendency for all the performers to adjust to audio input and provide immediate “jumps” in their timing so that the feedback loop isn’t experienced by anyone beyond a single beat)

Third: Up to this point, it’s just been about a simple Tap input going on. But this is really about taking a drummers organic performance and deriving a tempo … which can be done by a very intelligent algorithm, but not in real-time. There’s always a delay for analyzing and processing. There’s always a buffer. There’s a buffer even for a simple Tap algorithm, and this is much more complex.

Fourth: Say you have the most ideal solution technology will allow, You are still going to be asking the drummer to adapt his playing to suit that devised method. He’s probably not going to enjoy that constraint and giving up his full artistic expression. Maybe it’ll lead to great art through adversity, but likely it’ll feel more like a PITA.

When I compare that to a solution where the drummer (or anyone else in the band) simply has to monitor a click to keep everything on the rails, I just don’t see the juice being worth the squeeze, but I’ll gladly pay attention to your efforts and see what you find out, especially if you don’t just come to an abrupt end.

6 Likes

Thanks for sharing your experience on this subject!!

For now it sounds like we are foregoing any sort of tap tempo route, and will lean on embracing the experimental side of softpop’s envelope follower. I really like the idea of each performance having a dash of improv

1 Like

Update: Last jam session we got some decent grooves going using different clock division but still used an regular kick mic, so the dynamics would still cause the sequencer to jump ahead a step or two here and there.

The contact mics have arrived, so I will be testing those next.

My friend said he likes playing to a click, it turns out. So problem maybe solved.

With that said, I’m still toying around with this idea, so I think the thread can stay open.

Is there any way the Analog Four fits into this scenario with CV sequencing and audio input as an oscillator? I know the oscillators can use audio from the external input, but I’m not 100% sure on the limitations. Would that be more for melodic syncing vs rhythmic? Like with a guitar signal, perhaps? Could spin up a different thread.

Edit/

Nord Modular?