Here's the goal!
to create a tool completely inside of the current audiotool that is 1) easy to use! 2) sounds as real as professional drum vst's 3) can run with minimal PC specs
Here's where I need some great ideas
-I have the samples! ...but how do I make them easy to use? In the current drum template the snares are all in one machiniste. You have to know which one to trigger, which note that is for the specific "snare" machiniste and then repeat the whole process for the hi-hats on their machiniste 🤓... In other VST's you just go to the hihat note (always F#1, Ab1 or Bb1 for different articulations) and just tell it how hard to hit! Each snare sample is triggered on ONE NOTE and determined by how hard that note is hit! Here's my hope: A machiniste can be triggered by a single note and the velocity can determine the sample loaded.
- Can this work with one midi track? Most drum vst's work on one midi track. With the machiniste, we can only load 9 samples... However, with the note splitter and matrix there are possibilities to tie together multiple machinistes! ... but how do we make C1 and C2 not trigger multiple machinistes at the same time? Right now we have C-Ab repeat on each octave for all of the 9 samples you can trigger on a machiniste. For a project like this I will need at least 3 octaves! 😳I want to avoid the kick drum also firing off a mid tom every time it's selected if possible! If we can make this work on the general MIDI Standard Percussion key map, then mission accomplished! 🥳
-Can this run on a non-NASA PC? There are certain workarounds I could use with advanced gate/compressor side chaining! But that would end up with about... 200+ devices; most being hyper specific and probably not seeing use in the average track! I know some of y'all have great setups and could run that no problem. But I'll speak for the rest of us: MY COMPUTER SHIVERS AT THE IDEA OF CHROME RUNNING THAT MUCH.
There will be some limitations, there will be some frustrations, but we can probably accomplish something pretty remarkable! So please don't get all whiney in the comments if you can't think of a way to pull this off. ( some of y'all are way too pessimistic!) If you think there are better tools, approaches or methods that can be used in audiotool to give the community some great drums, then please share!
Create an account or Login to write a comment.
-
-
(link is only visible to registered users)
Okay, I'm not losing focus, but I am still running into this one wall of triggering different sounds based off of velocity.
It might be good to rig something like this up if I could ever figure how to split midi based off of notes.
Or it might be a good distraction so I don't get too frustrated!
-
I have an idea for doing different samples at different velocities. let's say you wanna do 3 different samples for each velocity map.
- a note splitter goes into 4 machinistes
- first three machinistes are used to play the drum samples. each output on each machiniste has a normal gate (stops the sample from playing when velocity is too low) and an inv gate (stops the sample from playing when the velocity is too high)
- last machiniste is used to trigger these gates based on velocity. each output on this machiniste controls the corresponding gates on the other machinistes
this would be 2 gates per drum samples + some splitters which might still be too much but I'm not sure if there are any smaller ways of achieving this
-
I just thought of an elegant way to get rid of half the gates in this approach. essentially one could get rid of the inv gates by playing the sample from the previous velocity step but with inverted polarity. this could then be baked into the sample for each velocity step. so
step 1: sample 1
step 2: sample 2 - sample 1
step 3: sample 3 - sample 2
etc
this way, when the next velocity gate is triggered all previous velocity samples will cancel out and only the highest triggered velocity sample plays
-
dog, you're getting credit when this gets finished. @mxcii was here
-
it's bulky, but that would work! Maybe two for kicks, probably 5 for snares (ghost notes at the lowest and rim shots at the highest) 3 for the hats and two varients for each crash/ride
That could actually be do-able!
-
-
discovery:
cutoff modulation on machiniste can be used to reduce overall number of samples
matrix on bypass only triggers notes, and does not repeat or arpegiate
note splitter velocity values do not translate to to pulv (pulv might not make a great "trigger" for gates)
graphic EQ can be used with a sharp Q resonance to translate notes from trigger to specific gates
not sure if just one note can trigger anything else in a machiniste... need to translate midi somehow or re-think the layout of this tool.
-
I'll invite you to the actual demo
-
im pretty sure there is a way to make the gate approach work but it's kinda hard to explain in comments - when I have time I'll try it out and let you know if it works
-
So, the gate approach works AMAZINGLY for hihats! But longer samples like cymbals or snare (lots of room reverb) get eaten up by the gates.
The inverted polarity approach doesn't work when different samples are being used with different frequency responses. I'm using different recordings of different "intensities" for each drum, so they don't line up perfectly in or out of phase... what might work better for this approach is clever active filtering, pitch and envelope control on the machiniste. Even though I might not continue to look into this as I pursue this vision, I still greatly appreciate the cleverness of it @mxcii 🙏
I need a certain level of randomness too to really sell the effect of the drums being a "real physical kit."
I've discovered other issues like how resonating cymbals don't really have an attack phase, they kinda do a bowed dynamic move into the next transient... It seems like what other drum vst's do is crossfade between random recordings at that velocity to get this effect and not have one million time specific samples for when you might hang out a crash.
-
6 individual midi tracks and 6 machinistes is about as good as I can get it for the thing to work consistently. I'm kinda just revamping the old template at this point, but with some smart adsr and active filtering I do think it's an overall improvement! The idea is to go into the "snare" part of the template and by moving up from C you're going up in intensity for a hit! Then velocity also will change the pitch and room response some too. My hope is to get the cymbals sounding good with this setup and then we'll be off to the races!
Next emo track will feature the prototype for the newest drum template!