Lately I’ve been using relatively large samples in Renoise, for example 14 stems (drum parts, bass, vocals etc et cetera) each 4.33 minutes long. With autoseek enabled and using the projects bpm in Renoise, creating enough empty patterns for at least the duration of the recordings and placing all the notes of the recordings on the first line of the first pattern, it all runs smoothly for possible remixing.
I feel there is room for optimization and automation here though which will help streamline, fasten up the remixing process, make it less tedious, save time and empty space.
Often with a lot of stems, the particular instruments only play at sections and not all throughout the recording (sometimes 3 minutes of nothing, some vocal grunt and another 1.33 minutes of nothing :)). Right now to save memory, one can manually cut away silences before and after parts and try to line up the sound starts at the corresponding pattern positions over the length of a track. I prefer cutting sounds to new instrument slots instead of using slice markers as sections often get processed differently / more flexibility.
Ideally I’d like a tool with gui where you can set the song length and length of the patterns you want to populate across it. It should have a process button which scans the recordings for silence (be able to set a minimum/maximum threshold length before cutting away). The tool should place the left over sound sections at the relative positions in the song arrangement. It should create new sample instrument slots for each sound section in between set silence limits.
I think this is scriptable, perhaps already half way done in paketti @esaruoho ? Will throw the idea around AI and see if I can get something working if time permits
funny that you should mention that, i just added Paketti Stem Loader as a feature, and this would be the next logical result (what you’re describing). here’s a description of it
```
Feature: Paketti Stem Loader - this will take the stems you provide it, and create new tracks, rename the new tracks accordingly, load the instruments in pakettified, introduce *Instr. Macros devices for each track, and read the BPM from the Wavefile header, or read the BPM from the folder the files are in, then calculate the length of the samples, and add the correct amount of patterns to the pattern sequence, and print, automatically, C-4 (instrumentnumber) 0G01 to every track that has been created, and enable Autoseek for each of the tracks.
```
i’ll have a think about what you’re describing and feasibility etc.
so basically, detect if the slice is empty or not, if it’s not empty, actually trigger it at the space inside the pattern sequence that makes sense to it. so like an “ intelligent stemplacer”.
Cool, that is a start! For me the cutting away relatively large parts of silence in between sections of sound, creating new sample slots for every sounding part and placing these at the expected positions across the arrangement is the crux of the feature request.
I think the hardest part would be scanning the samplebuffer, rms scanning blocks of audio to see if the volume drops below a set threshold (for example -60db) natively, as I guess this take a while for a lot of lengthy stems even in luajit? Perhaps instead of reinventing the wheel, I can check if the CDP tool can help with this. If I remember correctly there are a bunch of programs which strip silence in different ways. I need to figure out how to get the output, sounding chunks in the correct positions, line numbers in patterns across the Renoise arrangement.
For me the cutting away relatively large parts of silence in between sections of sound, creating new sample slots for every sounding part and placing these at the expected positions across the arrangement is the crux of the feature request.