JeromeB Contributions

[JeromeB]

btw. you wouldn't call it "revert" but "reverse"

Ok, thanks, I will correct this. (I speak English since I am registered on the forums and it's not my cup of tea )

[Blindsmyth]

He do you think it would be possible to adapt katjas transient analysis techniques for analysing the content of a table.
So that it would analyse all the transients in a table full of audio and spit out a number of indexes for each transients.

With that you could record a loop/fragment of audio and then kind of slice it to the transients and retrigger with incoming midi notes

[JeromeB]

Hi @Blindsmyth, sorry for the late answer.
Yes, I think it's possible, it's certainly easier than doing it in real time

I see four ways to do this :

1. record with your favorite sound editor, then slice by hand : yours eyes are the best algorithm. I think it's not your goal ;
2. record with your favorite sound editor and use it to slice the recorded sample : some good editors have a slicer plugin ;
3. record with axoloti, then press a button to slice the recorded sample, and play with slices : I think it's your goal ;
4. record and slice in real time. Today, demo patch have 4 channels, but you cant increase this by adding beatslicer objects. It's not your proposal but, I think, it works the same ways, no ?

But I find it interesting to code the point 3 !
Not now, I'm going on vacation three weeks, without computer..., but with my axoloti core, a smartphone battery and an STM32F429I-DISCOVERY board to control the axoloti board ! (touch screen and midi usb device configuration).

I will not forget to watch this on my return.

[Blindsmyth]

record with axoloti, then press a button to slice the recorded sample, and play with slices : I think it's your goal ;

yes that one would be suuper funky

But I find it interesting to code the point 3 !Not now, I'm going on vacation three weeks, without computer..., but with my axoloti core, a smartphone battery and an STM32F429I-DISCOVERY board to control the axoloti board ! (touch screen and midi usb device configuration).

Great great plan I'm going to do something similar by the end of the month with a looper/sampling axoloti device that I'm finishing at the moment.

Where are you going? Have fun!!

[JeromeB]

I'm going to do something similar by the end of the month with a looper/sampling axoloti device that I'm finishing at the moment.

I'm looking forward to see it !

Where are you going?

Turkey, hopefully.
This is not the best time but I will in the quiet part, on the beach in front of Greece.

[JeromeB]

Hi @Blindsmyth,

Sorry, I have not had enough time to work on the topic. I have just a prototype (attached to this post). I think the solution is not very far from your need :
- record and slice in realtime, in the same table (one table for slices and one table for indexes) ;
- you can store slices on sdcard ;
- you can play selected slice (with a new "slice player" object).

Todo :
- add pitch input to slice player ;
- add reverse mode input to slice player.

Beatslicer_DEV_IV.axp (95.6 KB)

You can make a looper with slices. Then you adjust loop length without change slice pitch :
Beatslicer_DEV_II.axp (24.6 KB)

[JeromeB]

New test patch. I just use table on parent and a subpatch for polyphony.
Reading table in parent patch from polyphonic subpatch is a great feature.

Beatslicer_DEV_V.axp (20.1 KB)

[Blindsmyth]

He thx for the effort. I am really deep in my multitrack looper / fx setup so I didn't really try it out yet, but I will test it soon

[JeromeB]

No problem when you have time !

[Blindsmyth]

Hey
I finally got to try your beatslicer out. It's is pretty nice, especially the DEV patches you sent to me!

However I'm a bit unsure of how to use it atm. My my main goal is a multitrack Audio Looper that can also do the magic stuff like mangling, pitching etc.
Step one is building the multitrack looper, I did so using the looper blocks of @rbrt. The next step would be to integrate the mangling elements, like your beat detection slicing. However this would be combining two different systems, it would be the most efficient to have all in one system and i'm a bit unsure how to do that.

@rbrt you see any possibilty of integrating @JeromB's work into you looper blocks? It would be quite amazing

[rbrt]

hi Simon,
I will check out @JeromeB 's stuff real soon,'cause I just tried to patch a live-slicer some days ago..and I wasn't really satisfied with my own objects..
especially ,there has to be some kind of x-fading between the slices...

[rbrt]

...but right now I'm really busy on some new midi and sequencer stuff..

[JeromeB]

Thanks for your feedback.
When I would have time, I want to write a midi looper. My goal is to record each slice detection in a midi recorder in order to have the same function as a looper but with new capacities like removing a slice, change pitch (with and without change length of the slice), change sequence tempo without change pitch of each slice, reverse slice, retrig slice, etc. All of this in live mode. But my ideas on the matter are not too clear, I am lacking time to investigate this.

[Blindsmyth]

@rbrt
that would be very very dope

@JeromeB
cool I think we want to to very similiar stuff! writing positions of transients to a table would be a great first step.
All the other stuff you describe is I think achievable with patching.

But for me I already have the looper environment, I only need the information where in the table the transients are, then I could do what I want with this information.
so it's pretty great of @rbrt to show interest in this

[mtyas]

Hi @JeromeB I must say I really enjoy using your objects, so thanks and bravo for the hard work.

I was playing around with the turing.axo object today, it's really amazing and great fun to use live (and just plain useful I must admit), but I was wondering if there was a way for the bit buffer to be adjustable so that we could have 3 / 5 / 7 / 9 bar loops instead of always 4 ?

[JeromeB]

Hi @mtyas. Thank you very much for your feedback.
I think it's possible to have an adjustable buffer length.

But, I work on an alternative : a mix between an euclidean sequencer and an turing sequencer.
I use a cpu efficient euclidean rhythms generator inspired by an old puredata patch.

Turing sequencer produce an periodic random number.
This number is used as an array index, of pentatonic scale.
Euclidean algorithm is used for the rhythm.
I have a patch for that, as an attachment of this post.

I have to go on a business trip. I'm looking at how to change the length of the buffer when I get back.

turing_euclidean_dev.axp (14.0 KB)

PS : Recently I write all my sequencers for a STM32F429 card. That way I can keep some CPU for signal processing and I have a display. Requires only a few resistors, an optocoupler and a diode.

[JeromeB]

Euclidean + turing sequencer on an external STM32F49I-Discovery board.

MDP32 in order to control STM32F429I board.
Midi out to axoloti board.

[lokki]

hi there, i realise this is an old post...hehe.
i was wondering if there is a "simple" way to completely empty out a beaslicer object, so that the buffer is zeroed out and all audio output stops? i would need this to build a "crazy" drum machine. great slicer object really!!

[JeromeB]

Hi,

I can't remember the code in detail, it's old object, but I think that is possible.
I intend to do clean my objects repository next week.
I'll let you know, of course !

[lokki]

thanks for considering!!
another bug i found is that in your beastlicer example with the turing sequencers the beastlicer does not seem to react to multiplications of the clock when in reverse mode. this is a rather strange bug since those are two separate objects.

EDIT: maybe it is just the retrig inlet that stops working in reverse mode...