Feature/voxpop

[teknopaul]

PR adds an instrument to make playing voice samples easier.
It allows importing multiple cue points that can be created with labels in Audacity and exported as .txt.
The Instrumnet imports a .wav and the cue points as a .txt file and then setups up sampleBuffers internally for each cue point.
Currently selected cue piont's index and text is shown in the GUI.

To select the cue point during playback user can chose between 1 of 3 modes
Sequential - selecting the sample one after the other in order, e.g.a song chopped in to verse1, chorus, verse2, verse3
Automation - separate automation track selects the cue point by index, 1,2 3 4
Piano - note chosen selects which sample to play (naturally pitch variation is lost in this mode

[Rossmaxx]

Just curious, what's the difference between this plugin and the new slicer plugin at #6857 ?

[RoxasKH]

Yeah, i'm having kind of an hard time understanding its purpose too.

You need to load 2 separate files (which you have to go through audacity to create) to do something that at my current understanding seems could be easily pulled off just using the sample tracks in 1.3, as slice aren't that frequent with vocals, or well if you're doing chops, SlicerT has been merged as well.

So i'm most likely trying to understand the usecase where this would help comfortably improving the workflow.

[teknopaul]

Not seen that plug-in. Where do I get it? Potentially it's the same thing but green.

…

On Sun, 19 Nov 2023, at 16:23, Rossmaxx wrote: Just curious, what's the difference between this plugin and the new slicer plugin at #6857 <#6857> ? — Reply to this email directly, view it on GitHub <#6994 (comment)>, or unsubscribe <https://github.com/notifications/unsubscribe-auth/AAGGM6RWPHYMKEQCZAV7KRDYFIP53AVCNFSM6AAAAAA7RYEJOOVHI2DSMVQWIX3LMV43OSLTON2WKQ3PNVWWK3TUHMYTQMJXHA4DMNJSGI>. You are receiving this because you authored the thread.Message ID: ***@***.***>

[teknopaul]

Audacity does not export cue to wave (yet) a Linux user

…

On Sun, 19 Nov 2023, at 16:26, Roxas wrote: Yeah, i'm having kind of an hard time understanding its purpose too. You need to load 2 separate files (which you have to go through audacity to create) to do something that at my current understanding seems could be easily pulled off just using the sample tracks in 1.3, as slice aren't that frequent with vocals, or well if you're doing chops, SlicerT has been merged as well. So i'm most likely trying to understand the usecase where this would help comfortably improving the workflow. — Reply to this email directly, view it on GitHub <#6994 (comment)>, or unsubscribe <https://github.com/notifications/unsubscribe-auth/AAGGM6WWVN5M5JI52KECVULYFIQLXAVCNFSM6AAAAAA7RYEJOOVHI2DSMVQWIX3LMV43OSLTON2WKQ3PNVWWK3TUHMYTQMJXHA4DONJWGE>. You are receiving this because you authored the thread.Message ID: ***@***.***>

[Rossmaxx]

Not seen that plug-in. Where do I get it?

The latest master. It was merged just a week ago.

[teknopaul]

OK finally managed to rebase and look at SlicerT

Kinda cool, but not the same thing. Looks designed for chopping up loop samples like rex, an algo for finding beat start points will not work reliably for voice because of vocal sounds like 'S' and 'W'.

I loaded my e.g voice sample and this is what I see with slicer, its split per word (without text)

Rather tha split by what you would want to trigger, i.e. a whole phrase / sentence.

Slicer does not seem to allow pitch shifting? or sample selection with automation? i.e. it seems it only supports one of the modes in Voxpop (piano). The UI is like all lmms plugins 250px across. For the whole wave that is too cramped and you don't get tools for finding 0 Z positions. The idea of importing cues from Audacity is that you get a full screen GUI and better zoom tools and labeling, start and end points and a standard format as output.

The music I've been writing with Voxpop I could not port to slicer.

I could try to pourt some stuff from voxpop to add to slicer but I dont think the goal is the same, slicer is a looping tool , voxpop is voice sample player for "one shots" I dont think a feature merge would make sense.
I plan someday to add key mappings to voxpop so it can be used as multisampler with configurable pitch range per sample. This will resolve a few use-cases logged for "multi-instrumnets"

I could try to port cue sheet import from voxpop to slicer, looks like slicer does not have end points or text tho.

[teknopaul]

For comparison here is same sample in voxpop

phase currently selected is "sssswipe da funk" with a long S so the sample start point is not on a peak.

The green digital numeric display (showing 3) is Automatable, you can drag drop it to an Automation track to select cue point, when in Automation mode. This way you can "slice" a sample and still pitch shift when you play it with the piano roll.

The stutter feature is as per AFP and again is automatable, you can schedule turning on and off that fceature in a song.

End point selection is on or off per the LED light when the fiel is loaded, but end points are not visualzed

[tresf]

@teknopaul thanks for the submission and for explaining the difference that this plugin has over others #6994 (comment).

it silly to publish a plugin architecture and docs on how to use it and then not accept user generated plugins.

We do! Can we keep this open?

[teknopaul]

Voxpop is not being developed on its own branch.

If it is wanted somewhere else, I would need to create a new branch and separate stuff out that is not wanted.