[LAU] [ANN] Sampleworks 0.1.0

Hey all!

I am excited to announce SampleWorks, a cross-platform* sound sample
manager inspired by the concept of a light table.

The main objective of the project is  to facilitate a highly efficient and
intuitive workflow in working with large collections of samples.  I'll
attach a copy of the README at the end of this message to give you a little
more info.

PLEASE NOTE that the current version is just a demo, but useful
functionality will be coming soon! But if you want a taste of what's in the
works, visit the GitHub repo at https://github.com/caelia/sampleworks

* It's theoretically cross-platform, but so far has only been tested on
Linux

README.md
----------------------------------------------------------------------------------------

# SampleWorks

SampleWorks is a cross-platform sound sample manager inspired by the
concept of a [light table](https://en.wikipedia.org/wiki/Light_table). The
main objective of the project is to facilitate a very efficient workflow
when organizing and processing collections of samples.

**IMPORTANT: the current version is just a demo!**

![Image of SampleWorks GUI](swgui.png)

## Running the demo

1. Clone the repo and `cargo run`. If that's not enough info for you, please
be patient - we're just getting started here!

2. In addition to compiling a bunch of code, SampleWorks needs to generate
waveform images for 59 audio samples. It's pretty slow in debug mode.

3. Once the GUI comes up, you can: (1) click on a waveform image to play the
corresponding sound; (2) stop playback by clicking again; or (3) click on a
different image to play a different sound.

*Please note that this program has only been tested on my own Manjaro Linux
system. Let me know if it doesn't work for you!*

## Features planned for v1.0

* Quick A/B comparisons

* An organizer window for sorting and grouping related samples

* Comprehensive metadata management. Metadata will be used for filtering and
sorting, and can be set through:
  - extracting embedded metadata from files
  - automated analysis of audio content
  - tagging of individual files or groups of files by the user

* Automated export of sample packs, ensuring consistency and conformance
to the requirements of the target system or device (e.g. maximum length
or file size, file format, stereo or mono, etc.)

* The ability to extract and process regions of a source file. So, for
example, if you have a recording that includes multiple drum hits, you
will be able to select a region of the source file with your mouse, apply
changes to just that region, and eventually export it as a new sample.

* Some form of editing/filtering capability (but see Open Questions).

## Open questions

* Should SampleWorks have its own sound editing/filtering capabilities? I do
not want to reinvent the wheel, so it would be nice to leverage an existing
sound editor such as Audacity, but I'm not sure this can be done in a way
that
supports the efficient workflow that I'm aiming for.

* What, if anything, should be the relationship between SampleWorks and
other
audio production software such as DAWs? Does it make sense to have a plugin
version of SW?

* Is it feasible to generate waveform images that use color to distinguish
different pitches? Or should we maybe use spectrograms?

* How much work should we put into analysis tools? For example, it would be
possible and maybe desirable to do a visual comparison between the waveforms
or spectrograms of two similar samples. Is that worth the extra effort, or
should we just focus on listening?

* What's the best way to manage in-process sound objects? E.g. if the user
extracts a region from a source file, it becomes its own object in the GUI.
If that object is cloned (in order, for example, to compare how two
different
FX plugins affect the sound), then another object is created. So a
medium-sized
project can potentially have hundreds of sound objects, some of which may be
duplicates. From the POV of storage and memory efficiency, it would be best
to represent each object as a reference to an audio source + changes, but at
some point a new audio file should be saved to disk, for two reasons: (1) to
prevent discrepancies between the audio referenced in a project and that
which
is saved on disk, and (2) to preserve content in case of a crash.

* Given that the objects to be displayed will include both sound files on
disk,
and derived objects, consisting of files or regions of files with
modifications
applied, what is the best way to assign object IDs?

## Last but not least ...

Thanks for your interest! If you have any feedback, please use the
Discussions and Issues sections of this repository.

_______________________________________________
Linux-audio-user mailing list -- linux-audio-user-cunTk1MwBs/[email protected]
To unsubscribe send an email to linux-audio-user-leave-cunTk1MwBs/[email protected]