What software do you use to convert all your samples to 44.1khz 24bit wav when loading them up on the OT?

Ive been using Pro Audio Converter for 10 or so years for work stuff, which involves me batch converting large folders of files. Can set it to replace the source files (which is often what I do) and the file names always stay the same

Too bad it’s OSX only

Using XLD on Mac.
Does the job.

I use ffmpeg to batch convert files. I made a little automator script that scans a directory for all the media files, then converts them all into the appropriate WAV file and puts them in their own folder. I can also make the folder it sends to on the octatrack itself so all I have to do is right click in a folder and go to “send to octatrack” and it converts everything in that folder and puts it on the octatrack’s drive as long as it is plugged in. Here’s the script for anyone interested. It is a little janky, and doesn’t work in all situations, and will definitely need to be modified for your folder names and things. But it’s pretty useful when it works.

Thumbnail document.wflow (6.3 KB) Info.plist (767 Bytes)

1 Like

I usually use a shell loop in the command line to convert and rename, but lately I started using this graphical program, which handles big batches well and might be faster than my command line tools:

https://www.ocenaudio.com/

I’m pretty happy with it.

OSX only

I’m looking for a cross-platform solution

So I wrote this script, that is executable from Windows, OSX or Linux.
I will put it at the root of the AUDIO directory, on my CF Card.
Once i copied the samples I want > Double click on the script to convert all files found as needed and nothing else to do : D

It also get the list of supported file format from sox command line, and convert only those.
It’s no tied to the OT and can be easily configured to convert differently

image

Gonna put it on Github and update it as I improve it

4 Likes

You can still use ffmpeg, but all the commands around it will be different depending on your operating system.

Freac is also an option maybe? Has options for save to original file source folder and overwrite?

https://www.freac.org/

2 Likes

Yes that’s one of the best option I found out there

Main issue with it is that you can’t filter out files and convert only those that don’t match your specs. Also it does not do normalization I believe

So it’s not really fitting my workflow since you can’t reasonably drop your whole 64GB Audio folder

Reaper - file - Batch file/item converter
Drop your folder. It does not matter how big it is from my experience. You don’t even need to know how to use reaper at all.

2 Likes

That’s a good idea! Just copy files in whatever sample rate they happen to be, then convert in place on the CF card.

I kinda want to expand on this idea and have it standardize file names too. Oh and maybe I could use multiproc in python to get it done faster than my typical shell loop around ffmpeg

Not hard to do ! Based on sub dir name ? Or what logic ?

If you want fast better use Golang or some lower level language (C++, Rust, Java)

I’m gonna time it with a large amount of file, i’m thinking it may be a good use case for https://binpa.sh/ !

Timed it running on CF Card directly, with 7,58 GB of samples (6 124 Files):

  • 4 minutes, 58 seconds
  • 3 minutes, 37 seconds
  • 3 minutes, 40 seconds

With pa.sh:

  • 4 minutes, 3 seconds
  • 4 minutes, 0 seconds

So basically no change.

Most of time is spent on checking sample rate (soxi -r filename) of all files found recursively. Maybe there is an alternative faster command out there. Gonna try ffprobe :

  • 12 minutes, 7 seconds

So sox is x3 times faster than ffprobe

Also tried with mediainfo :

  • 5 minutes, 50 seconds
1 Like

So most of the time is spent on disk I/O, yeah? The CF card is probably pretty slow. I was talking about multiproc, but that probably won’t have any great effect if I’m gonna run it directly on the CF card.

Super easy! Use it all the time.

I don’t think so. I will test on local ssd to see if there is any difference

Only one way to find out ^^

There are quite a few ways to do it with python and it might depend on which lib is used.

It might be faster on SSD because you mentioned that sox read command is taking all the time… maybe it’s just taking long on disk (CF card) reads. You can also check if it’s disk I/O with something like iostat on linux/macOS. I’m sure there’s something similar on windows, but I don’t speak windows.

I did some test on local ssd, converting all files to 24b, 44.1kHz and normalizing them to -1dB (2957 files, 5,53 GB):

  • Reaper: 1 min 28 sec
  • iZotope RX: 9 min 41 sec
  • Shell script: 4min 05 sec
1 Like