Since the Model:Samples only plays mono samples, I kept having to convert stereo sample packs before loading them via Transfer. Got tedious pretty fast.
So I built a Mac app called “It’s mono, yo!” — drag in your WAV or AIFF files, pick your output settings, and batch convert them all to mono at once.
Just shipped v1.2 with configurable bit depth (16/24/32-bit), sample rate conversion, and no file limits. Handy for prepping a whole library in one go.
Finding it hard to spot what your app does that transfer doesn’t. Transfer converted everything I ever threw at it, stereo, varied bit depths, sample rates.
media human audio converter is generally good for converting stuff w/o any meta data and other stuff in the formatting to get in the way of compatibility with different hardware devices.
Fair point — Transfer handles a lot. This is more for prepping files independently of Transfer, especially if you’re feeding samples to multiple devices or want to batch convert a whole library with specific output settings before importing anywhere.
Not sure if this is the “difference” but one thing transfer does NOT do is mixdown stereo to mono. It just takes the left channel and thats it. So if you have a stereo sample panned hard right, Transfer will just import silence.
This is the reason why I’m batch-converting samples in Audacity first, before going to Elektron transfer.
There’s a terminal command for Mac and Linux you can use for this sort of thing. It’s called ffmpeg. It’ll also convert anything into anything so you can change file formats as well if you need to.
Its a good option for programmers but if you can get it installed then ai can just tell you how to use it.
Quick update for anyone who tried this back in March: 1.4.0 now imports whole folders and keeps the folder structure on the way out, so you can point it at a full sample pack instead of dragging files in one at a time. That folder step was the tedious part I originally built it to avoid, so it feels a lot more finished now.
Also worth repeating what came up earlier in the thread: Transfer (and most quick converters) just take the left channel. This does a proper summed mixdown so you don’t lose whatever was panned to the right. That distinction turned out to be the whole reason it exists.
Still open source (MIT) if anyone wants to check the downmix maths. Happy to throw a few App Store promo codes at anyone who’d rather not pay the buck, just say the word.
I’m here! Just been lurking for a while.