Multitrack stem creation, advice pls

MarkGuitar

Hi
As we now have a pretty stable and really nice ST4 I am going to start exporting my stems from Ableton and loading them into ST4.
Couple questions.

Sample Rate
Currently all my files and gear (mixer etc) are set to 48k Did I read here somewhere that the ‘native’ in ST is 44 and to use 48 requires some kind of resampling by ST? Maybe I misunderstood? But if so, does that mean ST will run better with 44k stems?

Wav vs MP3
There was a discussion about this a few weeks ago. Just double checking. Is it definite that there is no overhead in ST if I use MP3? Ableton definitely suggests using Wav as it internally converts compressed formats to Wav on the fly if you use them and thus there is an increase in processing and thus a higher risk of skipping, artifacts etc. Hoping this is not the case with ST.

Thanks for any guidance on this. Exporting 10 or more stems for each of 400 songs is a long job. I only want to do it once! 😄
Thank you

Riff

MarkGuitar For Sample Rate, I think the main thing is to keep all of your files as 44.1 or 48, and to set ST4 to match.

I've been using MP3 exclusively. Not sure of the backend overhead vs Wav files. Definitely no sound difference.

peter

MarkGuitar But if so, does that mean ST will run better with 44k stems?

No, the important thing to note is that whatever sample rate you choose, stick to it. Set ST4 to that sample rate and make sure you don't have songs with a mismatched sample rate. It does not matter whether you choose 44.1 or 48 from ST4 perspective. If your audio interface only supports 48k, use that for anything else.

MarkGuitar Is it definite that there is no overhead in ST if I use MP3?

No noticable overhead. The processing needed levels out between the larger WAV files needing no processing to load and the smaller mp3 files needing decoding. Regarding audio quality: as @AnthonyB mentioned: do your own double blind test and decide for yourself. I did that and could identify 128kBit mp3 pretty accurately (on studio headphones) but starting with 160kBit and above I did not hear any difference. Test results have been completely random for me. So I decided to encode everything with 192kBit years ago and still have no issues with these files.

Recently I did some double blind tests with m4a files and that has superior encoding at lower bitrates to mp3. Now I have moved all my live tracks to use 128kBit m4a and that works pretty well for me while saving lots of space.

musicadi

MarkGuitar I would say that 48k is only relevant if you use plugins on your tracks/master, so you can have more headroom before distortion.

Not sure about sample rate conversion, but you will definitely want to use the same sample rate for both your files and ST4, unless you want surprises 🙂

As for wav vs mp3, the extra storage needed is probably not worth it in a normal situation, and most devices have hardware decoding nowadays, so mp3 would definitely be my choice, Anything above 160 kbps sounds great. I mean, for 4000+ wav tracks at 48 kHz, you’d definitely need a device with 256Gb, and they are… a lot more expensive 😛

EDIT: not sure about this, but logically, if you use 10 wav tracks per song, your device needs to read about 20 mbps, from storage, as opposed to, say, 1,6 mbps if you use 160kbps mp3 🤷‍♂️

Riff

MarkGuitar Exporting 10 or more stems for each of 400 songs is a long job. I only want to do it once!

Tell me about it! The monotony of ordering tracks, changing names, colors, bus routing, and transpose ability for every single track many times over for each song will be rough. I'd love for some automation abilities down the road. It took me several days for 350 songs.

AnthonyB

Sample rate - more a case of going with whatever you can most easily standardize on across your gear. Usually your mixer is the determining factor 🙂 Are you going to be able to hear the difference between 44.1 and 48kHz sample rate audio files? Not likely. Then as Riff said, just 'pick and stick', or to reference Ghostbusters, don't cross the streams otherwise bad things might happen!

Wav vs MP3 - The summary of that conversation was that Peter confirmed with his dev tools that there is no additional overhead going with MP3 over WAV files. I unscientifically observed the same. This is all despite the fact that in theory there is additional processing involved in uncompressing MP3 files (WAV is uncompressed). Maybe this is offset by the much smaller file size which reduces the 'cost' of the storage read operation. Either way, it might all be true in theory, but from a practical stand point, there is no observable or measurable performance difference.

From a 'can you hear the difference between uncompressed WAV file versus a compress MP3 (where MP3 is a lossy compression algorithm)'? The answer is most likely you can't tell the difference, but do your own test (try to make it a blind test to remove any bias) to validate this. Then also factor in the environment you'll be using these files. In a loungeroom on pristine speakers with no background noise? Pppffttt... not likely. Most likely in a noisy venue through quite average FoH speakers. I'm a bit of a 'because I can, I will' type of person but even I decided to switch from WAV to MP3 simply so I could reduce storage requirements on my iPad.

If you are doing 10 stems per song for hundreds of songs, file sizes are going to become an issue pretty quick even with the largest available iPad!

Anakin

peter I confirm what Peter said: m4a files are lighter and work perfectly. I have everything in m4a, and each song has at least 8 separate tracks.

Riff

Of course Scott Uhl made a video about using Wav vs MP3 for backing tracks. I always appreciate his approach.
https://www.youtube.com/watch?v=Bw_NwsMZnmY

jaytrax

Firstly it all depends on the original source material

If you purchase your backing tracks for example from Karaoke-version, these are MP3 at 320

The ideal format is leaving as is in MP3

However if you purchase tracks elsewhere or make your own from scratch and the source is Wav
Then you can use Wav
However Wav take up a lot of storage
So in this case M4a is smaller than MP3 and is a higher quality format

But remember there is NO benefit trying to change an MP3 to Wav etc - it just wastes space, you can’t “increase the quality”

MarkGuitar

Thank you to everyone who offered help here, much appreciated.

Just to wrap up, I have decided to export from Ableton as follows

Sample rate 48k as all my gear is set that way and I don’t want to be having to change stuff every time I switch over to using ST during my transition away from Ableton.

I will export MP3 at 320. Would have liked to use M4A but Ableton only offers MP3 as a compressed file option 🙄 . And it’s going to be a long enough process without adding a file conversion into the workflow. Any new songs will go straight from Cubase to ST with no Ableton (hoorray!) so for them I can use M4A as Cubase developers are in the 21st century 😄

Thanks again for the advice.

Riff

An option I use frequently is using Audacity for file conversions. I know you said that you didn't want another step in the workflow, and I get that. But what you can do is export from Ableton as Wav or Mp3, and then batch convert and save them all at once to m4a before sending them to ST4.

peter

I use XLD for batch file conversions. Very good little utility: https://tmkk.undo.jp/xld/index_e.html

AnthonyB

Yeah, I was going to say that there are a ridiculous number of audio converter tools (free and paid) available that can do batch processing. That way you could just point at your folder (and subfolders) of existing .WAV files and have a tool convert them to whatever format you want, retaining the original filename.

Like @Riff I use Audacity as it is a very solid and free (although starting to risk going down the path of 'bloatwarey' which makes me worry about the future). I use it for any audio editing (including offline pitch shift for any key change requirements), quickly verifying my DAW exported stems before I transfer them over to ST, etc.
While Audacity can do batch processing (conversion) using it's macro feature (https://support.audacityteam.org/audio-editing/macros/macros-palette#apply-macro-to) there are much easier tools that let you 'point and convert' so to speak.

I can't speak for MacOS as I'm a Windows user. I've used AVS Audio Converter many times very happily.