2gb limit on recorded file size.

[RalphShnelvar]

I was recording audio via the tape recorder and was astonished to learn there is a 2GB file size limit on recorded audio via the tape recorder.

A) is there a way to break that limit?

B) Of the 8 channels recorded, only 3 have data. How can I suppress recording them?

[Vincent Burel]

yes, there is a limitation on mp3, wav and aiff to 2GB file size (historic constraint).

you have to use the BWF format (wav compatible) to be able to go over this limit.

the number of channel recorded is defined by you in the recording option dialog box!

[Killy]

This is really confusing.
I thought WAV has 4GB (not 2GB) limitation and BWF has the same limitations as WAV, just additional metadata. And it is RF64, another format built after BWF, that removes file size limitation...

Application is better to support writing in chunks, starting another file after reaching file size limit.
And provide a clear statement of limitations somewhere. Something like:

(File type combo box)
- WAVE (split in 2GB file chunks)
- BWF
- AIFF (2GB file size limit)
- MP3 (2 channels limit, 2GB file size limit)

(Also, some compressed format for composite audio would be nice to have. MP3 is limited to 2 channels only, and so multichannel record is limited to WAV-like formats and consumes disk space quite quickly.)

[Killy]

This comment is a summary of a discussion happened in twitter, with some extra.
(twitter is a mess, I don't know if there is a reliable way to see all branches in the thread)

The issue I got with the recording feature:
The task was to make a workflow for recording long streams with voice and other sound sources stored separately. Quality isn't a concern, so compressed format would be preferred.
Maybe it's just hit a spot in between casual and professional use. Considering this topic got no traction, it isn't everyone's problem. But I'll just document what I've got for anyone who might face the same issues.

Now the points in the discussion:

1) no mention of file format limitations anywhere in the interface. This can lead to recording being stopped in the middle of process and some footage being lost.
- not acknowledged.

(You all remember you favorite youtubers saying something about stopped recording due to any sort of technical mishaps and explaining in words what happened between last recorded moment and the moment when stopped recording was noticed, aren't you?)

2) implementing workarounds that will allow to record more stuff with existing formats - starting new file after reaching size limit, writing different channels into different files (like MT32-SPLite can do already).
- not acknowledged.

3) adding file formats free of existing limitations:
CAF - being similar to RF64 it doesn't add anything to existing formats;
OPUS - lossy compressed format with multitrack support;
FLAC - lossless compressed format.
- no plans for OPUS or FLAC support because of CPU requirements for encoding, especially in multitrack situation.
- I hope someone can back me up with encoding benchmarking - OPUS versus other formats. I know one of the goal of this format is to be used on low performance devices, but I wonder how much it will actually take to encode stuff in real time.


Recording solutions when built-in recorder isn't enough:

1) MT32-SPLite. I can use it myself, but I can't enjoy using it and I can't recommend it to anyone else, when getting records on disk is all that's needed. Being a tool targeted for professionals, it is very complicated for average user.
- complexity not acknowledged

I'll try to elaborate here:
- absolutely alien, non-native GUI - if you're unfamiliar with the app, you don't know where to look for things you need, what's interactable, how to interact with things...
- optimized for touch panels in professional environment. Popping up keyboard is quite distracting. No usual on-hover reactions and tooltips.
- mandatory storage setup requiring two drives.

2) Carla + MRecorder (picture). Still complicated for average user, but more straightforward solution for someone like me. Silently switches from WAV to RF64 after reaching size limit - might be a bit of confusion (in case when opening with software not supporting RF64) but doesn't cause footage losses.

3) recording software with ASIO support, like Ocenaudio. The easiest way to go so far.

(this is extra, not a part of twitter discussion)
4) in case you recording in OBS or similar tool - learn it's own capabilities. OBS is actually quite versative in audio recording.
First, it allows to record multiple audio inputs to different tracks inside output file (won't work for streaming though). Looks like this is the only way so far to have compressed audio stored straight away.
Second, it supports VST plugins, so MRecorder can be used to capture separate inputs.