Skip to main content

Notice

Please note that most of the software linked on this forum is likely to be safe to use. If you are unsure, feel free to ask in the relevant topics, or send a private message to an administrator or moderator. To help curb the problems of false positives, or in the event that you do find actual malware, you can contribute through the article linked here.
Topic: Weird behavior of this audio track  (Read 872 times) previous topic - next topic
0 Members and 1 Guest are viewing this topic.

Weird behavior of this audio track

Hello, i have this audio track that on Android plays with no sound when played on the speaker, while it's full volume when listened on the headphones. On iPhone it's good with the speaker too (using Telegram on iPhone and Whatsapp/Files on Android 11). The really weird thing is that i tried to downmix it as mono and i obtain a track with nearly no volume.

The only thing i know is that this file is ripped from a vynil.

Edit: used Audacity for downmixing.

Edit2: i found that the iPhone is using a stereo speaker set while Android may use a mono speaker and downmix the audio to that. Really weird track

Edit3: maybe i found the reason. Looking at the audio waveform i notice the left channel is the right channel inverted

Re: Weird behavior of this audio track

Reply #1
Down-mixing a stereo recording to mono is frequently problematic.  Even if you don't get total cancellation (as in this case), the result may lack some frequency components where that element of the track happens to be in anti-phase between the L and R channels.  Not recommended if it can be avoided.

In the studio, each individual voice or instrument is recorded to a separate track with the minimum possible cross-talk.  The multi-track recording is then mixed down to L and R channels as desired, with each track given an appropriate position in the stereo field.  This might be suitable to then down-mix to mono, because there will be no phase difference per track in the L and R channels.

However, if it is a live stereo recording made with just two microphones, there is a very good chance of anti-phase signals according to the frequency and position of the sound source.  That is not a problem when played back in stereo, but can't reasonably be mixed to mono.  Synthesised sound could be even worse, if phase is used to exaggerate the stereo image.

If a recording is ripped from vinyl, the L and R channels come from the stylus cartridge and the stylus itself is moving in directions L+R and L-R.  The cartridge (by its construction) extracts L and R from those.  According to how it is wired, that might produce L and R or L and -R (perm any combination of + and -).  I suspect the rip has recorded L and -R (or whatever), and as the bulk of the signal will be common to L and R, when you sum L and -R of course you get nothing.  The solution, as you have found, is to invert one of the channels before summing BUT unless you are dealing with a recording which contains a strong stereo separation, you will be just as well off (and possibly better) just to take either L or R as your mono channel and not mix them.
It's your privilege to disagree, but that doesn't make you right and me wrong.

Re: Weird behavior of this audio track

Reply #2
Quote
Edit3: maybe i found the reason. Looking at the audio waveform i notice the left channel is the right channel inverted
Audacity can "fix" that, but assuming the original was stereo, you can't get the stereo back.

In Audacity you can Split to Mono or Split Stereo Track and that allows you to edit left & right separately.   If you split to mono you'll need to delete one track so they don't cancel when you export-as mono, or you can leave it "stereo" and invert one channel.  (A true-mono track will play through both speakers.)

Quote
The really weird thing is that i tried to downmix it as mono and i obtain a track with nearly no volume.
Exactly what the mono phone speaker is doing.  ;)     (The iPhone MAY have stereo speakers.)

Re: Weird behavior of this audio track

Reply #3
I wonder how they did do this.

As for Audacity, is its default option to apply dither? Asking because it doesn't seem that L+R = zero. I didn't analyze it further.

Re: Weird behavior of this audio track

Reply #4
As for Audacity, is its default option to apply dither?
I believe it is.  If you recall, I had trouble with my nominally silent tracks (generated in Audacity) actually having an amplitude which then got normalised in the waveform mini-bar.

I've not examined the sample given (too much else to do), but my references to silence above should be read as "almost silence".
It's your privilege to disagree, but that doesn't make you right and me wrong.

Re: Weird behavior of this audio track

Reply #5
The file's left and right channels are not mathematically identical with polarity inversion so the result of mixing is not a perfect null. It could be caused by misconfiguration in the recording process, for example, some analog mixers have polarity switches on individual channels.
https://mackie.com/en/blog/all/using_invert_polarity_button.html

There is also a chance that the vinyl record was deliberately produced in such a way for some unknown reasons.

In either case, analog signal transmission is never identical, when reaching the stereo ADC, left and right channels must have some differences regardless of polarity.

OP can either...
[1] discard either channel and get a mono file.
[2] invert the polarity of one channel but keep the stereo format.