There’s a mystery that’s been puzzling me for a few months now.
When I participate in Zoom calls, I record the call into an MP4 video file for later playback. During the live session, the people on the other side assure me that I sound clear and crisp and there are no audio issues.
After completing the recording and listening to the resulting MP4 video, everybody’s voice sounds hi-fi and normal in the recording (similar to what I heard in the live session), but mine sounds lo-fi!
I can’t ascertain what I myself sound like in the live session so I can only go by other people’s feedback. If something was lo-fi they would have told me, so I trust their feedback. But then why does my own audio in the recorded call (and no one else’s in that call) become lo-fi in the file?
Also, if it was some general compression or loss of quality, it would have affected all participants’ audio equally, but that’s not the case. The participants’ audio quality doesn’t deteriorate in the recorded MP4 but mine does.
The recording is done on a laptop with a built-in, no-frills sound card, I assume.
You are doing the recording using the computer that you are using to participate in the call. The sound from the other participants is being played and recorded via the sound card. But your voice is being recorded from your microphone. I’m gonna guess that the sound captured directly from your mike is of a lower audio quality from what’s being played from your sound card.
That’s the only explanation I can come up with. Surely there are other Dopers who can offer a better explanation.
It may be more that problem that we rarely hear our own voice as other’s hear it. Most people hate how they sound in real life. It just isn’t what we hear and are used to.
You could try recording your voice using a simple recording app and playing that back. Complicating matters, most conferencing systems have quite sophisticated echo cancellation built in, so there will be some intrinsic differences in the sound. But hearing a raw recorded voice without the conferencing app in the way will give you a baseline as to what you actually sound like.
Modern audio data compression algorithms are very good, and a dramatic loss of quality is unlikely. Worth noting that there are two totally different types of compression when it comes to audio. One is applied to the data to reduce the amount of data transferred, and they can be lossy or not. The other is dynamic range compression, which seeks to even out the sound level, boosting the volume of softer sounds and attenuating the volume of louder sounds. In a conferencing app this is very useful. In music, over application of dynamic range compression results in the flat over processed feel of much of modern pop music. This isn’t anything to do with data compression.
Years ago a friend had a VCR capable of dubbing. We did dubbing parties and at the first one when we played it back I sounded nothing like me. Other people said “who is THAT” even though we had all just dubbed it 1/2 hour ago. I have no idea how my voice sounded different but no one recognized it as my voice. Everyone agreed everyone else sounded normal so I dunno.
There is a distinct difference in what’s happening with the audio certainly - the other participants audio is arriving as a stream and is being played as audio; the outgoing audio from the participant at this end is not being played out locally - it’s being captured and sent.
I seem to recall hearing of someone else having this problem with recording a zoom call - in the end, they joined from an additional computer which had its microphone muted, and the sole purpose of that computer was to record the session, which included their own participation on their main computer - which was just part of the whole stream, from the point of view of the recording computer.
Correct. The audio is not the same as in the preview. Here’s an example: The 1st part is from the preview from my Camera app (mic check), the 2nd is from the actual Zoom-recorded MP4. In the preview recording, the sound quality is good, but not in the actual recording.