I have bombed the most important episode of my new podcast and desperately need some help.

Hey guys,

I hope that someone will be able to help me.

I just started podcasting and have been fortunate to record an episode with someone who I am looking up to, and she is well recognised in the world of leadership: Kim Scott.

Unfortunately what happened(my theory) is the cable went bit loose and the quality of the audio deteriorated significantly.
This same happened during another episode, only this time I can hear it happening.
I have attached the audio where you can hear first few seconds being OK and then it all goes south.

I hope that there is hope for me and its possible to improve the audio and make it an OK quality at least. There is no way to invite Kim again and record it again, and I need to find a way to fix it.

I would tremendously appreciate your help!

The spectrogram track view clearly shows that after the first couple of seconds, there are virtually no frequencies above 4 kHz.

First Track000.png
That can happen if the sample rate is reduced to 8000 Hz, but that does not happen with a normal microphone recording.
How / what are you recording?

Was she in the room with you? I don’t see any way to rescue that sound. There is no clean, clear sound to boost.

There is a Hollywood solution. Find somebody who sounds like Kim Scott and have them read the same words in a studio or quiet room.

I double record important interviews.

This is my phone sitting on the table with the microphone up in the Voice Memo app. I could have used this sound if the regular recording failed.

This is a broadcast interview and I did the Los Angeles section. The Rode microphone on the right is the backup. My microphone and recorders are on the left. Both sound files recorded the show.


I have used the snowball yeti microphone, and it was recorded virtually over the zoom. I was hoping to hear that there is a way to save it/boost it.

What you are suggesting is that I could record everything that I have said again, and then mix it all together again?
I could transcribe the episodes, and record again my part, and then mix it up somehow.

Time-consuming but for the episode like this one, worthy!

I could in the future have a zoom call and recording audio via zoom, and turn on audacity as a second source? Would I need two mics?

Steve, what would be the reason behind losing the frequencies?

I have done it on my Chromebook, and it happened few times already and today I did an episode on the PC and not once this issue occurred(so far)

it was recorded virtually over the zoom.

So that’s a Zoom recording? That’s super good to know because this quality problem has come up before where the chat company (apparently) destroyed the sound quality.

not once this issue occurred(so far)

You can’t do that. If somebody forced me to guess at this problem, I would say internet congestion and connection problems at that exact instant caused the reduction in quality. That muffled voice takes less data than a full, clear quality voice. I’d be shocked if the quality corrected itself as the call went on. Nobody is going to upgrade your data pathway if they don’t have to.

Were you doing this around 2AM in your time zone? That’s when some internet systems go into testing and backups. That’s when your YouTube stops working.

Also see: with everybody working from home, internet traffic has gone nuts and some services are failing to keep up.


That’s my guess too.

If the conferencing software detects that the network connection is not fast enough, it switches to a lower sample rate so that there is less data per second required over the network.

I wouldn’t. It is technically possible for the software to detect the data rate and dynamically change the bit-rate up or down. The Zoom conferencing software is modern and technically advanced, so it might do that. I think it’s less likely that older software such as Skype would do that, but that’s only a guess.