Audio dropout - any ideas?

[wireless]

This is a snapshot of an audio recording via external microphone and audio interface and viewed in the Sequencer in Pitch Edit mode. It looks like some kind of dropout has occurred during the recording, since a section of the audio has gone AWOL!
It evidences in the playback as a slight hiccup in the audio.

I haven't come across this before, but suspicion must fall on my recent drop in the size of the audio buffer from 512 samples to 128 samples in an effort to reduce the latency being heard between recorded instruments - initial trials suggested that this would work. Under Windows, one never knows when the computer's interest will wander to some other less important task....

Audio dropout highlighted.jpg (177.33 KiB) Viewed 696 times

Any ideas as to whether or not this is repairable? I was wondering about a) slice the missing section followed by a slight stretch to maintain the beat or b) cut and paste from an adjacent section?

Luckily it's not valuable and I could just re-record in this instance, but from a techniques point of view I'm interested to find out if there are any new skills I could learn....

Thank you

[selig]

It’s difficult to say what’s going on with the audio since all you’ve posted is a shot of the pitch edit page (and the audio file is TINY in that picture).

But it doesn’t look like a complete dropout of audio AFAICT.

As for a fix, a combination of a short time stretch and a short crossfade should be able to repair it I would think.

Pitch edit may still take issue with it, so it may take a few tries if that’s the case (and pitch edit is important for this track).


Sent from some crappy device using Tapatalk

[wireless]

Thanks for thoughts Selig - much appreciated.

Closer inspection reveals that it's somewhat stranger than a dropout - maybe more phase related? Perhaps variation in mouth position to microphone?

The snapshot shows the audio above the pitch edit screen, with the 'issue' apparent at location of the time cursor. The pitch edit screen is greyed at that section, initially leading me to suspect a dropout, but the audio waveform is trying to tell a story too. That part of the vocal was slowed right down and the 'effect' is audible as a kind of phase shift rather than a dropout. I guess some experiments with the mic are in order (sE Electronics sE5 condenser mic).

Audio dropout highlighted 2.JPG (39.28 KiB) Viewed 672 times

[selig]

Thanks for thoughts Selig - much appreciated.

Closer inspection reveals that it's somewhat stranger than a dropout - maybe more phase related? Perhaps variation in mouth position to microphone?

The snapshot shows the audio above the pitch edit screen, with the 'issue' apparent at location of the time cursor. The pitch edit screen is greyed at that section, initially leading me to suspect a dropout, but the audio waveform is trying to tell a story too. That part of the vocal was slowed right down and the 'effect' is audible as a kind of phase shift rather than a dropout. I guess some experiments with the mic are in order (sE Electronics sE5 condenser mic).

Audio dropout highlighted 2.JPG

OK, it’s a vocal right? I’m going to go out on a limb and guess it’s a condenser mic, possible a large diaphragm condenser. If that’s the case, it could well be that what you’re seeing is a response to humidity from your voice.

This is something I’ve seen before, and if that’s what it is there are a few ways to help eliminate it. For one, work in a dry environment when possible (no open windows in a humid part of the world). Also, use a fabric pop filter.

Finally, sing “across” the microphone rather than directly into it, especially when working close to the mic. To do this, imagine starting with the mic directly in front of your mouth (6” or so away). At this point we can say the mic is at the 12:00 position on an imaginary clock (when viewed from above). Now move the mic from the 12:00 position to either a 10:00 or 2:00 position, rotating it around the imaginary circle so that it is still aimed directly at your mouth. Some engineers even suggest that many singers sing more out of one side of their mouth than the other - if so, move to that side!

And if this is a humidity issue, it will be difficult to “prove” you’ve solved it with only one take. You’ll have to judge the success or failure of this over time.

Alternatively, there could be some issue with the mic or preamp along the lines of a failing component (tube?).

But from what I’m seeing it’s unlikely it’s a buffer issue (that would be more like a razor cut with totally missing audio), or a cable issue (which would show static or spikes in the waveform).

One other thing to check to see if it could be a humidity issue, is to play that section while viewing the spectrum display. If there is a large low frequency “event” at the dip, it could be a humidity issue (or also a component issue).

Maybe others have additional input, corrections, or clarifications, which I welcome!


Sent from some crappy device using Tapatalk

[wireless]

Hi Selig,
Many thanks for taking the time and for such a considered reply. I knew when I asked that the problem was a bit esoteric, and so all the more thanks for responding.

Yes - all you've said makes sense. I know that I should be using a pop filter but am not as yet - other things have been competing for funds. I guess I need to prioritise this some more. Down a few levels in my sub-conscious I'd been considering a Vocal Booth like the one below, or the Aston Halo as they are often bundled with a pop shield.

Great idea about singing 'off axis' and I'll give that a go.

Vocal booth.JPG (54.05 KiB) Viewed 596 times

Regards