Hi Terry

. . . . until the first spoken word begins, the resulting playback is amplified enormously. I can practically hear individual molecules of air in my breath! . . .

This sounds like the auto gain control starting at max gain.

Has this just started happening or is it since trying the mic boost etc from your previous audio issue?

Does your AKG USB mic have an automatic gain control setting possibly through its drivers?

John EB

Hi Terry,

I presume that you are on MEP2017, correct? I'm using VPX, but the result should be similar.

Try turning on the mixer and watching the level meter when you play back the recording. Do you see a peak at the beginning, meaning the level jumps way up into the yellow or red? A bar will remain at the highest level.

What are your recording parameters? DAT, normalize after recording, stereo?

Have you tried different length recordings? I have found that with short recordings, about 5s, with normalize after recording on, the level comes out higher than a longer recording of 10s or more, and I have to adjust the volume downward quite a bit. But, I have not had your problem.

I do have different problem and that is sometimes I get a slight hum in a recording, none in another one recorded a few seconds later. 

Also, one reason for using Music Editor or AML is to do volume automation. I find it easier in these tools than in VPX/MEP. The reason for doing the automation or modifying the volume curve is to try to even things out. I find that some words and phrases come out louder than the rest of the dialogue, and I reduce the level at these points. They are easy to locate with a big waveform on the screen. These spikes affect the normalization of the audio as the spike sets the high point and the rest of the dialogue comes out lower. At least, that is what I found.

I have been struggling to understand audio for a couple of years now, and it is still difficult.

P.S. Yes, something has gone wrong with this forum interface and it is splitting words.

John CB

(Third attempt to post. Problem inserting image.)

OK, that worked this time. Hopefully self-explanatory. See also my reply to John EB.

I'm still using MEP 2016 but it's exactly the same in MEP 2017.

Terry, East Grinstead, UK

Hi Terry

. . . . until the first spoken word begins, the resulting playback is amplified enormously. I can practically hear individual molecules of air in my breath! . . .

This sounds like the auto gain control starting at max gain.

Agreed. Yet I can find no AGC or equivalent anywhere as you see.

Has this just started happening or is it since trying the mic boost etc from your previous audio issue?

It is recent and does indeed follow the playing around with all my settings.

Does your AKG USB mic have an automatic gain control setting possibly through its drivers?

If it does then so far it's eluding me! I've shown all the settings I assume relevant in my consolidated screenshot. Apart from that obscure message about 'further installation' I see no clues. Except possibly something relating to speech recognition, which I played with after your earlier post. (BTW, what's the logic in the 'Configure' button opening that window?) I just tried going in there again and note that on closing it the mic volume had been reset from 44 to 34 dB, if that's relevant. (I've changed it back to the necessary 44 dB.)

My workaround is to zoom and trim out that 'wait' section, but it's a bit of a nuisance.

Alternatively: https://www.dropbox.com/s/4iq2jfkvsvffvsu/AKG%20Recording%20AGC%20problem.jpg?dl=0

Terry, East Grinstead, UK

Hi Terry,

Probably nothing to do with your problem, but on the parameter screen in MEP/VPX, I use DAT to get 48kHz so that the file is the same frequency as that recorded by the camera and used in MEP (project settings). Also, on the second parameters screen, check Mono unless you are using a stereo mic, otherwise you will only have sound on one channel (probably left).

Hi John,

AKG mic is definitely stereo.

Just before having to leave my PC for the day, and following lots more playing with settings while making those screenshots, I appeared to make a recording which has a much lower volume level during that initial 'silent' period. If so, I've unknowingly made a change somewhere. A change for the better. I'll keep fingers crossed that this persists until I get back on the case tomorrow! And will check out that 44 v 48 issue.

First, that apparent improvement I described in my last post was premature. The symptoms remain the same if conditions like volume setting etc are unchanged.

But today I made what I hope is a helpful discovery.

I stopped playback at the start of a pause between sentences.When I then restarted playback the high amplification of my breathing was then heard clearly. I had thought this was only at the initial startup section, before I began speaking at all, because all intermediate pauses were then normal when played back continuously.

Does this help the audio experts to diagnose the issue please?

So this unwanted amplification only gets into the exported recording at the startup. But it's objectionable because even after deleting the breathing, the first fraction of a second of speaking is too loud.

My revised workaround until I resolve it is to start every recording by saying "Start", and then cutting that out after recording the rest, all of which is then OK. Still a pain.

And one more clue...

If I export the sort of recording we're discussing (i.e. with what should be a 'silent pause' at its start) then play that in MPC HC with that app set to continuously repeat, it's only in the very first playback that I hear the amplified breathing etc. All other repetitions are normal.

Terry, East Grinstead, UK

Hi Terry

. . . . it's only in the very first playback that I hear the amplified breathing etc. . . . .

Is the mic plugged in all the time?

It sounds to me like that your mic is continuously live and you are actually hearing your breathing live.

John EB

Hi John,

The mic is plugged in, yes, but nothing gets recorded until I click the record button!

And it seems from the additional info described in my last few posts that this is neither an MEP issue nor even a recording issue. It's about playback. I've called it a day now (and looking forward to a glass of wine) but tomorrow I'll get back on it.

One thing I'll try is playing an exported recording via my TV. I suspect the 'noisy silence' won't be present. Somewhere in the mysterious innards of either Win 10 or the associated built-in Realtek audio stuff there must be a setting that's bringing every audio output up to some sort of 'normalised' level.

Terry, East Grinstead, UK

Hi Terry

. . . . but nothing gets recorded until I click the record button! . . . .

I have a suspicion what you are now hearing is the mic picking up live sounds, I do not think they are on the recording given your comment:

. . . . . All other repetitions are normal. . . .

Are you using headphones to listen with? If you are then unplugging them and using the speakers should result in a feedback screech - I do not recommend doing this for safety reasons.

Try unplugging the mic or turning it down in the Stereo Mixer, IIRC you enabled this during the audio testing from the previous topic, and listen to the recording again.

HTH

John EB

Hi John,

I have a suspicion what you are now hearing is the mic picking up live sounds

Mic is certainly "picking up live sounds" - but only when it's been instructed to do so!

I do not think they are on the recording

You've lost me there! The initial section before I start speaking is most definitely on the recording. (How could it fail to be?) It's background and noise, including the sound of my quiet breathing.

Are you using headphones to listen with?

No. I'm listening to the speaker setup I described earlier. No 'feedback screech'.

I think the opening para of my first post still hopefully describes the symptoms reasonably clearly:
"Immediately after starting a mic recording, and until the first spoken word begins, the resulting playback is amplified enormously." together with the extra info in my 19/05/2017, 15:37 post.

Here's another example, this time playing back the exported recording in GoldWave:

(I exported a WAV, but also saved an MP3 as that's the only format allowed here.)

Note, as reported before, even that first section plays back correctly if the clip is allowed to repeat. It's just the first play of the first 'quiet' section that gets amplified. Why?

But maybe it's my misunderstanding and this is 'normal' and unavoidable? What do others find when playing a mic recording like that?

Terry, East Grinstead, UK

Hi Terry

It was this comment that prompted my suspicions

. . . . in MPC HC with that app set to continuously repeat, it's only in the very first playback that I hear the amplified breathing etc. All other repetitions are normal. . . . .

. . . . The initial section before I start speaking is most definitely on the recording. (How could it fail to be?) It's background and noise, including the sound of my quiet breathing. . . . .

If the second and subsequent playbacks when repeated are normal ie there is no breathing then I could not see how this was actually in the recording, until now.

Check that you do not have the Loudness equalisation effect in the Windows playback sound control turned on:

Right click the speaker icon, Playback devices, Speakers, Properties, Enhancement tab, Loudness equalisation

HTH

John EB

Gentlemen,

Well I'll be... I noticed this problem when playing back through other programs, like Groove Music (whatever that is and from where it came, I don't know). But, I didn't realize that it was a problem. The sound starts loud and then drops. I checked my settings and sure enough, I have Loudness Eq checked, even with my headphones. Unchecked, the problem went away. So thanks, John EB. I'll file that one away for future reference.

Terry, note that all of your recordings go somewhere, and that somewhere is indicated on the parameters screen. Mine go to: B:\Documents\MAGIX\Video Pro X\My Record

Take a look at what you have there. I have many with just a date because I was doing recording tests (sometimes not tests) before or without saving the project with a proper name. Not a good idea. The result is that I have hundreds of recordings that probably are not in any project or were not used, and I probably have some that were used in a project but were recorded before deciding to save the project with a proper name. I often do 2 or more takes trying to read a sentence without stumbling, so I have unused wave files sitting there. Thus I will have to pay more attention to housekeeping of these files in the future.

The point here, is that your recording is a wave file that is sitting in a folder. No need to export to wave again from MEP, just copy or use (or listen to) the original.

In case you haven't tried it yet, try Music Editor 3 by right-clicking on an audio object and selecting Edit wave externally. Use a test clip that needs some cleaning using tools that are more sophisticated than those embedded in MEP. Try some old recordings that have a background hum and try to reduce it. Play with the tools in ME3. You can actually save a project and open ME3 as a standalone and clean and master files, etc., with it. When you exit, save, and the edited file is now the one that you were using. Again, note where that file is located so that you can use it elsewhere.

Have a good evening!