Fledgling voice artist seeking counsel

Special Interest Groups Audiobook Production
101

I meant to do this earlier.

Prepare a show normally plus leave a count of three or four “room tone” at the beginning. This is you staring blankly at the microphone and making no noise at all. You can do this at either end of the performance, but you will forget to do it at the end.

Past maybe a gentle volume change, I expect #1 is the only audible step if the show is recorded in studio conditions. The test protocol should only take a couple of minutes if that — if you pass. If the show falls off the rails anywhere in the process, then we enter rescue mode which could take days.

The steps depend on each other.

\

  1. Apply Steve’s High Pass filter.
    — low pitched sounds can be almost inaudible and take up a lot of blue wave room. Get rid of those immediately. I don’t agree with the 130Hz High Pass filter idea. That would have the exact opposite affect of those commercials to increase your manhood.

  2. Normalize to 0.
    — An interim step. We assume any DC vanished in #1.

  3. Run the RMS inspector.
    — Hopefully, you pass RMS with 3dB to spare and can Effect > Amplify to -3 and go straight to #4.
    — Assume RMS came out OK, but not great, so you have to leave it there. Then apply gentle peak limiting, the show peaks not to exceed -3.

  4. Run inspector for noise level.
    — Everyone is going to fail this one and I’m not sure where to go from here. This could point to the four volume set “How to reduce noise.” This will almost certainly point off-page.

  5. Cut off any beginning and ending pops caused by #1 and the room tone silent segment.

  6. Listen to the work and if it still sounds like you reading your story, Export a WAV for your own show archive and then export a submission for publication approval.

Koz

102

Also: It would be nice if you could record a single impulse e.g. by clapping, hitting some hard material (like two stone bars) or something similar.
This can be used to either analyse, attenuate or boost the room reverberation.

103

How about two #2 wooden pencils? It shouldn’t be too difficult to round up two of those and their impulse is not likely to overload the microphone.

“We need to be careful not to get Mars and Venus mad at us or they’ll stop sending us candy bars and pencils.”
- Glendora — Afternoon Kids TV show -

This can be used to either analyse, attenuate or boost the room reverberation.

I’m wondering where you’re going with this since the room of a newbie is going to echo and we can’t get rid of echoes in software.

Also, can we agree that echoes are a sound quality issue between the performer and ACX? I point to bgravato trying to record his guitar with a single microphone over weeks and this very thread threatening to overtake Bruno’s 18 forum chapter record. We passed technical standards full steam a while back and now we’re working on sound quality issues.

Who’s going to write the Pencil Tick Analysis Tool?

Koz

104

Anything that produces an as narrow as possible impulse, some use even a pistol or a balloon.

Says who? It is done all the time in mobile devices. However, in communication systems, it is blind deconvolution whereas we would have the impulse response of the room.
I actually do not want to remove the echoes though, rather the coloration superimposed by the reverberation.

Do you dare to make a guess in the wild?
We have already the plot spectrum analysis tool. Steve has even written a plug-in that converts the exported data to an importable EQ-preset. That’s all about sound quality imo, since it will give the voice more presence and clarity in the best case.

105

That’s all about sound quality imo, since it will give the voice more presence and clarity in the best case.

Yes, but divorced from ACX sound file standards.
Maybe after you achieve ACX compliance, then a second pass for theatrical quality. I did build that into the last step.

Anything that produces an as narrow as possible impulse, some use even a pistol or a balloon.

Both of which would overload the microphone, sound channel or both. The test result would be a combination of room characteristics and damaged microphone the latter not part of the actual performance and invalid.

Koz

106

A popping balloon might destroy a ribbon microphone notoriously sensitive to sudden massive air movements. Koz

107

The impulse source hasn’t to be in front of the microphone, for this purpose finger-snipping or the two pencils are enough.
Louder claps should raher be made where the microphone catches least direct sound, i.e. behind it. Balloons and such things are better suited for large rooms such as a concert hall.

There are other possibilities if you use a mobile player. Such as white noise, noise in one third/octave steps or following the bark scale, chirp tones, or an digital impulse train. They have the advantage that they can exactly be subtracted from the recording and that they originate at the mouth position.
The disadvantage is that characteristics of the player itself are transferred as well.

108

Greetings all,

I hope you are all well and for those in California, surviving this sweltering heatwave.

Thanks to birthday presents I received an extra box of the Auralex tiles and the Auralex Mudguard. I was able to finish the back wall with tiles, put some on the door and even some on the bottom of the shelf which is above my microphone. The whole room has tiles now. I’ll take some pictures tomorrow.

The Mudguard arrived today. I was out for most of it showing my visitor around and so just completed the tests which were requested of silence, speaking, clapping and pencil tapping.

Unfortunately, the Mudguard did not perform the wonder I had hoped. The buzzing noise persists. Also, when I resorted to taking the laptop out into the hall and moving it as far from the audio closet as possible, despite the microphone being in the same place and the volume on Audacity still set the same, I can barely be heard in the recording AND the buzzing is still present. I tried without the USB extension (the original just about reaches) but no difference.

Ugh, this is frustrating.

My visitor taught me about something called Chokes which were used before HDMI came along. Does anybody know if they would silence a USB cable so we could rule that out?

Anyway, for what it is worth, the tests are here on my OneDrive:

https://onedrive.live.com/redir?resid=FF7251924FF7642D!43644&authkey=!AK_9aA0m-UdA7V4&ithint=folder%2c.wav

Thank you.

Edit: Unsurprisingly, my insomnia has a strong hold on me tonight. Tis now 04:14 and I just had a lightbulb moment, something I ought to have thought of much sooner. I dug out the microphone attachment for my old gaming headset and recorded the room with that. I’ve added it as file 5.

I don’t hear the buzz.

This must surely mean it is the USB cable or my Blue microphone itself, right?

109

It’s not uncommon for USB leads to have ferrite chokes in-line:

Looking / listening to your recordings, even on the quietest one the peak level of the clap is trying to punch its way well over 0 dB. This is one of the inherent problems of USB microphones, they need to be able to handle quite high signal levels because some users will be recording loud things, but that means for low level spoken word recordings, the level is rather a lot lower than desirable, so the noise floor is relatively higher. If you were singing loudly a couple of inches away from the mic, then relatively, the noise level would be almost insignificant.

110

Thank you, Steve. I’ll see if I can find an inexpensive one.

What did you think of the recording with my headset microphone?

Cheers.

Edit: This will arrive on Friday. I really hope this sorts it out and is the last money I have to spend.

http://www.amazon.com/gp/product/B003MQ29B2/ref=oh_details_o00_s00_i00?ie=UTF8&psc=1


Cheers.

111

A really stupid question, but are you sure you didn’t record the internal Laptop mic instead of the Snowball?
Firstly, it sounds exactly as with mine, device noise and all. Secondly, the forth example (extern) sounds as if you’ve recorded behind the door.
My second guess would be that the audio comes in over line-in instead of mic line, but it is actually over USB, so there’s naught to this assumption. I wonder if there is any setting we haven’t set up properly yet. The main hum (60/120 Hz etc) is very prominent.
My Usb does not suffer from this problem, instead it has a constant DC offset, therefore, I somehow doubt that this cable is the source.
Does running on battery/plugging in the laptop change anything?
I think the clapping won’t give us much information as long as we are not able to get rid of the buzz.
However, the booth itself seems to be very good. There are some spectral peaks at 6, 20, 60 113 120 and 157 Hertz but I don’t know if they are all due to the interference. It seems that the place will be perfectly suited for your narrations eventually.

112

I guess, you are not performing Flamenco accompaniment as a rule?
Your hand claps are what a Spanish would call “Sorda”, i.e. dark. But we need a tight sound. This “Clara” is produced by clapping 3 or more fingers on the palm of the other hand. It is like praying but with one hand shifted down.
Here comes a comparison between your clap and mine.

After normal time , the claps are repeated and stretched by the factor 10.

Take especially note of the frequency development of the two claps.
My sitting room colours the spectrum towards the highs, whereas your clap stays practically in the same (low) region.
This is actually good since we don’t want coloration. However, as I said, the impulse has to be tight or “bright” to include all frequencies equally.

In the next file, I’ve applied my impulse response to one of Koz recordings. The goal would eventually be to get the first spectrum by eliminating the room characteristics from the second sample and thus to give more presence to the voice.
It is of course not a proper before/after comparison since there are two room impulses that clash. Also, my recording isn’t high quality and there are a lot of noise sources around me (the roaring of the fire for instance).
But I hope you’ll get the idea.

113

Robert!

There are no stupid questions, there ARE however, fledgling voice engineers / actors who forget amidst all the moving about and switching to a whole new room to re-disable the built-in microphone!

That is exactly what it was! The built in was also recording. Wow.

I’ll still use the Choke since it arrives tomorrow and wasn’t expensive but please have a listen to the new test I uploaded to the same folder. I think Robert has squished the buzzing insects!

https://onedrive.live.com/?cid=ff7251924ff7642d&id=FF7251924FF7642D!43644&ithint=folder,.wav&authkey=!AK_9aA0m-UdA7V4

Cheers.

114

Ahh, what a relief.
There fit whole worlds inbetween those recordings.

I think you’re ready to go and to record the “real thing”.

There are some frequencies I would cut (remember the boominess I complained about weeks ago?).
It is now really near to perfection and you might only notice the difference in direct comparison.
I’ll pick the 1 kHz/1.25 kHz region as an example, how I would search for offending frequencies.
I open the equalizer and boost some frequency under suspicion extremely high. You’ll immediately hear the resonance in the preview.
You decrease the dB value until the resonance disappears (and a little bit further).
This real time recording shows how I do it (in the equalization dialog)

The A/B comparison is at the very end.
You can try it yourself, for example with 315 Hz.
The highpass filter would of course be applied first (at about 100 Hz, 6 to 12 dB roll-off).
I’m looking forward to listening to your first chapter.

Ciao
Robert

(and yes, you’ve guessed my second name right enough…)

115

Your hand claps are what a Spanish would call “Sorda”, i.e. dark. But we need a tight sound.

Are we up to two pencils yet? Easily repeatable sharp impact with common objects that doesn’t overload the microphone or amplifiers.

The earlier problem was caused by an incompletely damped microphone enclosure causing the “talking in a box” sound. That should be nearly completely gone with the new configuration.

Probably the most discouraging thing I found was frying mosquito noises from an interface that doesn’t use a USB cable.

That’s not good news. That’s where I got my sound sample from.

We should be careful not to “tune out” Fledgling Voice Artist’s natural vocal tones.

The comparison might be between a free-air snowball which will accumulate sloppy room echoes, but probably no tonal problems, and The Studio.

I have a non-sound project I need to pay attention to.

Koz

116

Greetings,

well, my friend has returned to England and I have finally recorded some real work.

https://onedrive.live.com/redir?resid=FF7251924FF7642D!42773&authkey=!AEacDqviT__VLho&ithint=folder%2c.wav

I still need to do more work on my own glottal pops but otherwise, I think the only problem I have is how fast the room warms up. Heh. I have to step out rather often.

As always, I look forward to your expert feedback.

Cheers.

117

Oh, that is coming along, isn’t it?

The second Chapter 1 still has that hum in it, but it responds perfectly to Steve’s equalization filter.

Even better. It sounds like you and perfectly natural. No honk or other environment problems.

I think it’s submission ready (given the hum filter). The peaks are at -3, the background level is -60 (by my measurement), etc. etc.

Oddly, I didn’t write down the location of that filter. !@#$%

Searching…

~~

Oh, and welcome to the LA heat wave. Those of us without air conditioning usually end up at The Mall, Public Library or McDonalds. My west-facing bedroom stays cool for a while, so it’s late afternoon before I have to escape.

Koz

118

Anyway. Here it is attached.

Launch the show and then Effect > Equalization: Save-Manage-Curves. Import.

“LF rolloff for speech.”

Koz
LF_rolloff_for_speech.xml (299 Bytes)

119

Thanks, Koz! :slight_smile:

I’m so excited now. I’ll be auditioning on Voices.com like mad and continuing with the book in between. :slight_smile:

And yes, the bloody heat wave. My mate was here for 10 days so he, unfortunately, had to suffer some of it too. He almost melted.

Thankfully it has cooled off again for the moment.

I have A/C in three rooms (including the home theatre room) so please, feel free to come by. I’d love for you to see the actual audio room anyway. There might be things which spring to mind that I could make even better. Plus I’d be able to put a face to the name. :slight_smile:

Cheers.

120

Greetings,

I have added a new processed version and I have also sent it off to ACX for their opinion.

These are the steps I took:

  • Record in mono.
  • Normalize to 0.
  • High pass filter.
    80 Hz
    12 dB
  • Noise reduction:
    12
    0.00
    200
    0.00
  • EQ base boost
  • Add Steve’s Limiter
    -3
    10
  • Amplify to -3 dB.
  • EQ LQ Rolloff for speech import
  • Export as WAV.
  • Export as MP3.
    192 kbps CBR.

Are these still the steps all would agree on now that I have the room sorted and the angry insect buzz annihilated?

I know the EQ base boost isn’t technically needed but I feel it gives my voice a little more warmth (what do you think?).

Cheers.