Tag Archive for technical issues

Learning Hurts My Brain

I’m spent more time than I want to admit (at least to my writing self) working through the basics of creating a reasonable vocal track.  God, I’m even using terms like “vocal track” and “gain.” Don’t be fooled.  I am stilling scraping forward to even a rudimentary understanding of what the heck I’m actually doing.

This then is my final pass.  I’ve recorded half of J. M. McDermott’s “Death’s Shed,” editing as best I can.  I will record the second half this weekend, using my current level of middle-school-boy finesse.

So what are my current “best practices” and why am I even telling you my theoretically patient reader?  Basically, I am terrified I’ll forget how I managed to produce even this mistake riddled track.  Except for the softness (which I can’t seem to fix without distorting the audio), it sounds pretty good.  Could be delusion on my part, of course.

Death’s Shed by J. M. McDermott

How I Arrived at This Fabulous Production Value

Mic: Snowball.

Location: Mudroom.  The mic sits on the far right corner of the desk.  The laptop, prone to long outbursts of crazy-burring fan-noise, is pushed as far to my left elbow as possible.

Me: About a foot and a half from the mic.  My khaki-green water bottle sits at my feet.  A thick orange extension cord runs from the desk, through the bottom crack in the door into the kitchen.

Software: Audacity 1.3 for Windows with the Spitfish plug-in and the LAME encoder installed.

Settings: I start with a gain of 9 while recording.  After I finish recording, I select a stretch of silence and run the denoise effect, then I run the amplify effect setting the peak to -0.1.   Then I run the Spitfish effect to try and fix my sibilance.  Finally, I adjust the gain so that the monitor shows the track running at a max of about -0.6 (close to the right edge without any red lines appearing).

Tip from Jim Kelly:  I record as a single track. Whenever I make a mistake, I pause for a long stretch so I will be able to find and delete the error later, then I resume speaking.


Audio Quality Tests: Yes, It’s a Testing Tizzy

My first recording seemed like it went well–for a first recording.  And really that’s the problem.  If SB Press (and later on, perhaps, just little ol’ me on my own blog) publish audio, it needs to be better than a “good first attempt.”

So I went ahead and purchased a snowball mic.  Jim Kelly, podcaster extraordinaire, agreed to listen to my first attempts with the old mic and give me an assessment. (Yes, he is a terrific guy.)  He noticed two big things 1) too much background noise and 2) my s’s were far too prominent.   He suggested I adjust my distance from the new mic to see if it helped.

So here are the three sample tracks  1) old first attempt (old mic) 2) new mic at a distance of 1 and 1/2 feet and 3) new mic at a distance of 3 feet.

1. Original track (snippet).

Death’s Shed by J. M. McDermott.

2. Snowball mic at 1 and1/2 ft (snippet).
Death’s Shed by J. M. McDermott.

3. Snowball mic at 3 ft (snippet).
Death’s Shed by J. M. McDermott

My Two Questions:

How is the background noise?

Have I tamed those s’s enough?
I think not. I’m trying to figure out how to work an Audacity plug-in called Spitfish that de-esses tracks, but so far no luck….

I really had no idea I had this “s issue” until yesterday.  Now I hear it every time I speak!

More Technical Details of My Recordings Than You Really Want to Know:
So here’s the current issue, Spitfish isn’t picking up any sound to de-ess. I think it’s because my recording levels are too low. When I record with the input setting to maximum in Audacity, the recording level hangs around -24 decibels (not the -6 decibels suggested in the user manual). The audio peaks seem to sit somewhere between 0.15 and 0.2 (not the 0.5 “ideal”).

I’ve got my mic on setting #1 (for single voice recording). Because of the sibilance issue, I’m experimenting with recording between 1 and 1/2 ft to 3 ft from the mic with a gain setting of +10 (tracks samples above).