User:BirdHopper

From Wiktionary, the free dictionary
Jump to navigation Jump to search
Wiktionary:Babel
en This user is a native speaker of English.
/ʑ/
IPA-1
This user has a basic understanding of the International Phonetic Alphabet.
Search user languages or scripts

About Me

[edit]

I recently became very interested in phonetics and IPA transcriptions. My knowledge is very limited so far, but I learn more each day. I figured Wiktionary would be a good place to practice the skill and motivate me to get better at it. If you happen to be an IPA expert and take a look at my edits, I would really welcome some constructive feedback.

In addition to transcriptions, I'm adding audio pronunciations to pages that need them. I have a "Snowball iCE" microphone, which is pretty decent, so I figured I may as well use it for something. I use Audacity for all my editing. Doing all my work in Audacity is a little tedious, but worthwhile since it produces superior results compared to tools like SoX, which has poor noise reduction capabilities. To ease the workflow, I set up a simple "chain" consisting of most of the steps I do each time I record a word. Here is what I do.

Before each recording session, I create a noise profile. I record about 10 seconds of ambient noise, select it, go to Effect > Noise Reduction, and click Get Noise Profile. Next, I start the repetitive process of recording words.

  1. Record myself saying the word several times
  2. Apply custom "Word Pronunciation" chain of commands (go to File > Edit Chains to set one up)
    1. Noise Reduction (15 dB, sensitivity 12, frequency smoothing 6)
    2. Bass and Treble (bass -8, gain 2)
    3. Normalize (usually -4 dB; totally arbitrary)
    4. Sound Finder to isolate each word (silence level 26, duration 0.1 sec, label before duration 0.1, label after duration 0.2)
  3. Trim out the best sounding pronunciation by clicking one of the labels produced by Sound Finder and hitting Ctrl-T
  4. Select the "silence" before (and after) the audio and do Generate > Silence, because noise reduction isn't perfect
  5. Fade in and out from the silence for a smooth transition to and from the audio
  6. Select the audio and export selection as ogg

You'll have to play around with the settings, as each microphone and environment is different. The command chain really sped up my workflow. To make it even faster, I added some keyboard shortcuts: Alt-Shift-S for adding silence, Alt-Shift-< to fade in, and Alt-Shift-> to fade out. Feel free to contact me with any questions, comments, or concerns. I'm always wanting to improve. Also, I'm happy to elaborate on my Audacity workflow if anyone wants me to.

[edit]

This is a link to stuff I'm working on at the moment.