summaryrefslogtreecommitdiffstats
path: root/2022/organizers-notebook/index.org
diff options
context:
space:
mode:
Diffstat (limited to '')
-rw-r--r--2022/organizers-notebook/index.org31
1 files changed, 19 insertions, 12 deletions
diff --git a/2022/organizers-notebook/index.org b/2022/organizers-notebook/index.org
index 1b431945..60303abd 100644
--- a/2022/organizers-notebook/index.org
+++ b/2022/organizers-notebook/index.org
@@ -1606,7 +1606,7 @@ We’ve simplified the process down to these steps:
You might want to increase the size of the waveform by pulling on the
bottom of the bottom of the track.
-[[file:audacity-demo-resize.webm]]
+ [[https://media.emacsconf.org/misc/audacity-demo-noise-reduction.webm][audacity-demo-noise-reduction.webm]]
2. Find a moment of quiet in the video, and select it.
@@ -1624,6 +1624,8 @@ We’ve simplified the process down to these steps:
enough (Noise reduction (dB): 12, Sensitivity: 6.00, Frequency smoothing
(bands): 3).
+ [[https://media.emacsconf.org/misc/audacity-demo-noise-reduction.webm][audacity-demo-noise-reduction.webm]]
+
6. Tools → Apply Macro → Alpha
Before you can apply the Alpha macro, you need to save its content to
@@ -1637,19 +1639,21 @@ Normalize:ApplyGain="1" PeakLevel="-3" RemoveDcOffset="1" StereoIndependent="1"
Compressor:AttackTime="0.1" NoiseFloor="-50" Normalize="1" Ratio="2" ReleaseTime="1" Threshold="-30" UsePeak="0"
#+end_src
-7. Save the file to disk with libopus (.opus format)
+7. Export → Export Audio… → Opus Files (.opus format)
+
Use the following settings:
-#+begin_quote
-Bit Rate: 64 kbps
-VBR Mode: On
-Compression: 10
-Application: Audio
-Frame Duration: 20 ms
-Cutoff: Disabled
-#+end_quote
+ [[https://media.emacsconf.org/misc/audacity-export-settings.png][audacity-export-settings.png]]
+
+ #+begin_quote
+ Bit Rate: 64 kbps
+ VBR Mode: On
+ Compression: 10
+ Application: Audio
+ Frame Duration: 20 ms
+ Cutoff: Disabled
+ #+end_quote
-[[file:audacity-export-settings.png]]
**** Model-based denoising filter
If you can’t manage to get a good result with Audacity, chances are it’s
@@ -1666,6 +1670,9 @@ this means a model that minimizes background noise and maximizes clear
voices (the speakers’).
This is the model we’ve been using:
+
+[[https://media.emacsconf.org/misc/audio-denoiser-model-mp.rnnn][audio-denoiser-model-mp.rnnn]] (download link)
+
Source: [[https://github.com/GregorR/rnnoise-models][rnnoise-models]], Model: [[https://raw.githubusercontent.com/GregorR/rnnoise-models/master/marathon-prescription-2018-08-29/mp.rnnn][marathon-prescription]]
You should always apply the filter on the original’s audio, as opposed
@@ -1680,7 +1687,7 @@ Make sure to modify the ~DENOISER~ variable and adapt input/output.
DENOISER="/path/to/audio-denoiser-model-mp.rnnn"
input="original.opus"
output="denoised.opus"
-ffmpeg -i "$input" -af "$DENOISER" "$output"
+ffmpeg -i "$input" -af "arnndn=m=$DENOISER" "$output"
#+end_src
There’s no need to customize the libopus export information; the default