I modified the code as you instructed, appending a note to the name to differentiate it from your original plugin, and tried it out. It perfectly clamped the regions below the threshold to perfect silence, allowing me to use the disjoin command to separate out the individual audio pieces for sliding about the timeline. This will speed my editing up greatly!
Thanks so much, stevethefiddle! Not being a LISP/Nyquist programmer, the obviousness of your simple code change was only apparent after the fact. I appreciate your effort!
This next bit is off-topic, but demonstrates a use case for Audacity. In case any reader my be curious what I’m up to, I’m using this with audio streams from the LiveATC.net archives. If you’re an aviation enthusiast, you might already know the site, which fosters a community of folks who capture and stream air-band communications from nearby airports. Students can find the site invaluable for learning how ATC controls traffic, and developing an “ear” for comprehending the often rapid fire instructions.
To follow the action where multiple control positions might be available for listening, it’s helpful to jump from channel to channel to hear how ATC commands manage the flow of individual aircraft. Often, communications overlap in time, so this modified filter allows me to build a timeline of several recorded ATC channels, and slide any overlapping individual transmissions to areas of silence in the other channels, while keeping relatively close to the original timeline. The project can then be mixed down into a single audio file, and you can then listen and build up a mental picture of the traffic being managed, and how ATC is guiding flights to maximize airport flow. This can be a helpful learning aid for student pilots who will eventually need add their voices to this din and understand and be understood by their fellow airmen.
Quite a lot of changes so we’ve skipped a version. Most of the changes are to the help files, but there are also a couple of significant changes in the code.
Gate to absolute silence:
This is the main new feature in version 2.4
To “gate to silence” set the “Level Reduction” to below -96 dB.
Due to the size restrictions on help screens in Nyquist, there are now 4 built in help screens. I’m quite tempted to rip them out altogether and just include a separate help text file. NoiseGate_version_2-4.zip (5.5 KB)
Noise gate 2-4 seems to work nicely: the hiss at the start of this recording is reduced …
Odd thing I noticed is the envelope of the “after” waveform looks slightly different from the “before” …
although I can’t hear a difference in the audio there are very slight changes in the before and after spectra of the speech.
NB: I did not select the “low cut”, which would affect the speech waveform,
the only operation applied was the noise gate which should not alter any of the speech, only reduce hiss in the “silence” at the start …
That’s because of the “Gate Frequencies Above” setting. You will always get that type of effect occurring when using frequency filters - perhaps it needs rewording to indicate that in most cases you would leave that set at 0 ?
No, the audible differences are much less than the visible difference because the visual difference is primarily due to changes in phase, whereas human hearing is relatively insensitive to the phase.
The audio above the noise threshold will be totally untouched if you are not using the filters (Gate Frequency Above set to 0).
Nyquist plug-ins are written in Nyquist code, with a simple graphical interface provided by Audacity (for more information, see: http://wiki.audacityteam.org/wiki/Nyquist_Plug-ins_Reference).
This interface is very basic and does not provide the ability to change menu presets. We are hoping for more versatile interface options in a future version of Audacity, but that is not likely to be in the near future unless we acquire a C++ developer that is particularly interested in developing this part of Audacity.
I don’t have a clue what I’m doing wrong, but no matter how I set the settings, I don’t get silence. I downloaded and installed the new version but still can’t get absolute silence like I get clicking the silence button. I feel stupid and frustrated. I assume this works but I can’t get it to work.
This version has been fully updated, so it requires a fairly recent version of Audacity.
It has all of the features of the version on the wiki, and uses exactly the same gating engine, but in my tests it is about twice as fast, and hopefully will handle long tracks better (I’ve tested with over an hour of audio with no problems).
The one thing it hasn’t got is built-in help. This is because I’m hoping that it will be included in the next Audacity release, in which case the help button will link to the manual.
Please give it a good test, and leave feedback in this topic. noisegate.ny (5.52 KB)
I have been unnaturally successful at getting the previous version (2.2 ??) to work as a final pass at noise reducing a voice performance. Are the new action settings more or less similar going forward?
I used gentle Noise Reduction of the Beast (6, 6, 6) on a mastered performance which gets the noise values separated from the show enough so Noise Gate has good elbow room to do its thing, but not so stiff that it turns words into honky, bubbly trash. I used Gate at only 12dB reduction which avoids the “hissing words” pumping problem and between the two tools can turn a relatively noisy (and failing) voice performance into a clean, mastered chapter with -72dB noise floor.
Imagine my surprise. ACX apparently can’t tell what happened.
Always a plus.
You still can’t record in a boiler factory, but with a little head start, you don’t need the home studio you did before.
So long as the new version doesn’t have bugs. If you have time to play with it / test it, then please do, and let me know how it goes.
I’d be particularly interested to know if the performance improvement that I see on Linux is also on Windows / Mac.