[BUG]

[jeffmielke]

Debugging checklist

[x] Have you read the troubleshooting page (https://montreal-forced-aligner.readthedocs.io/en/latest/user_guide/troubleshooting.html) and searched the documentation to ensure that your issue is not addressed there? [x] Have you updated to latest MFA version (check https://montreal-forced-aligner.readthedocs.io/en/latest/changelog/changelog_3.0.html)? What is the output of mfa version? [x] Have you tried rerunning the command with the --clean flag?

Describe the issue A clear and concise description of what the bug is.

This may be more of a misunderstanding than an issue. I would like fewer silence intervals to be inserted during the closure phase of fluently-produced voiceless stops, and I was expecting that setting silence_probability to a value less than 0.5 would do this, but all values appear to increase the number of new silences. Here is an example of the number of silences added in one tier of one textgrid with different values:

default settings: 22 silences are added silence_probability=0.0: 31 silences are added silence_probability=0.3: 31 silences are added silence_probability=0.5: 31 silences are added silence_probability=0.9: 39 silences are added

I expected that 0.0 might produce no new silences and 0.5 would produce 22 new silences since 0.5 is the default. So I don't understand what I'm seeing.

For Reproducing your issue Please fill out the following:

1. Corpus structure
   • What language is the corpus in? English
   • How many files/speakers? 8/8
   • Are you using lab files or TextGrid files for input? textgrid
2. Dictionary
   • Are you using a dictionary from MFA? If so, which one?
   • If it's a custom dictionary, what is the phoneset? arpabet
3. Acoustic model
   • If you're using an acoustic model, is it one download through MFA? If so, which one? english_us_arpa
   • If it's a model you've trained, what data was it trained on?

Log file Please attach the log file for the run that encountered an error (by default these will be stored in ~/Documents/MFA).

no errors

Desktop (please complete the following information):

• OS: [e.g. Windows, OSX, Linux] Linux
• Version [e.g. MacOSX 10.15, Ubuntu 20.04, Windows 10, etc] Ubuntu 20.04
• Any other details about the setup (Cloud, Docker, etc)

Additional context Add any other context about the problem here.

[mmcauliffe]

Also one thing to note is that there is also the --boost_silence parameter which should control the acoustic likelihoods generated. The --silence_probability parameter is only affecting the lexicon/decoding graph, so it might also be worth playing around with --boost_silence 0.1 (--boost_silence 1.0 is no boost, less than 1.0 is reducing silence likelihoods, and greater than 1.0 is boosting silence likelihoods).

[jeffmielke]

Thanks. I didn't know about that one. I'll try it!

On Tue, Jul 16, 2024, 2:35 PM Michael McAuliffe @.***> wrote:

Also one thing to note is that there is also the --boost_silence parameter which should control the acoustic likelihoods generated. The --silence_probability parameter is only affecting the lexicon/decoding graph, so it might also be worth playing around with --boost_silence 0.1 (--boost_silence 1.0 is no boost, less than 1.0 is reducing silence likelihoods, and greater than 1.0 is boosting silence likelihoods).

— Reply to this email directly, view it on GitHub https://github.com/MontrealCorpusTools/Montreal-Forced-Aligner/issues/824#issuecomment-2231865515, or unsubscribe https://github.com/notifications/unsubscribe-auth/AH3Q3BIVZLDGXEALL7P6533ZMWG2DAVCNFSM6AAAAABKXIPDBOVHI2DSMVQWIX3LMV43OSLTON2WKQ3PNVWWK3TUHMZDEMZRHA3DKNJRGU . You are receiving this because you authored the thread.Message ID: @.*** com>