Try again the sound of the audio was not clear

[jeaprk0504]

testfile : testpro.py wav file : examle - suun3.wav

I used your example... but I have some trouble...... Try again the sound of the audio was not clear......

$ python3 testpro.py [<parselmouth.Data object at 0x7f8c43061030>] Try again the sound of the audio was not clear [<parselmouth.Data object at 0x7f8c4116ad18>] Try again the sound of the audio was not clear [<parselmouth.Data object at 0x7f8c41172d18>] Try again the sound of the audio was not clear [<parselmouth.Data object at 0x7f8c05ec50a0>] Try again the sound of the audio was not clear [<parselmouth.Data object at 0x7f8c050321f0>] Try again the sound of the audio was not clear [<parselmouth.Data object at 0x7f8c0def68f0>] Try again the sound of the audio was not clear [<parselmouth.Data object at 0x7f8c43061030>] Try again the sound of the audio was not clear [<parselmouth.Data object at 0x7f8c4116ad50>] Try again the sound of the audio was not clear [<parselmouth.Data object at 0x7f8c41172d50>] Try again the sound of the audio was not clear [<parselmouth.Data object at 0x7f8c05ec50a0>] Try again the sound of the audio was not clear [<parselmouth.Data object at 0x7f8c050321f0>] Try again the sound of the audio was not clear [<parselmouth.Data object at 0x7f8c0def68f0>] Try again the sound of the audio was not clear [<parselmouth.Data object at 0x7f8c43061030>] Try again the sound of the audio was not clear [<parselmouth.Data object at 0x7f8c4116ad18>] Try again the sound of the audio was not clear [<parselmouth.Data object at 0x7f8c41172d18>] Try again the sound of the audio was not clear [<parselmouth.Data object at 0x7f8c05ec50a0>] Try again the sound of the audio was not clear [<parselmouth.Data object at 0x7f8c050321f0>] Try again the sound of the audio was not clear Try again the sound of the audio was not clear Try again the sound of the audio was not clear

I think myprosody/dataset/essen/myspsolution.praat have some warning

<line no. 245> int = Get number of intervals... 2 if int<2 warning$="A noisy background or unnatural-sounding speech detected. No result try again" appendInfoLine: warning$ exitScript() endif

int is '1'... when I run testpro.py, myspolution.py executes run_file function. but run_file returns warning.... so exception is enable....

How can i fix it???

[mbcse]

any solution found

[Shahabks]

<line no. 245> int = Get number of intervals... 2 if int<2

These lines assure the quality of sound-transforming to word-chunks . I cannot speculate why you run into that error when you simulate the sample ...... I need more information to diagnose you issue.........However, if you run to the mentioned warning, perhaps the audio files come with a noisy background, or the audio files are too short, or the files do not contain spoken words or contain stuttering (repeated words)

[cburst]

Hi, I'm running into this issue as well. I just wanted to test this script out. The recording sounds quite clear to me, it's about 20 seconds long. Is that not long enough? https://www.dropbox.com/s/9ymtqgzg1ps1p4o/wecopy.wav?dl=0

[Shahabks]

Sure, 20 secs long audio is long enough to be analyzed. Generally, if the algorithm does not pick up spoken languages in the audio (e.g. noisy backgrounds, sounds of non-spoken languages, repeating short phrases) it will give that warning out. To detect a spoken language, it needs at least 10 to 12 seconds of speech

[cburst]

Thanks for the prompt reply. Unfortunately, I'm still getting that error. Is there anyway to debug the script so that I can see where it fails? I'm not really a python expert to be honest (it actually took a lot of work to get the script to run at all; i.e., manually identifying the sound device etc).

[Shahabks]

Are your audio files in *.wav format with the specifications mentioned in ReadMe? ... From what you said about "..manually identifying the sound device" make me think that yet your sample audio file needs to be compatible

[cburst]

Thanks again for the prompt reply. First, all audio files are generated using the script. Secondly, I have tried 42 khz and 48 khz at 16 bits with no success. Should I try a different frequency/bit rate combination? 323-1147.wav.zip

[cburst]

perhaps i'm missing some dependencies? i noticed that the my-voice-analysis library had not been installed (though the script is still not working for me, it did take longer to fail though)

[cburst]

also, my-voice-analysis works GREAT! (tried with a 44 khz file as indicated) perhaps this is all I need for my project. i still want to get SpeechRater working though.

[cburst]

also, tried 44 khz 24 bit, still not working

[cburst]

ok, i figured it out, all of the language models were missing, managed to get most of them from the myprosody github, but still missing the my_classifier.pickle where might i find that?

[Shahabks]

I hope your problem is solved. Please let me know if there is any other

[bagustris]

The problem still exist for me, although the wav file clear enough.

[Shahabks]

it could happen when the length of recorded audio is short ..... If you want to capture short audio files (less than 10 sec., if my mind goes right) you need to change the filter parameters

On Wed, 10 Jun 2020 at 11:18, Bagus Tris Atmaja notifications@github.com wrote:

The problem still exist for me, although the wav file clear enough.

— You are receiving this because you modified the open/close state. Reply to this email directly, view it on GitHub https://github.com/Shahabks/myprosody/issues/8#issuecomment-641678895, or unsubscribe https://github.com/notifications/unsubscribe-auth/AGTX33X2CEQA56OD27XZJG3RV3UNPANCNFSM4IS6KDFA .

[ahalyaj]

I too had the same issue regarding my audio file. And I constantly get this message("Try again the sound of the audio was not clear") after testing every functions (for eg: mysp.myspsr(p,c)). Can you explain the reason why I get this issue even though the audio is in wav format and not of short duration? Can you please check my audio file, and tell if it works for you?

my-voice-analysis.zip

[felixbur]

Actually i had this problem and what helped me was to copy the myspsolution.praat file to the folder containing the audio. I guess the error message could be improved.

[tux103]

Hi worked it out you have to remove the file extension on the file name. I changed the file name form "myrcording.wav" "myrcording"