Some voices in 10.8 don't support [[inpt TUNE]]?

I was hoping for confirmation that some of the newer tts voices with OS X do not recognize the [[inpt TUNE]] type of fine tuning.


For example, from a commandline:


say -v Alex "example should say 'testing.' [[inpt TUNE]]t {D 80; P 107.8:0 152.4:94} 1EH {D 115; P 141.6:0 133.6:4 128.1:9 123.6:17 124.3:26 140.6:91} s {D 70; P 141.1:0 188.3:86 190.4:93}

t {D 60; P 184.4:0 167.5:92} =IH {D 75; P 159.3:0 146.6:13 137.4:27} N {D 90; P 101.2:0 95.7:11 93.8:22 94.6:89} [[inpt TEXT]]"


works as expected, but the following does not:


say -v Moira "example should say 'testing.' [[inpt TUNE]]t {D 80; P 107.8:0 152.4:94} 1EH {D 115; P 141.6:0 133.6:4 128.1:9 123.6:17 124.3:26 140.6:91} s {D 70; P 141.1:0 188.3:86 190.4:93}

t {D 60; P 184.4:0 167.5:92} =IH {D 75; P 159.3:0 146.6:13 137.4:27} N {D 90; P 101.2:0 95.7:11 93.8:22 94.6:89} [[inpt TEXT]]"


I was hoping someone could confirm this is a voice problem vs me doing something dumb. If I'm not being dumb, anyone know of a workaround (other than using a different voice)?

iMac, OS X Mountain Lion (10.8.3)

Posted on May 22, 2013 4:53 PM

Reply
2 replies

May 22, 2013 7:06 PM in response to seanth

I wrote a short python script that read /System/Library/Speech/Voices/ and made lists of all voices that are "Compact" and those that are not (and don't have a corresponding "Compact" version). I can confirm that all the "Compact" voices are unable to use the [[inpt TUNE]] info.


Not that this helps me, but it helps narrow the field.

May 22, 2013 10:52 PM in response to seanth

I wrote a seies of tests based on the embedded speech commands found at https://developer.apple.com/library/mac/#documentation/userexperience/Conceptual /SpeechSynthesisProgrammingGuide/FineTuning/FineTuning.html#//apple_ref/doc/uid/ TP40004365-CH5-SW10


Four voices were tested to try and cover the history of speech with OS X: Victoria (10.0), Vicki (10.3), Alex (10.5), and Moira (10.7). From the results its pretty clear the newer 10.7 voices (well, at least Moira and Tom) are either not as feature rich by design or broken:


CommandVictoria (10.0)Viki (10.3)Alex (10.5)Moira (10.7)
chary y y y
ctxt WSKPy y y n
ctxt TSKPy y y n
emphy y y n
inpt PHONy y y n
inpt TUNEy y y n
nmbr LTRLy y y y
nmbr NORMy y y y
pbasy y y/n* n
pmody y y n
ratey y y y
slncy y y y
volmy y y y

*: Alex correctly increased the baseline pitch, but only for the word immediately after the command


Python used to test the voices:


import os

theTests=['[[char LTRL]] cat [[char NORM]] cat',

'[[ctxt WSKP]] GPS provides [[ctxt WORD]] coordinates. [[ctxt NORM]]'

'[[ctxt WSKP]] The post office [[ctxt WORD]] coordinates [[ctxt WSKP]] its deliveries. [[ctxt NORM]]',

'[[ctxt TSKP]] Your [[ctxt TEXT]] first step [[ctxt TSKP]] should be to relax. [[ctxt NORM]]'

'[[ctxt TSKP]] To relax should be your [[ctxt TEXT]] first step. [[ctxt NORM]]',

'Do [[emph +]] not [[emph -]] over tighten the screw.',

'My name is [[inpt PHON]] AY1yIY2SAX [[inpt TEXT]]',

'[[inpt TUNE]] k {D 90; P 150.0:0 150.0:67 150.0:96}',

'Please call me at [[nmbr LTRL]] 5551990 [[nmbr NORM]].',

'Here is a number: [[nmbr NORM]] 127 [[nmbr NORM]].',

'[[pbas 127.0]]Baseline pitch',

'no Pitch modulation. [[slnc 500]] with [[pmod 0]]Pitch modulation',

'[[rate 2000]] Speech rate',

'And now [[slnc 2000]] silence',

'Loud and [[volm 0.5]] soft']

theCmd="say -v %s '%s'"

for aTest in theTests:

os.system(theCmd % ("Victoria", aTest))

os.system(theCmd % ("Vicki", aTest))

os.system(theCmd % ("Alex", aTest))

os.system(theCmd % ("Moira", aTest))

This thread has been closed by the system or the community team. You may vote for any posts you find helpful, or search the Community for additional answers.

Some voices in 10.8 don't support [[inpt TUNE]]?

Welcome to Apple Support Community
A forum where Apple customers help each other with their products. Get started with your Apple Account.