Speak[] function using Wolfram Cloud

Question

I wrote a program for practicing note identification on a guitar:

note1 = {"A.", "B", "C", "D", "E", "F", "G"};
note2 = {" Sharp", " Natural", " Flat"};
string = {"first  ", "second  ", "third  ", "fourth ", "fifth ", 
   "sixth "};
list = StringJoin[#] & /@ Tuples[{note1, note2}];
list2 = Flatten[
   Table[RandomSample[StringJoin[#] & /@ Tuples[{string, list}]], {i, 
     1, 1}]];
For[i = 1, i <= Length[list2], i++,
 Speak[list2[[i]]];
 Pause[2.5];
 ]

I do not have a Mathematica license, so I want to use Wolfram Cloud. The problem is that the server times out before the for-loop gets through the entire list. Is there anyway to send the speak command to the server, then have it send back the entire list in audio format?

If you don't have a Mathematica license, and Wolfram Cloud is not a solution for you, you can still use Wolfram Engine and Jupyter Notebook as a front-end, both free to install and use (some restrictions apply). — rhermans, May 23 '22 at 09:30

rowsi · Accepted Answer · 2022-05-23T14:59:57.203

This works for me on Wolfram Cloud:

note1 = {"A.", "B", "C", "D", "E", "F", "G"};
note2 = {" Sharp", " Natural", " Flat"};
string = {"first  ", "second  ", "third  ", "fourth ", "fifth ", 
   "sixth "};
list = StringJoin[#] & /@ Tuples[{note1, note2}];
list2 = Flatten[
   Table[RandomSample[StringJoin[#] & /@ Tuples[{string, list}]], {i, 
     1, 1}]];
pauseTime = 2;(in seconds)
audio = AudioJoin[
   AudioPad[#, pauseTime] & /@ SpeechSynthesize /@ list2]

To speed the spoken part up, add AudioTimeStretch with an appropriate factor. I had to rewrite the code a bit, as Wolfram Cloud doesn't seem to like too many nested Map-functions. Adjust audioSpeed to taste:

Clear[note1,note2,string,list,list2,pauseTime,audio,audioSpeed];
note1 = {"A.", "B", "C", "D", "E", "F", "G"};
note2 = {" Sharp", " Natural", " Flat"};
string = {"first  ", "second  ", "third  ", "fourth ", "fifth ", 
   "sixth "};
   list = StringJoin[#] & /@ Tuples[{note1, note2}];
list2 = Flatten[
   Table[RandomSample[StringJoin[#] & /@ Tuples[{string, list}]], {i, 
     1, 1}]];
pauseTime = 2;(*in seconds*)
audioSpeed=0.9;(*Values < 1 for speeding up*)
   speedaudio=AudioJoin[Riffle[AudioTimeStretch[SpeechSynthesize[#],audioSpeed,Method->"Speech"]&/@list2,AudioGenerator["Silence",pauseTime]]]

This is great thanks! The speaking seems a little slow though, is there any way to speed those parts up? — ngc1300, May 23 '22 at 13:40
Do you mean the pause between each called note and string is too long or the spoken part itself is too sluggish? — rowsi, May 23 '22 at 14:09

Speak[] function using Wolfram Cloud

1 Answers1