Indigo Domotics

Please add all future feature requests to this thread.

Hi Dave,

It should be nice if it would be possible to adjust the words per minute rate. I think that it’s to fast at the moment and would like to have the possibility to slow it down when you have longer and more complex announcements.


Skickat från min iPad med Tapatalk

tazswe wrote:

Hi Dave,

It should be nice if it would be possible to adjust the words per minute rate. I think that it’s to fast at the moment and would like to have the possibility to slow it down when you have longer and more complex announcements.


Skickat från min iPad med Tapatalk

Thanks for the suggestion. I'll add this to the list of things to take a look at the next time that I'm in the plugin code. I need to think about how best to implement such a feature. The reason: it's a bit more complicated than it may appear which will become evident in a moment. I need to figure out the best way to expose the following information in the plugin device.

As it turns out, you already have this power! Apple's embedded speech system has ways to adjust the spoken strings that can make some significant changes in the way that the speech is rendered. For your circumstance, you could try something like this (the value is in words per minute):

Code: Select all

say "[[rate 150]]This is a sentence with seven words."


You can play with this within the announcements device settings pane or enter the text directly in Terminal with the `say` command and adjust the rate value to taste. Take small bites at the apple as these adjustments can have weird outcomes if you're not cautious. I'd guess the default rate is around 175 WPM. There are other useful codes as well. You can find references to these modifiers and an example on the Tutorials page of the plugin wiki.

Hi Dave,

Thanks a lot, it’s working perfect.

Tried a lot of settings and you can’t go below 145 unless you want her to sound sleepy.

Next project will be to replace a dot with a comma in the weather announcements.
We don’t have the same decimal separator and it can make it sound weird in an announcement.


Skickat från min iPad med Tapatalk

Glad to hear that's working well for you. Yes, the comma separator issue has come up before. I was never able to find a way to universally handle this through code. I'm frankly surprised that the system settings forLanguages and Regions doesn't handle this. With more current OS versions (I'm on Catalina) there is an ability within the Languages and Regions pane for making application-specific adjustments; however, I've never toyed with those so I can't be much help there I'm afraid.

Hi Dave,

Played around with different settings and found something wired when I tried to put in a moment of silence.

There is a remarkable difference between [[slnc 900]] and [[slnc 1000]] and almost no difference between 900 and 400, it’s like it’s not working with values less than 1000 milliseconds?


Skickat från min iPad med Tapatalk

I think it may be a matter of perception. I wrote a simple script to investigate and--while it's difficult to discern the difference in the steps audibly--there's clearly a difference between the beginning and end of the loop. This script demonstrates it:

Code: Select all

from os import system
import datetime as dt

for r in range(1000, 0, -100):
   now = dt.datetime.now()
   system('say one [[slnc {}]] two'.format(r))
   print(dt.datetime.now() - now)


0:00:02.280117
0:00:02.172002
0:00:02.073821
0:00:01.967781
0:00:01.874026
0:00:01.775612
0:00:01.670471
0:00:01.573126
0:00:01.477721
0:00:01.371177

While not an exhaustive (or extremely accurate) test, it's clear that the statement is getting progressively shorter. There shouldn't be any effect of running the statement through the plugin because the speech synthesis is interpreted by the operating system. I'd suggest using something that sounds right to your ear regardless of the actual value you use.

I would like to request a feature that allows the announcement string spoken by the built-in speech synthesizer to be exposed for use by other plugins.

Currently, when the "Speak Announcement" button is clicked on the "Configure Announcement Device" screen, only the built-in speech synthesizer is used. I would like the ability to route the announcement string through another tool so that it can be spoken by a different speech synthesizer plugin. This would help in the integration of alternative speech synthesizers, as it would allow users, as they construct Announcements, to hear their complex announcements strings thru their preferred tool.

anyone wrote:

I would like to request a feature that allows the announcement string spoken by the built-in speech synthesizer to be exposed for use by other plugins.

Currently, when the "Speak Announcement" button is clicked on the "Configure Announcement Device" screen, only the built-in speech synthesizer is used. I would like the ability to route the announcement string through another tool so that it can be spoken by a different speech synthesizer plugin. This would help in the integration of alternative speech synthesizers, as it would allow users, as they construct Announcements, to hear their complex announcements strings thru their preferred tool.

That's interesting. I agree this would be a good addition to the plugin, but I'm not sure how I'd do it exactly. I'll have to think about that one for a bit. Fully open to suggestions!

Even tho I don't think you made ur app to be a tts but just generate the announcements. Can you make a version that plays a sound before or after the spoken announcement using this app https://www.macupdate.com/app/mac/63375/play-sounds

Would appreciate it

killerdan56 wrote:

Can you make a version that plays a sound before or after the spoken announcement using this app https://www.macupdate.com/app/mac/63375/play-sounds

Would appreciate it

I'm not sure I could. I looked over the app's minimal documentation and it said even support for AppleScript is rudimentary which means to me that it's not easily scriptable.

You're right about the plugin -- it's targeted towards constructing strings, which are saved into device states. Then users can script those states in any way they choose.