How to identify TTS files

Configuration Voice Assistant
1

Hi all!

I have a project which sends notifications to my nest hub.
Picture / text + voice in a same time.
I’m using NodeRed and castv2 node, creating the TTS file with VLC telnet.
The problem is: the generated mp3 file has to be in the www folder otherwise the castv2 node can’t play it.
So I’m trying to identify the corresponding mp3 file and copy to www folder.
It worked 'till the last few updates.
Now I can’t get the file names.
The HA system has some kind of list of the generated ttts- text pairs because if I generated a tts and a few days after giving the same text, the system won’t generate again the mp3, but uses the old one.
So not always the “youngest” file is what I want.

My question is: is there any method or list / directory where the system stores the text - tts pairs?
I tried the REST API as well, but it gives back different filenames, than what names are in the /tts directory.

I also wanted to know how long - in seconds -is a tts or mp3 file.
Is there any simple way to get that?
Now I’m using some terminal command, which needs some packages to install - after every system reboot - .

Excuse me for my poor english!

2

Are you sure it writes them out as files (as opposed to just caching them in memory)? I wondered the same thing and I couldn’t find any files that looked like they were TTS cache files.

3

You can use the Chime TTS integration’s chime_tts.say_url action to generate your TTS audio. The action’s response object includes both the URL to the generated file as welll as the duration of the audio file.