K3
September 17, 2024, 3:16pm
1
Original Discord Post by hyrohyken | 2024-09-17 15:16:23
Environment :
Convai : Free Plan
ElevenLabs integration in Convai via API Key
Using the voice model : Claire
The voices has been set and its working pretty well
We only have one problem, it seems that the voice goes lower and lower throughout the speech
Our goal is to make an AI Character that delivers quite long speech (approx 1000 letters)
I have tested the same speech on the ElevenLabs interface and the behaviour isn’t the same
Any clue on how to fix this ?
K3
September 17, 2024, 4:16pm
2
Reply by joel_jamaze | 2024-09-17 16:16:25
I experience testing a unreal engine application with eleven labs pro and convai pro the same this affternoon ……
K3
September 17, 2024, 8:11pm
3
Reply by k3kalinix | 2024-09-17 20:11:12
Hello <@354588744143339520> & <@255863933125132288>,
Could you please share Character ID & Session IDs?
K3
September 18, 2024, 6:53am
4
Replying to k3kalinix’s Message
Reply by k3kalinix | 2024-09-17 20:11:12
Could you please share Character ID & Session IDs?
Reply by hyrohyken | 2024-09-18 06:53:40
Of course !
Character ID
c28b1dc6-74c7-11ef-adcc-42010a7be011
Session ID
6aab7786fa429bd940a65b04cf5f5ef7
K3
September 18, 2024, 12:18pm
5
Reply by k3kalinix | 2024-09-18 12:18:51
Thanks for the details.
K3
September 27, 2024, 12:40pm
6
Replying to k3kalinix’s Message
Reply by k3kalinix | 2024-09-18 12:18:51
Reply by hyrohyken | 2024-09-27 12:40:48
Hello ! Any news about this issue ?
K3
September 27, 2024, 12:41pm
7
Reply by k3kalinix | 2024-09-27 12:41:23
<@547328106927620106>
K3
September 28, 2024, 5:52am
8
Reply by d_acharyya | 2024-09-28 05:52:38
Hi <@354588744143339520> we are currently looking into it
K3
September 28, 2024, 5:56am
9
Reply by d_acharyya | 2024-09-28 05:56:01
Can you please confirm one info :
K3
September 28, 2024, 6:49am
10
Reply by d_acharyya | 2024-09-28 06:49:53
Also which model did you used while trying out on Elevenlabs platform
K3
September 28, 2024, 3:00pm
11
Replying to d_acharyya’s Message
Reply by d_acharyya | 2024-09-28 05:56:01
Reply by hyrohyken | 2024-09-28 15:00:42
This issue is happening on every response from a voice from ElevenLabs
We used the latest avalaible model and « Claire » voice
K3
September 29, 2024, 5:22am
12
Reply by d_acharyya | 2024-09-29 05:22:18
Thanks for confirming
K3
October 2, 2024, 7:09am
13
Reply by hyrohyken | 2024-10-02 07:09:39
Hello ! Any updates ?
K3
October 3, 2024, 9:12am
14
Reply by d_acharyya | 2024-10-03 09:12:14
We were able to recreate the issue on unreal platform, but its only happening for your voice. We tried with a couple of other elevenlabs custom voices and there were no volume issues.
K3
October 3, 2024, 9:13am
15
Reply by d_acharyya | 2024-10-03 09:13:01
Can you please give it a try in the convai website and let us know if you are getting the issue there too ?
K3
October 3, 2024, 12:07pm
16
Reply by hyrohyken | 2024-10-03 12:07:35
We are getting the issue from all our custom voices in the Convai Platform and in our UE project
What Elevenlabs model are you currently using for your REST call ?
K3
October 3, 2024, 1:59pm
17
Reply by d_acharyya | 2024-10-03 13:59:05
Its eleven_multilingual_v1
K3
October 3, 2024, 1:59pm
18
Reply by d_acharyya | 2024-10-03 13:59:24
Its not a REST call, we use the streaming version that uses the websocket
K3
October 3, 2024, 3:00pm
19
Reply by d_acharyya | 2024-10-03 15:00:16
When you said : latest avalaible model , were you referring to eleven_multilingual_v2 ?
K3
October 3, 2024, 4:22pm
20
Replying to d_acharyya’s Message
Reply by d_acharyya | 2024-10-03 15:00:16
Reply by hyrohyken | 2024-10-03 16:22:28
Yes, and the 2.5 aswell