Discussions
Inquiry on Real-Time Text Streaming Capabilities
I'm utilizing a Large Language Model (LLM) and I'm interested in streaming text inputs to enhance real-time processing with D-ID. Can D-ID support streaming inputs where it begins processing as soon as a word is submitted?
poor video resolution on streams
i've seen this posted multiple times, but i haven't noticed a satisfactory answer. i would like to use the talks/streams/{id} end point to stream videos using text and an eleven labs voice. i've copied the code from https://github.com/de-id/live-streaming-demo and modified it for my use case and it works, but the video quality is poor (low resolution, blurry). the audio is fine. I know that my network is not the problem. at the time i was running the test i was getting 81Mbps download speeds. I turned on chrome://webrtc-internals/ to capture statistics for multiple streaming runs and i was not seeing any packets dropped and the incoming video bitrate was between 250k and 320k bps, so it does not seem that the stream coming from D-ID is taking advantage of all the bandwidth available. What are my options for improving the quality of the video?
session_timeout
i tried to increase the session_timeout to 300 in the talks/streams endpoint, but the call failed. i can see that the documentation says you need to have proper permissions. how do i get those permissions?
How can I increase the size and change the position of the embedded AI agent on my site?
I've created an AI agent using the pro plan and embedded it into my website. However, the AI is closed, too small, and located in the far right corner of the website. How can I keep the Ai agent open so that it's a larger size with the preloaded questions visible to the user, and how can I change the position so that it's located above the fold, right above the Book a Demo button? Here's a link to the site with the AI agent: https://www.limitlessgrowthsolutions.com/
Unauthorized
import requests
Seamless Transition for Clips
I am programming a smart interphone. I have written a code that gives me responses formatted as JSON. How can I send them to my API?
hello
Video breaking while streaming avatar
Occasionally, the streaming video and audio may experience interruptions during playback, which could be due to issues with the WebRTC connection i guess? Are there methods available to optimize this and prevent the video from breaking?
English words in German Texts are mispronounced
We use D-ID Studio with German language output. Unfortunately, all English words within the German text are pronounced in German as well, which sounds very awkward. Is there a possibility to mark those words within the German text which should be pronounced in English?
In html one can, for example, use a language tag to make screenreaders pronounce English words properly.
How do I change the logo to Ai?
If I set the logo to anything, I get a "400 bad request", I am using axios and am trying to create options like this: