#How can I add more samples to an already created professional voice clone

1 messages · Page 1 of 1 (latest)

rose stagBOT
#
AI-generated Answer for daria076199

Short answer

  • If your Professional Voice Clone is still in Draft (not yet verified), open My Voices > More actions (three dots) > Edit voice to return to the creation flow and upload additional samples. (help.elevenlabs.io)
  • If it’s already created/verified, add samples via the PVC API, then retrain:

API steps

  1. Add samples to the PVC
  • POST /v1/voices/pvc/:voice_id/samples with your audio files (multipart). Optional: set remove_background_noise. (elevenlabs.io)
  1. (Optional) Clean up or select speakers per sample
  • POST /v1/voices/pvc/:voice_id/samples/:sample_id to apply noise removal, choose speakers from separated audio, and trim. (elevenlabs.io)
  1. Start retraining
  • POST /v1/voices/pvc/:voice_id/train to fine‑tune the clone with the newly added samples. (elevenlabs.io)

Notes

  • ElevenLabs’ April 21, 2025 update introduced these PVC sample-management endpoints (add/update/delete, speaker separation, verification, train). (elevenlabs.io)
  • Typical PVC training takes roughly a few hours once queued. You’ll be notified when ready. (help.elevenlabs.io)

If you need the exact request shapes or SDK calls, the API reference pages above include runnable examples. (elevenlabs.io)

Sources:
What does the status of my Professional Voice Clone mean? – ElevenLabs
Add samples to PVC voice | ElevenLabs Documentation
Update PVC voice sample | ElevenLabs Documentation
Train PVC voice | ElevenLabs Documentation
April 21, 2025 | ElevenLabs Documentation
When will my professional voice clone (PVC) be ready? – ElevenLabs
Introduction | ElevenLabs Documentation

crystal flume
#

How do I add more samples to the PVCAPI Andrew retrain?

rose stagBOT
#
AI-generated Answer for daria076199

Short answer:

  • Use Add samples to PVC voice to upload the new audio files to your existing PVC (your “Andrew”) by voice_id: POST /v1/voices/pvc/:voice_id/samples (multipart form: files=[...]). (elevenlabs.io)
  • If any clips have multiple speakers, run speaker separation and choose the correct speaker(s), then update the sample(s) with selected_speaker_ids. (elevenlabs.io)
  • Retrain the PVC to incorporate the new samples: POST /v1/voices/pvc/:voice_id/train (optionally set model_id). (elevenlabs.io)

Minimal Python SDK example:

  • Get Andrew’s voice_id (via API or copy from dashboard). (elevenlabs.io)
  • Add samples, then retrain.

Example:
from elevenlabs import ElevenLabs
from io import BytesIO

client = ElevenLabs(api_key="YOUR_API_KEY")

voice_id = "ANDREW_VOICE_ID"

1) Add new samples

files = [BytesIO(open("andrew_sample1.wav","rb").read()),
BytesIO(open("andrew_sample2.mp3","rb").read())]
client.voices.pvc.samples.create(voice_id=voice_id, files=files)

Optional: separate speakers, pick the right one(s), and update the sample(s)

client.voices.pvc.samples.speakers.separate(voice_id, sample_id) # start

client.voices.pvc.samples.speakers.get(voice_id, sample_id) # status

client.voices.pvc.samples.update(voice_id, sample_id, selected_speaker_ids=[...])

2) Retrain to apply the new data

client.voices.pvc.train(voice_id=voice_id, model_id="eleven_multilingual_v2")

Refs: Add samples endpoint, speaker separation + update endpoints, and train endpoint. (elevenlabs.io)

Notes:

  • A PVC must be verified before training; use the CAPTCHA verification endpoints if the voice isn’t verified yet. (elevenlabs.io)
  • The cookbook shows the full end‑to‑end PVC flow (create, add samples, separation, verify, train) using the SDK. (elevenlabs.io)

Sources:
Add samples to PVC voice | ElevenLabs Documentation
Start speaker separation | ElevenLabs Documentation
Train PVC voice | ElevenLabs Documentation
List voices | ElevenLabs Documentation
Add samples to PVC voice | ElevenLabs Documentation
Get PVC verification captcha | ElevenLabs Documentation
Professional Voice Cloning API Cookbook | ElevenLabs Documentation

crystal flume
#

Where do I find add samples to PVC?

rose stagBOT
#
AI-generated Answer for daria076199

In the API docs: Voices → PVC → Samples → “Add samples to PVC voice” (POST /v1/voices/pvc/:voice_id/samples). It expects a multipart form with one or more files under files. (elevenlabs.io)

Helpful references:

Sources:
Add samples to PVC voice | ElevenLabs Documentation
Add samples to PVC voice | ElevenLabs Documentation
Professional Voice Cloning API Cookbook | ElevenLabs Documentation
Train PVC voice | ElevenLabs Documentation
How do I find the voice ID of my voices via the website and API? – ElevenLabs

crystal flume
#

Is there a way to add samples without using the API? And if I need to recreate my voice clone and have sent a two year notice. How long do I have to wait before adding a new one after deleting the original?