#Is there a transcript template for the best outcome possible when one wants to make a PVC? I want t

1 messages · Page 1 of 1 (latest)

spare houndBOT
#
AI-generated Answer for rolandmedici

When creating a Professional Voice Clone, it's important to focus on providing high-quality audio samples. While there isn't a specific transcript template mentioned in the available information, here are some key points to consider for the best possible outcome:

  1. Upload at least 2 hours of high-quality audio featuring a single speaker.
  2. Ensure the audio is free from background noise and excessive reverb.
  3. Avoid distractions during the recording, as the AI will mimic the audio precisely.

Regarding creating a Professional Voice Clone in Bahasa Indonesia, it's crucial to note that the current policy supports only one language per clone. The system is designed to work with audio samples exclusively in the selected language.

If you're planning to create a clone in Bahasa Indonesia, make sure all your audio samples are recorded only in that language. Mixing languages or including words from other languages might lead to errors or rejection of the samples.

Remember, you're limited to one Professional Voice Clone per account, unless you have a Business or Enterprise subscription. If you need to create clones in multiple languages, you might consider setting up separate accounts for each language.

sturdy wren
#

If there isn't any template for the transcript, is there tips for one? Or its fine to just say anything as long as its high quality?

spare houndBOT
#
AI-generated Answer for rolandmedici

There isn't a specific transcript template mentioned for creating a Professional Voice Clone (PVC). However, there are clear guidelines to ensure the best outcome:

  1. Audio Quality: Recordings must be high-quality, free of background noise, excessive reverb, music, or distractions. The audio should sound professional and feature only one speaker.

  2. Consistency: Use recordings with a consistent tone, style, and volume. Avoid mixing old and new recordings, as this can lead to inconsistencies.

  3. Duration: Ideally, provide at least 2 hours of high-quality audio. While successful clones can be trained with as little as 10 minutes, more data improves the results.

  4. Content: You can say anything, as long as the audio quality meets the standards. Multiple topics are acceptable, but ensure clarity and avoid excessive pauses.

  5. Recording Environment: Record in an acoustically treated space to minimize echoes and background noise.

For Bahasa Indonesia, the same principles apply. Focus on clear pronunciation and consistent delivery to ensure the AI captures the nuances of the language effectively.

sturdy wren
#

Do i need to record my voice on a special application like audacity? Or using the voice recorder from the phone (that is connected to a mic with good quality) is fine? Is there a specific format for the recording?