Let their voice be part of the goodbye.
A voice is often the first thing we lose, and the thing we miss most.
We turn the photos, phone clips, and recordings a family already holds into one service-ready memorial film. Voice projects begin with a 15-second preview. If it doesn't sound like them, we stop there.

Hear the voice before full production.
After complete, usable materials arrive.
Known wishes and close-family agreement.
Source files and voice models are deleted unless archived by request.
Pick by what you already have.
We look at what you have before taking the project on. The clock starts when the last usable file arrives, not when you first write.
What a farewell sounds like.
A 90-second demonstration in our production style: restored era photographs, gently animated moments, and a farewell letter read in a recreated voice.
A demonstration film. “Margaret” is fictional — imagery and voice are AI-generated to show our process and style. Client films are made only from a family’s own materials, with consent confirmed first.
Stay close to the family. We handle production.
Either we talk to the family directly, or they only ever hear from you. No volume commitment; every voice project still starts with the 15-second preview.
Photos remain. Voices disappear.
Photos get printed, framed, and kept. Voices are left inside a voicemail no one can bring themselves to delete. What people miss is often how someone sounded when they said a name.
A farewell, blessing, or thank-you proceeds only after the family approves the preview.
A service moves fast. A familiar voice can hold one quiet moment inside it.
We make an ending. Then we stop.
“One farewell. Not digital immortality.”
You hear it before you decide.
The limits are part of the service.
If they said they would not want this, we decline. If someone close to them objects, we stop.
No chatbot, ongoing avatar, or claim of digital immortality.
Voice processing runs on studio machines; source material and models are deleted within 30 days unless archiving is requested.
Every film is labeled as containing AI-generated content and licensed for the service and private family keeping.
Tell us the date and what material exists.
We reply within one business day with a feasibility assessment and the safest next step. English and Traditional Chinese are welcome.
Share the deadline and material overview first; files are requested only after consent and feasibility are clear.
Voice Legacy by Tzukao. Please don't send identity documents or sensitive files in a first email.
Can you work from the photos and clips we already have?
Yes. A Memorial Film normally begins with 20 or more photos and any useful phone clips. We assess the material before committing to a format or date.
Are a few voicemails enough?
Sometimes. We test the recordings and let the 15-second preview decide; if the recordings are too rough, we say so rather than promise.
What if the family disagrees?
Production stops. Known wishes and close-family agreement are required before voice work proceeds.