Chrisfirst restores JFK’s Rice speech to 4K at 60fps with Topaz
Chrisfirst presented a 4K, 60fps restoration of JFK’s Rice University speech using Topaz Starlight 2.6 and Adobe Podcast. Posts compare image details and show an audio spectrogram.

TL;DR
- JFK’s Rice University speech gets a 4K, 60fps restoration using Topaz Starlight 2.6 and Adobe Podcast, according to chrisfirst.
- The visual comparisons cover faces in one close-up and newly readable presidential-seal lettering in another.
- Audio cleanup gets a separate spectrogram comparison, with the original and processed tracks overlaid in chrisfirst’s demonstration.
Topaz’s Starlight 2.6 model notes include a control for backing off over-sharpened detail. Its desktop documentation also flags an archival gotcha: blurry lettering can be interpreted as texture.
A 4K 60fps Rice speech
Chrisfirst describes his starting footage of the 1962 address as grainy 30fps video, with hiss under the words.
The thread does not explain the 30-to-60fps conversion or specify processing settings. The named tools establish the picture-and-sound stack, but the post stops short of a reproducible recipe.
Faces and the presidential seal
The strongest visual test is the podium seal: chrisfirst says “President of the United States” becomes readable down to every letter, alongside added definition in the eagle’s feathers.
The face comparison extends beyond Kennedy to the people behind him.
Adobe Podcast and the spectrogram
Chrisfirst supplies a three-color key for the audio overlay:
- Yellow: the 1962 track.
- Blue: the processed track.
- White: where the two overlap.
He calls the remaining yellow regions noise. Adobe’s Enhance Speech FAQ describes pitch and volume adjustments alongside noise filtering, so spectral differences can reflect processing beyond removed hiss.
Local rendering and the full speech
Topaz’s desktop documentation lists these options for Starlight Precise 2.5/2.6:
- Cloud exports: supported on Mac and Windows.
- Local Nvidia or AMD rendering: at least 12GB VRAM and 32GB system RAM.
- Local Apple Silicon rendering: at least 16GB memory.
Topaz warns that local diffusion rendering takes a long time. Chrisfirst does not report a render duration for this project, but says a YouTube upload of the full 18-minute restoration is forthcoming.