Your dive photos
are memories.

Connect DeepSix to your Apple TV. Your trips appear as a browse shelf — site photos, depth badges, trip names. Pick a dive. Your photos fill the screen, one by one, with the dive profile running alongside them.

Every photo is placed at the exact moment it was taken. The scrubber dot moves. The depth readout matches the frame. The species badge shows what you were looking at. It's not a slideshow. It's a dive replay.

The tvOS home screen shows your dive history as a globe with pinned sites and a trip list. Select a trip — the view flies to your sites and shows every dive as a card. Site photo or satellite map, depth badge, date. Scroll through the Solomons trip and Toa Maru, Zeuna North Reef, Sand Island, and Heian Maru are each a card.

Your log is a travel record as much as a dive record. Zanzibar and the Jewels of the Indian Ocean. Solomon Islands. Micronesia. Komodo. On the TV, that context lands differently than on a phone screen.

Underwater photos lose red first, then orange, then yellow. By 20 meters your camera is recording a world that only exists in blue and green. DeepSix corrects for depth — not a flat filter, but a depth-aware algorithm that knows how much red was lost at 18 meters versus 5 meters.

On-device, non-destructive, free. Your photos in Photos.app are never overwritten — the correction is stored as an adjustment layer, reversible with a tap.

The Paralenz was doing its best. It corrected what it could in real time, in a sealed housing, with the compute available in 2015. The algorithm running in DeepSix in 2026 is better. And it doesn't care when the photo was taken.

Depth-adaptive color correction works on every photo in your library. The depth data is already in your dive log. The correction is better now. Your archive is worth revisiting.

The boat ride out. The surface interval. The debrief on the back deck. The beer at 4pm when everyone's talking about the manta they saw and nobody agrees on the wingspan.

DeepSix scans your Photos.app for the 45 minutes before and after each dive and tags those photos as surface memories. They're part of the record. They're available as hero images. They show up in your trip on Apple TV alongside the underwater shots. They just don't get placed on the depth profile — because they weren't there, and the profile should be honest.

Surface memories are day-wide. The boat ride out at 8am shows in every dive that day — not just the first one. Photos from your partner touring inland while you're underwater don't appear at all: DeepSix filters out any photo taken more than 5km from the dive site.

Critter ID doesn't run on surface photos. The algorithm knows you weren't photographing fish from the surface interval. Your API credits go to the photos that need them.

Every dive has a Hero — the single photo that represents it on the dive list and everywhere the dive appears — and Highlights, the curated set of up to five that lead the dive. They're separate decisions with separate badges: the hero wears a star, the highlights wear sparkles. DeepSix picks both automatically, but you're never stuck with its taste — long-press any photo to Set as Dive Hero, and tap the current hero to clear it back to the automatic pick.

The scene classifier keeps the galleries honest, too. Not everything shot on a dive day is a dive photo — the slate, the gauge check, the screenshot of the briefing. Those land in their own Non-Dive Photos bin instead of your galleries, and a three-way Dive / Surface / Non-Dive control in photo detail lets you overrule the classifier — your manual call sticks.

And a dive with no photos at all still gets a face: its list row falls back to the site's reference image, so a photo-less dive never scrolls by as a gray square.

Claude analyzes every photo from a dive and returns the species name, scientific name, family, and a confidence score. High-confidence IDs build your lifetime species database. Every Aetobatus narinari across every dive is connected.

GPS coordinates go with every identification request — a ray at Puluwat Atoll narrows to different candidates than the same shape in the Red Sea.

Your dive computer, camera, and Apple Watch all disagreed on the time. DeepSix measures the clock offset from each source and applies it to every photo match. Tell it which photo was first — it calculates the last, accounting for drift. Every photo between them is matched to depth and time automatically.

Photos already matched to another dive show orange conflict badges when you scan a new one. The system knows what it's already claimed.

The photographer on the dive has the good frames. DeepSix is building direct device-to-device transfer — no internet, no cloud, no cables. Two iPhones on the same liveaboard. Their photos go through the same clock-offset matching. They land on your dive at the right depth, the right GPS, the right minute.

Works with AirDrop, SD cards, and any app that can share to DeepSix. The pipeline is the same regardless of how the photo arrives.