Steam Frame Lab

Let's get comfort and performance settings dialed in for every game on Steam Frame.

Steam Frame runs on one fixed chipset and one fixed pair of pancake displays, whether a game is running standalone on the headset or streamed over the bundled 6GHz dongle from a PC. That means the comfort and graphics settings that make a game feel right are worth testing once, properly, and sharing — not rediscovering headset by headset.

That's the gap we're building this site to fill, the same way Steam Machine Lab does it for Steam Machine: crowdsourced settings from people actually wearing the headset, not guesses from a spec sheet.

Steam Frame isn't out yet, but we've already settled on exactly how we'll test each game once it is: a fixed benchmark pass, then comfort tuning, then publishing the numbers with whoever ran them credited. Below is where that plan stands today, and how to help once hardware ships.

What we'll test

Every category below gets the same treatment Steam Machine Lab gives its FPS tables: run by real hardware, credited to whoever submitted it, and published without a login wall.

Performance

Per-game frame rate, standalone vs. streamed

Frame rate at 72/80/90/120Hz on-device, and again over the 6GHz streaming dongle, so you know which games actually want a PC behind them.

Comfort

Comfort & FOV tuning per game

Snap turn vs. smooth turn, vignette strength, and how far you can push foveated rendering before a given game starts to look smeared.

Streaming

Dongle vs. Wi-Fi vs. USB streaming

Latency and visible compression across every PC-streaming path Valve ships, not just the bundled point-to-point dongle.

Battery

Battery life & thermals by genre

Front-box only vs. counterweight battery strap, runtime per game type, and where thermal throttling starts to cost you frames.

Controllers

TMR thumbstick feel & deadzones

Per-game deadzone and sensitivity recommendations for the tracking-ring controllers, the same way we log control schemes today.

Storage

Which storage tier fits your library

How many typical VR titles actually fit at each capacity Valve offers, so you're not guessing at checkout.

How testing will work

Same three-step process Steam Machine Lab uses for game settings, adapted for a headset instead of a console-like box.

01

Run a fixed benchmark pass per game

Standalone first, then streamed, at each supported refresh rate, on stock settings, so every game starts from the same baseline before anything is tuned.

02

Tune for comfort, then for frame rate

Comfort settings (turning, vignette, FOV) are locked in first since they're personal; performance settings are then pushed as far as they'll go without breaking that baseline.

03

Publish it, credit the submitter, take corrections

Every settings page names who tested it and on what storage/battery configuration, and stays open to community corrections — no locked-down "official" numbers.

FAQ

Is this an official Valve site?

No. Steam Frame Lab is an independent fan project, run the same way as its sister site, Steam Machine Lab. Valve has no involvement.

Has Steam Frame actually come out yet?

Not as of this writing. Valve has pointed to a launch window for the headset but hasn't shipped it, so every benchmark and settings page on this site will be built from real units once they're in hands — not from spec-sheet guesses.

Will standalone and PC-streamed settings really differ that much?

Almost certainly, yes. Standalone play is bound by the onboard chipset; streaming trades that ceiling for network and compression variables instead. We'll test and publish both paths separately rather than picking one as "the" setting.

How can I contribute once I have a headset?

Use the "Submit your data" link to send your benchmark run and settings. Include your storage tier and battery configuration — both affect results enough to matter.