Under the hood

For the people who read the specification first.

The settings window in full, the maths behind the file size estimate, the frameworks the capture is built on, and every number we would want to know before installing something.

One window, no tabs

This is the whole settings window.

Change anything on the left and the capture on the right follows, countdown included. The readout underneath is the same encoder estimate the app shows you before you hit record.

Superscreen Settings

Recording

Cursor

Show cursor in recording
Highlight clicks

Camera bubble

Camera
20%

Audio

Microphone
System audio

Files

Save to~/Movies/Superscreen
Next file nameTemplates arrive in 1.0
Superscreen Area 2026-08-23 at 16.24.mov
Superscreen
09:41
Recorder.swift
1import ScreenCaptureKit
2
3final class Recorder {
4func start(_ filter: SCContentFilter) async {
5let config = SCStreamConfiguration()
6config.minimumFrameInterval = .init(1, 60)
7config.capturesAudio = true
8config.showsCursor = settings.showCursor
9
10try await stream.startCapture()
11}
12}
superscreen.app/metrics
1638 × 835
00:00
Estimated outputHardware encode
Resolution
1280 × 800
Bitrate
7.2 Mbps
Per minute
54.1 MB
Per hour
3.17 GB
ScreenCaptureKit/AVFoundation/media engine/HEVC/H.264/native retina scale/60 fps/separate audio tracks/Rec. 709/QuickTime .mov/hardware encode/local only/

Mac only, on purpose

Built against the system, not around it.

The encode is not your problem

Capture runs on ScreenCaptureKit and the HEVC encode runs on the media engine, so a 60 fps recording of a full display leaves the CPU free for whatever you are actually demonstrating.

One container, done properly

QuickTime .mov, with the microphone and system audio on separate tracks. It opens in Final Cut, Premiere, Resolve, QuickTime and Preview without a conversion step.

Nothing leaves the machine

No account, no sign in, no upload, no telemetry. The app makes exactly one network request, and it is to ask whether a newer build exists.

Keyboard, from anywhere

  • 5Open the capture bar
  • 2Record a drawn area
  • .Stop recording
  • CToggle the camera bubble

The numbers

Specifications, without the marketing.

Capture

Sources
Display, drawn area, window, camera
Display scale
Native Retina, up to 2560 × 1440
Frame rates
30 or 60 fps, locked
Multi-display
Hot-plug, no restart
Window tracking
Follows moves, resizes and display changes

Encoding

Codecs
HEVC and H.264
Acceleration
Media engine on Apple silicon, Quick Sync on Intel
Container
QuickTime .mov
Colour
Rec. 709, 8 bit
Camera bubble
Composited into the same pass

Audio

Sources
Microphone and system audio
Tracks
Written separately, same file
Microphone
AAC, 128 kbps
System audio
AAC, 192 kbps
Sync
Locked to the video clock

Files and privacy

Naming
Mode, date and time, templates in 1.0
Destination
Any folder you choose
Account
None
Uploads
None
Network use
One update check

In more detail

The questions the specification does not answer.

Why only .mov?

It is the container macOS encodes to natively, it carries HEVC and H.264 without a repack, and it keeps microphone and system audio on separate tracks. Every Mac editor opens it directly. More containers arrive alongside the editing release.

How accurate is the size estimate?

It multiplies pixels per second by a bits-per-pixel figure for the chosen codec, then adds the audio tracks. Screen content is flat and compresses well, so real files usually land a little under the estimate and busy video lands a little over.

Does the camera bubble cost extra performance?

Barely. It is composited into the same encode pass rather than recorded as a second file, so it adds roughly its own share of the frame and nothing more. The estimate on this page accounts for it.

What happens if I capture a window that moves?

The capture follows it. Window capture tracks the window itself rather than a region of the screen, so moving it, resizing it or pushing it to another display keeps the recording intact.

Is HEVC hardware accelerated on Intel Macs?

On any Intel Mac with Quick Sync, which covers everything macOS 13 supports. Where hardware encode is unavailable the app falls back to software and warns you before recording rather than dropping frames silently.

Are the two audio tracks really separate?

Yes. Microphone and system audio are written as distinct tracks in the same file, so you can mute, duck or replace either one in an editor without touching the other.

Ready when you are

Record the next one properly.

One download, no account, no upload step. Open it, pick a frame, and the file is in your folder before you have switched windows.

18.4 MB · build 412 · macOS 13 Ventura or later

  • Free forever, with a small badge on exports
  • Notarised by Apple, no Gatekeeper detour
  • No account and no card until you remove the badge