Performance and cooling

Start with System Health

Add the System Health panel to your switcher layout, then tap it for details while the switcher is running. Its summary shows the overall performance state, thermal state, process CPU, and GPU frame budget. The detailed view identifies the primary constraint and shows recent changes during this app session.

• Thermal state moves from Nominal to Fair, Serious, or Critical as the device experiences more heat pressure. Nominal does not guarantee that every other part of the production has enough headroom.
• Source Workload separates Configured, Enabled, and Active demand. Active demand counts sources currently requested by Program, Preview, multiviews, and other live consumers. A disabled source does not create the same workload as a live one.
• Processing Load shows CPU, GPU frame budget, Metal working set, and app memory limit, along with display and compositor timing. These percentages describe workload relative to a budget or limit; they are not simple hardware utilization readings.
• Monitor Health and Output Health show whether active displays and destinations are sustaining their requested frame cadence. An inactive output is Unverified until it starts.
• Real-Time Delivery and Output Counters help you spot missed deadlines, failed frames, recording or streaming drops, stalls, and failures. Real-Time Delivery counts cover the latest one-second sample; recording and streaming counters accumulate during the current run.

Watch the pattern over time. A rising thermal state with a worsening performance state suggests heat is part of the problem. If thermal state stays Nominal, follow the panel's primary constraint and output health instead of assuming the device needs more cooling.

Keep the source workload manageable

1. Start with the sources you need on air. Add local cameras, network feeds, and playback sources gradually. Watch Active demand as you bring each one into Program, Preview, or a multiview.
2. Disable sources you will not use. A long configured list is less important than the sources actively requested, but disabling unused inputs makes the setup easier to manage and can keep them out of live demand.
3. Simplify monitoring. Close unneeded source monitors, multiviews, or other live previews. They can keep sources active and add rendering work even when Program Output looks simple.
4. Choose practical formats. If the device remains busy, test fewer simultaneous high-resolution or high-frame-rate inputs and a less demanding output format. Change one factor at a time so you can see what helps.
5. Rehearse the full production. Run the intended sources, recording, streaming, and displays together for as long as the event requires. Check System Health and the actual recording or stream, especially near the end of the test.

There is no universal safe number of sources. Device model, source formats, previews, effects, network conditions, and simultaneous outputs all affect the result.

Use Thermal Protection

Open Settings → Performance → Thermal Protection. Automatic limits nonessential preview and multiview updates to 15 fps when thermal pressure rises; Always On applies that monitoring limit immediately. The same screen shows Thermal State and whether the monitoring frame rate is reduced. You can also turn Thermal Protection off.

Program, recording, streaming, and transition timing keep their selected rate under this monitoring limit.

Give the device room to cool

• Keep the device out of direct sunlight and hot enclosed spaces. Allow airflow around it and avoid covering the back with insulating material during a long session.
• If charging adds heat, use a suitable power setup and consider removing a case that traps heat. Keep cables and connectors dry and unobstructed.
• An external fan or compatible phone cooler may help with sustained workloads. A thermoelectric (Peltier) cooler can become very cold: follow its maker's instructions and stop using it if moisture or condensation appears. Do not attach anything wet to the device.
• If the device displays a temperature warning, stop the production if you can, move it to a cooler place away from sunlight, and let it cool before resuming. See Apple's temperature guidance.

Cooling helps with heat pressure; it cannot fix an overloaded CPU or GPU, a weak network connection, or a failing destination. Use System Health to decide what to change.

Need help interpreting a warning?

1. In SwitchEffect, open Settings → Diagnostics and turn on Enable Diagnostics Logs.
2. Reproduce the issue, then note the warning and what System Health reports.
3. If support asks for a log, open the System Health details sheet and export the diagnostic log.

Email support@switcheffect.app with your device model, OS and app versions, source and output setup, and what System Health reported. Attach the exported log only if requested. Avoid sending private production content unless requested. See the Privacy Policy for what Diagnostic Logs contain and how they are shared.