Audio Hijack 4.6.0 Mac download

Audio Hijack's session grid is where it all comes together. You drop audio sources — Safari, FaceTime, a USB mic — into a pipeline, then add blocks for effects or transcription. The result is a recording setup you can save and reuse without touching a single setting next time. It works with anything that produces sound on a Mac. iTunes, DVD Player, Skype, internet streams, even analog inputs. Built-in templates get you started fast: one click for a podcast, another for a VoIP call. Post-processing tools like Denoise, Declick, and Dehum clean up the raw file, and you can export to FLAC if you want lossless output. Version 4.6.0 adds support for macOS Golden Gate and updates the Transcribe block with a recommended model. The underlying Whisper.cpp backend moved from 1.7.4 to 1.9.2, which means more accurate speech-to-text. If you need to capture system audio or any app's output reliably, Audio Hijack is the tool. It runs on Intel and Apple Silicon Macs with macOS 14.4 or later.

macOS 27 Support Without Workarounds

Version 4.6.0 is a maintenance release with two concrete changes that affect how you work day to day.

• macOS 27 (Golden Gate) support. If you're on the latest macOS, the app installs and runs without workarounds. No compatibility headaches when you update your OS.
• Updated Transcribe block. The new recommended transcription model saves you from hunting through model dropdowns. Select "Recommended" and the Whisper.cpp backend — now 1.9.2 — picks the best option. The result is cleaner transcripts with less fiddling.

Neither change rewrites your workflow, but both remove friction. You spend less time configuring and more time recording.

Key Questions Before Installing

Can I record a FaceTime call without hearing my own voice?

Yes. Audio Hijack routes the call audio into its pipeline and lets you mute your own microphone feed. Set up a template once, and every call records clean — no echo of your side of the conversation.

Works with Any macOS Audio Input

It records whatever macOS sees as an audio device. If your Bluetooth headset shows up in System Settings as an input, Audio Hijack picks it up. The same applies to USB microphones and built-in laptop mics.

What if I need to record from two apps at once?

The session grid supports multiple sources simultaneously. Drop Safari and Skype into separate channels, apply effects per channel, and export a mixed or separated file. Good for podcast guests who want to hear a YouTube clip while you talk over it.

Can I batch-process old recordings?

Audio Hijack is built for live capture, not batch processing. For that, you would use Rogue Amoeba's other tools. But you can re-route a file through the pipeline — play it from iTunes, pipe it into Audio Hijack, and apply Denoise or Declick as it plays. Not as fast as batch, but it works for a few files.

Does the Transcribe block work offline?

Yes. The Whisper.cpp backend runs locally. No internet connection needed for transcription. That matters if you record in a home studio with spotty Wi-Fi or need to keep audio data on your machine.

Workflow: Record a Clean Interview From Skype

This scenario: you are interviewing someone over Skype. You want a clean recording of both sides, minus Skype's compression artifacts, and a transcript for editing later.

1. Open Audio Hijack. Select the "Skype Call Recording" template from the startup dialog. It sets up two audio inputs — your mic and Skype's output — in a single session.
2. Hit the record button. The session captures both channels separately. Talk to your guest as usual.
3. When the call ends, stop recording. The raw file includes both tracks.
4. Apply Denoise to each channel. This removes background hum and keyboard clicks. Adjust the strength slider until the background drops out without making voices sound hollow.
5. Add a Transcribe block to the output chain. Set it to "Recommended" model. Run the transcription. It processes the local file and produces time-stamped text alongside the waveform.
6. Export the final file as FLAC. You now have a lossless archive of the interview and a text file that syncs to timestamps.

Save the session as a custom template. Next interview: open template, record, export. No setup beyond the first time.