Choose between Zencastr and Riverside by the guest devices and the files you need after recording. Zencastr is a concrete option for a supported Linux computer or a free session with up to six participants. Both document Android phones: Zencastr through Chrome, Riverside through its app. In either case, a finished conversation is not proof that the editor has received complete local tracks.
Start with the guest's actual device
Zencastr's requirements list Chrome, Edge and Brave on desktop, with Windows 10 or later, macOS 10.14 or later, or a current Linux version. On mobile, it supports its app on iPhone and iPad with iOS or iPadOS 16 or later. For Android, its help page lists Chrome. For mobile recording it names the iOS app or Chrome on Android.
Riverside lists Chrome, Edge and Safari 18.2 or later for computer recording. Safari users must remain on the Riverside tab, and some studio features are unavailable there. Its mobile recording route is the iOS or Android app, not the mobile browser. Linux and ChromeOS computers are listed as unsupported.
Local recording needs local space. Zencastr recommends at least 8 GB RAM and 10 GB free disk space. Riverside recommends at least 5 GB available to the browser, with an additional storage check for sessions exceeding 1.5 hours. Both advise against private or incognito recording. Zencastr explains that closing an incognito window can make local backup recovery impossible after an upload problem.
Ask guests about devices when arranging the session, not when they arrive. A link that opens a page is not proof that the device supports recording. Give a phone guest time to install the required app and select the intended microphone.
What Zencastr's free recording offer includes
Zencastr's recording page advertises unlimited recording time, separate participant tracks and up to six people total on the free plan. Computer guests can join by link without creating an account. The page describes local capture followed by progressive cloud upload, with local copies and server backups. It also describes automatic syncing to connected Dropbox or Google Drive accounts.
The current English pricing page labels the free offer Vibecastr and lists unlimited separate audio/video recording and download, with up to 4K video and 48 kHz WAV. These are published capabilities, not evidence that a guest's camera or room will produce professional footage. The same pricing page labels MP3 as 44 kHz, while the recording page groups WAV and MP3 under an up-to-48-kHz claim. Treat WAV as the clear 48 kHz option and inspect an MP3 export rather than assuming the descriptions are identical.
Riverside's status guide says finished high-quality tracks can be downloaded as WAV or constant-frame-rate MP4. It also distinguishes those local tracks from cloud recordings available for reference while uploading and processing continue. A cloud backup is useful, but it should not silently replace a promised local master in the editor's delivery folder.
Read the completion status before saying goodbye
Riverside documents four stages: Recording, Uploading, Processing and Ready. Guests should keep the browser or app open until the upload is confirmed successful. Once a track is Processing, its upload has finished and the guest can close Riverside. Ready means the high-quality track is available. Different participants can be at different stages at the same time.
If the host closed the studio without stopping recording, a track may still show Recording. The guide tells the host to return and stop it. If a guest left before uploading finished, the recovery route is to reopen the mobile app or visit Riverside's upload page. The status guide also describes partial-track processing beginning fifteen minutes after the call or recording ends. A partial track remains partial; check its displayed duration.
Zencastr's troubleshooting guide gives similarly practical instructions for a stalled upload: keep the recording tab open, check the connection and allow it to finish. After a crash or closed tab, reopen the recording from the Library and let pending uploads complete. It says local capture can continue through a brief connection drop, but this does not mean a remote conversation works without a network.
Zencastr's marketing page describes files as ready when a session ends. Its troubleshooting guide nevertheless documents unfinished uploads. Use the visible completion state rather than interpreting the marketing line as a guarantee that every guest can close immediately.
Choose a handoff, not just a resolution
| Decision | Zencastr | Riverside |
|---|---|---|
| Guest uses Linux | Current Linux listed with supported desktop browsers | Linux listed as unsupported |
| Guest uses Android phone | Chrome on Android | Android app, not a mobile browser |
| Guest uses iPhone or iPad | App with iOS/iPadOS 16 or later | Use the mobile app |
| Computer browser choice | Chrome, Edge, Brave | Chrome, Edge, Safari 18.2+ with restrictions |
| Clear free-session fact | Up to six participants and unlimited recording advertised | No equivalent free allowance established by the cited status guide |
| Delivery check | Complete pending local-track uploads | Separate Uploading, Processing and Ready states |
Imagine a hypothetical three-person interview where everyone has a supported computer and the editor wants separate WAV tracks. Zencastr's advertised free recording scope makes it a reasonable first rehearsal. If one guest can only use an Android phone, the question is whether that guest can open Chrome (Zencastr) or install an app (Riverside). If a guest can only use Linux, the device argument points the other way.
Neither recommendation ranks microphone quality or transcription accuracy. The useful question is whether the required people can record, finish uploading and deliver usable tracks through the documented workflow.
Rehearse the delivery in ten minutes
Use two devices that match the intended guests. Spend two minutes checking browser or app support, microphone selection and free storage. Avoid private browsing. Make sure the guest understands that stopping speech is not the final step.
Record for two minutes. Say a distinctive sentence at the beginning and another at the end. Include a brief overlap between voices, followed by silence. These markers help you check the delivered files without pretending the exercise is a formal quality benchmark.
Spend three minutes following the normal stop and upload procedure. Note the actual status for each participant. If an upload takes longer, keep waiting rather than closing merely because the exercise's ten minutes have elapsed. Completion depends on file size and connectivity.
Use the remaining three minutes to download available tracks and inspect the beginning and end. Open the files together in the editor you intend to use and check alignment. Confirm that the delivery contains the requested local tracks, not just a convenient backup. Write guest instructions using the displayed status wording.
Where Notibo belongs after the handoff
Notibo can turn an allowed recording into a transcript, summary and key points after you have verified the source files. It accepts audio or video uploads up to 50 MB. A large production master may therefore require a suitable smaller working file, while you retain the original for editing and checking.
Notibo is not the remote recording room, a video editor or a bot that joins the call. It does not provide transcript timestamps or promise speaker labels. Processing happens in the cloud, with EU storage and US processors. Choose one main language for the material; language changes within a recording can cause errors. Use Notibo when notes are an additional deliverable, and keep Zencastr or Riverside responsible for the recording workflow you rehearsed.
Related comparisons
- Riverside vs Zoom: recording a remote interview
- Trint vs Sonix: choosing a transcript review workflow
This is a documentation-based comparison by Notibo, checked on 20 October 2026, not a measured accuracy benchmark. The examples are suggested trials, not reported test results.