Riverside Review

Reviewed by Tom · Founder & lead reviewer

The problem it solves

Remote interviews live or die on the guest internet connection, and the tool recording the call usually has no way to improve it. A conventional call recorder captures whatever arrives over the wire: compressed audio, dropped frames, the moment the video froze while the best answer of the hour was being given. The guest sounded fine to themselves and terrible in the file.

That single failure is the most expensive thing in interview-based content, because it is unrecoverable. You cannot re-edit a frozen frame or restore compressed audio, and asking a guest for a second hour is a favour you get to ask once.

This product inverts the model. Every participant is recorded locally on their own machine at high quality, and the files upload progressively in the background while the conversation continues. What you end up with is studio-grade raw material from a conversation that happened over a shaky connection, which is a different category of problem solved from better compression.

Who it is for

The recorded fit is podcasters and video creators who record remote interviews and want broadcast quality without a studio. Remote and interview are the two load-bearing words. The value is proportional to how often you record someone who is not in the room with you.

A second strong fit is anyone publishing conversation-shaped video: expert panels, recorded webinars, customer interviews used as marketing. Separate tracks per participant make those far easier to cut than a single mixed recording.

The weak fit is a solo creator who mostly records themselves. If you control the room and the machine, local recording software you already own achieves the same result for nothing, and the subscription is buying a problem you do not have. Heavy post-production teams are a partial fit: they will record here and finish somewhere else, which is a reasonable use but changes the value calculation, because you are paying for the capture layer alone.

How the pricing is structured

We could not confirm current prices against the live pricing page for this update, so open it before buying. The figure recorded for this listing is nineteen dollars a month.

The structural detail that decides the real cost is recording hours. Lower tiers meter them, which is recorded here as a limit: usage-based recording hours on lower tiers require planning for heavy schedules. A weekly interview show is a very different consumer from a research project recording several conversations a day, and the sticker price treats them the same way only up to a point.

The honest way to price it is per episode rather than per month. Divide the monthly cost by episodes published and compare that to what the alternative costs when a recording goes wrong - not the usual case, the bad case. The subscription is insurance against an unrecoverable failure, and insurance is always mispriced if you compare it against the months when nothing broke.

Day-to-day use

Guests join from a browser link. No installs, no account, no technical support call before every interview. That matters more than any feature on the list, because the guest experience is the part of a remote recording workflow you cannot control and cannot rehearse.

For the host the session feels like a video call with a record button. Afterwards you get separate audio and video tracks for every participant, which is what makes post-production flexible: you can cut one person cough without touching anyone else, and you can mix voices that were recorded at different levels.

The friction is specific and recorded here, and it is worth knowing in advance because it is alarming the first time. The live call preview can lag even when the recordings are fine. You watch a stuttering guest, assume the episode is ruined, and then find clean files afterwards. New users need to be told this, or they will abandon a good recording mid-session.

Recording quality and the local-recording trade

Each participant is recorded locally in up to high resolution, so a bad connection does not degrade the final tracks. That is the whole architectural argument, and it is sound, but it has consequences worth understanding.

The first is that the recording lives on the guest machine until it uploads. Progressive background upload covers most of this, and the operational rule that follows is simple: do not let a guest close the tab the moment the conversation ends. Thirty seconds of chat while the last of the file goes up is the cheapest insurance in the workflow.

The second is that local recording is only as good as the local hardware and room. Studio-quality capture of a laptop microphone in a kitchen is a faithful recording of a kitchen. The architecture removes the network as a variable and leaves the microphone and the room exactly as important as they ever were, which is why asking a guest to use headphones remains the highest-return request you can make.

The AI editing layer

Recording is half the job and the AI tools handle a surprising amount of the other half. Transcription is fast and accurate, filler words can be stripped in a click, and the clip finder locates shareable moments and formats them vertically for short-form feeds.

Text-based editing is the most consequential of these: you cut the conversation by deleting sentences from the transcript. For interview content specifically this is the right abstraction, because an interview edit is mostly a decision about which answers survive, and that is a reading task rather than a waveform task.

The vertical clip feature deserves a realistic expectation. Automatic clip selection is good at finding moments with the shape of a highlight and indifferent to whether the moment is actually interesting, so treat the output as a shortlist to approve rather than a publishing queue. Even so, a shortlist of candidate clips is the part of short-form repurposing that people abandon first, so having it generated matters.

Limits and reasons to choose something else

Three limits are recorded and all three are honest. The live preview can lag even when recordings are fine, which unnerves new users. Editing depth still trails a dedicated editor for complex multi-camera projects. And recording hours on lower tiers need planning if your schedule is heavy.

The middle one sets the boundary of the product. This is a capture tool with a competent editor attached, not an editor with capture attached. Multi-camera work with real timeline requirements will be finished elsewhere, and that is a normal workflow rather than a failure.

When something else wins: if your editing needs are deeper than your recording needs, Descript is the stronger editor and the two are compared directly in Riverside vs Descript. If the problem is background noise on calls rather than recording quality, Krisp AI is the cheaper and more direct fix. And if you only ever record yourself, local software you already have is sufficient.

Alternatives and adjacent tools

The nearest alternative is Descript, which overlaps on text-based editing and transcription but comes at the problem from the editing side rather than the capture side. If remote guests are central to what you publish, the capture architecture is the decisive difference; if most of your material is already recorded and the work is assembly, the editor matters more. The head-to-head is Riverside vs Descript.

Adjacent rather than competing: Krisp AI for cleaning audio before it reaches the recording, ElevenLabs when a narration track or a pickup line is needed without booking the room again, and Canva Pro for the cover art, thumbnails and social crops every episode needs.

For how a publishing workflow fits together end to end, the shortlist in the best AI tools for podcasters covers capture, editing and the repurposing layer in one place.

Common questions

Why does the call look bad while it is recording? Because the live preview and the recording are separate things. The preview travels over the network; the recording does not. Lag in the preview does not mean the files are damaged, and new users should be warned about this before their first session.

Do guests need to install anything? No. They join from a browser link, which is the single biggest reason remote recordings get cancelled less often.

What happens if a guest closes the tab immediately? Uploads happen progressively, so most of the file is usually already up, but the safe habit is thirty seconds of small talk before anyone leaves.

Is it enough to replace an editor? For interview content, often yes. For multi-camera projects with real timeline requirements, no.

Does high-resolution recording need a strong connection? No, and that is the point - it needs a reasonable machine and a little patience while files upload.

Should you buy it?

If you publish interview-based content - podcasts, recorded conversations, expert panels - this is the strongest tool in its lane and the subscription earns its keep quickly. The reason is not the feature list. It is that it removes the single biggest unrecoverable risk in remote content, and every other improvement you could make to the workflow is smaller than that one.

Solo creators who mostly record themselves can get by with simpler setups and should, because the architecture that justifies the price is solving a problem they do not have. Heavy post-production teams will still finish in a dedicated editor, and should treat this as a capture layer and price it accordingly.

For everyone in between, the decision comes down to one question: how often do you record someone who is not in your room? If the answer is weekly, buy it. If the answer is twice a year, hire the studio for those two days instead and keep the subscription money.

Best for

Podcasters and video creators who record remote interviews and want broadcast quality without a studio.