What an office needs that a bedroom machine doesn’t
Most white noise machine reviews you'll find are built for sleep — quiet dark rooms, one person, a solid eight hours of silence. A desk is its own beast: you're trying to bury a coworker's voice or keep yours private in a space where people are constantly moving and doors keep slamming, usually all day long. That's not the same problem at all, which means what you're actually looking for shifts completely. Battery life is irrelevant. What you need is sound that stays consistent and doesn't wear on your ears, and it absolutely can't turn into an obviously looping track by mid-afternoon.
Mechanical fan vs digital sample loops
This is the one distinction worth understanding before you buy anything on this page. The Yogasleep Dohm Classic and the SNOOZ both use an actual spinning fan to generate sound — it’s non-looping by nature, because it’s not a recording, it’s a motor. Independent testing from Reviewed measured the Dohm Classic at 41–52 dB and the SNOOZ at 46–87 dB across its ten volume steps, which is the only outside-measured acoustic data I have for anything in this category, and I’m not about to claim I have more.
Everything else here — the Homedics SoundSleep, the Dreamegg, the Calm Me — plays a digital sample on a loop. That’s not automatically worse. It gets you variety (rain, static, ocean, whatever), and a decent loop is long enough that most people never notice the seam. But if you’re the type who notices a Spotify song looping in an ad, pay attention to that when you’re reading a product page, because I can’t measure loop length for you — nobody’s published it.
What actually matters when picking one for a desk
Skip past any spec that screams "bedroom gadget." Sleep modes, scent diffusers, and mood lighting don't belong at your desk. What actually counts: can it get loud enough to blanket a genuinely chaotic open office (not just handle a quiet room), does it sit on your desk or plug straight into the wall where space matters, and do you need your phone synced up just to make it work — the SNOOZ requires that for its scheduling, which some people find annoying rather than useful.
If you’re buying this specifically to stop a nearby coworker’s meeting from bleeding into your calls, that’s a masking problem, not a soundproofing one — nothing on this page or any desk fan will block sound, it disguises speech intelligibility by adding a steady layer over it. The Calm Me is built and marketed for exactly that office-privacy use case rather than sleep, which is why I’ve got it ranked for shared desks specifically.
The Dohm Uno situation
I want to flag this one directly because most roundups list the Dohm Uno as simply “the cheaper Dohm” and move on. It’s rated 4.2 against the Classic’s 4.6 — a real, consistent 0.4-star gap, not rounding noise. The documented reasons, straight from Yogasleep’s own support pages: it’s single-speed and some owners find it too quiet to mask anything in a real room, dust can build up on the internal motor and change the sound over months, and some housings run warm (Yogasleep says that’s normal and there’s a fuse). Save $13 if you must, but go in knowing why it’s cheaper.
How I picked these
I did not run any of these in a lab, and I don’t own a decibel meter I’d trust for this. What I did: read the Amazon product pages and spec sheets for every model here, checked which acoustic figures had actually been independently measured versus which were manufacturer claims, and cross-referenced Yogasleep’s own published support notes for the Uno-versus-Classic gap. Where I couldn’t verify a number — loop length, exact sound counts, filter specs — I left it out rather than guess. If you want the fuller sound-machine lineup including the higher-volume options, I’ve got a separate page on loud white noise machines and what’s actually measured versus advertised. For the rest of the desk comfort category, start at /comfort; if your actual problem is a noisy meeting room rather than a noisy neighbor, that’s more of an /audio or /presentation fix than a white noise one.