Last January, I stood in a storage room with 48 sealed boxes. Every box said "HP Reverb G2" on the side. The purchase order listed it as hp-reverb in our inventory system. A regional hospital network was expecting a VR training lab, and some of those headsets were going to be used for virtual reality exposure therapy sessions. My job was to certify that every unit matched the approved HP Reverb G2 VR headset specifications before it came anywhere near a patient.
I'm the quality and brand compliance manager at a VR training company. I review roughly 200 headsets a year. During our Q1 2024 audit, I rejected 12% of first deliveries because of spec mismatches, which sounds aggressive, but for this industry it's necessary. Five minutes of verification beats five days of correction.
The order that almost slipped through
The first sign of trouble was a serial number. We had an approved list from our vendor: specific model numbers, firmware versions, and lens spacing calibration ranges. The 48 boxes all had the right model number printed on the outside. But when I pulled one headset out, the serial number didn't match the approved sequence. It wasn't randomly different. It was one digit off from a batch that had a known tracking issue.
The vendor told me it was "the same thing, just a newer firmware." That was a red flag. I've heard that phrase before. What they didn't say was that the firmware image on those units was an older build—one that had a patch for SteamVR tracking missing. From the outside, a VR headset is just a display strapped to your face. Inside my checklist, it's a tracking system, a pair of precision lenses, a cable management problem, and a software stack. A small firmware difference can change how a headset behaves in a clinical simulation.
Why the spec sheet matters more than the box
The HP Reverb G2 specs are the reason our procurement team specified it in the first place. According to the HP Reverb G2 product documentation we use as our baseline (hp.com), the key specs are covered by dual LCD displays at 2160x2160 per eye, a 90Hz refresh rate, a 114-degree field of view, four tracking cameras, six degrees of freedom, and integrated spatial audio. Those numbers are solid for enterprise PC VR training. But raw specs only help if they're actually in the unit you receive.
Most buyers focus on resolution and completely miss the details that cause problems. Interpupillary distance. Refresh rate. Firmware. Controller tracking. Cable length. The question everyone asks is "what's the screen resolution?" The question they should ask is "what happens when a controller loses tracking for half a second during a therapy session?" Because that half-second can break the immersion and, more importantly, the trust of the person wearing the headset.
The surprise that changed the review
I did what I normally do: I got one unit fully charged, updated the firmware, and connected it to our SteamVR test rig. The first test was the standard boot and display sequence. The second test was tracking. In the middle of the routine, the right controller floated away from the virtual hand for a solid second. Then it snapped back. On a consumer game, that might be an annoyance. In an exposure therapy environment, that kind of glitch can make someone feel instantly unsafe.
So we rejected the entire batch. The vendor had to redo the order at their own cost. It delayed the project by about three weeks, and I won't pretend that wasn't painful. But the alternative was worse: delivering 48 headsets that didn't meet our own checklist, then waiting for something to fail during a live session.
What is virtual reality exposure therapy?
This project also forced me to explain something to our internal stakeholders. What is virtual reality exposure therapy? In plain terms, it's a clinical technique where a therapist uses a simulated environment to expose a person to a feared or avoided situation in gradual, controlled steps. The therapist controls the anxiety levels, so the hardware has to respond predictably.
That changes how you review a headset. Refresh rate and tracking accuracy are not just technical specs. They're clinical variables. If the image jitters or the controller drifts, the patient might associate the scene with anxiety, not the gradual learning the therapist is trying to build. I'm not a therapist, and I can't speak to clinical efficacy. But I can verify the hardware. And I can tell you that the checklist for a therapy deployment is stricter than the checklist for a gaming setup.
Why not a PS5 headset or Galaxy XR headset?
At one point, someone in procurement asked why we just didn't buy a PS5 headset or a Galaxy XR headset and save money. I have mixed feelings about that question. On one hand, I understand the appeal of a lower unit price. On the other hand, our training modules are built on SteamVR and PC VR infrastructure. A PS5 headset is designed for a closed console ecosystem. A Galaxy XR headset belongs to a different product category entirely, with different tracking assumptions. It's not about which one is "better." It's about which one matches the approved workflow. The HP Reverb G2 is what our spec list asks for, and SteamVR compatibility is a large part of that.
The checklist that paid for itself
The result was a new 12-point incoming inspection checklist. Serial numbers get verified against the approved batch. Firmware versions get recorded before and after update. Tracking tests run for a minimum of two minutes. IPD adjustment gets exercised on both ends of the range. Audio output is tested for both speakers.
That checklist grew out of a mistake I made two years earlier. I still kick myself for not documenting an email promise from a vendor. Now, if it's not in the checklist, it doesn't exist. The process is boring. It's repetitive. It's also the cheapest insurance I know.
Bottom line: for high-stakes VR training, the checklist needs to be longer and more boring than people want it to be.
And before you ask: this won't apply to everyone. My experience is based on enterprise PC VR deployments with a specific target population. If you're buying a headset for home use or console gaming, your experience might differ. I can't speak to that setup. I only know what happens in our lab and what the FTC guidelines expect us to substantiate when we make claims about suitability.
Per FTC advertising guidelines (ftc.gov), claims about what a product can do for therapy or training need evidence. In my role, that evidence starts with a documented hardware review. It's not just internal paranoia; it's a compliance issue.
Lessons for anyone buying VR headsets
If there's a lesson beyond this specific product, it's this: don't trust the outer box. Check the actual unit. Write down your specification baseline before you receive anything, not after. If a vendor says "it's the same thing," ask for the exact version numbers and put them in the order contract.
I can't promise that will prevent every problem. But it would have prevented this one. The HP Reverb G2 is a capable headset when the specs are correctly verified. What I won't do again is approve a batch based on the box art alone.