I’ve been a pretty vocal supporter of the Even Realities G2, so as competing brands launch what is essentially the same product with small points of differentiation, I am very excited to see what they improve on… Fast forward to July, and XGIMI (the parent company of MemoMind) reached out about taking a look at their upcoming competitor glasses – the MemoMind Ones. So obviously I had to break them down…

These are a lightweight pair of smartglasses, but only by comparison to the very bulky Meta RayBan Displays – sitting at 48g measured, they are 11.4g heavier than my G2 daily drivers. This isn’t a huge deal, and surprisingly I found myself not really dealing with any headache during all day use, likely due to a larger footprint on the nosepads chosen (the Even Realities G2s have a small nosepad and can get pretty annoying after 4+ hours).

They also have a very low clamping force to the head, with the temples only applying up to 1N / 3kPa max at a relatively wide 18cm head-width. This low pressure is actually an issue with some glasses where the temples are relatively straight to accommodate a rigid 2.5D PCBA, but in the MemoMind One’s it actually feels just right. There is a very aggressive curvature to the rear of the temples that hugs the occipital bone (back of the head) nicely and makes them feel very securely positioned.

I’m getting into more and more audio testing, but baseline module amplitude is still a fun test as a relative data collection to understand which glasses can get loud (some too loud…) and where they are noticeable given the open-ear speaker nature for most of these. In the case of the MemoMind One’s they are definitely too loud at max volume – 77 dBA in-ear is pretty much danger territory for long stretches, and you will also give folks around you a 60+ dBA listen to your music/phone call. Around the medium setting these get much less noticeable at 3-6ft (~1-2m), but still output a very loud 63 dBA in-ear.

One new data collection I’ve started is the full-spectrum intensity collection with a calibrated mic. I put any data collected in reference of one of the better small form-factor speaker modules out there: my iPhone speaker (sorry Samsung folks). Really what I look for here is a relative plateau across the entire frequency range when outputting white noise (roughly equal intensity across all frequencies), but I’m sure some actual audio engineers can tell me why that’s the wrong thing to look for. Yes, there’s some non-linearity to our hearing as well… but from my few hours on Wikipedia and diving into NCBI open-access papers, I’m not sure how that translates to a better ‘shape’ to the spectral output.

Rant aside, MemoMind actually has one of the best smartglass speakers we’ve tested (so far). The output intensity is pretty even from 100 Hz to ~5 kHz, with steep dropoffs on either end which is pretty characteristic of these tiny speakers.

MemoMind actually has two audio profiles – the main / loud ‘Harman’ Mode and the ‘Quiet’ Mode, which I expected to have some sort of equalization settings that change the spectral response a bit. It turns out ‘Quiet’ Mode does just that – makes the entire output quieter… I measured a very similar spectrum that is just 5-10 dBFS lower across the entire audible range, with no obvious changes in the relative octave intensities.

Next up is battery, and this one was fun to assess since Memo has an excellent blog post detailing their own battery drain assessments and explaining what ‘mixed use’ actually means in their all day claims – nice and transparent!

My findings generally agree with theirs, albeit I tend to translate everything to an actual power draw (mW) value where they measure the actual capacity drained (mAh) of the battery. In reality, both are fine measures since generally most of these small batteries operate at ~3.85V, so it’s a simple conversion… And regardless of units, the glasses definitely hold up with the highest power draw of 215mW for navigation still comfortably giving you >4 hours of always-on time. In the realistic case where the glasses are on standby for a few hours, you’re listening to music for a few hours, using the display for some notifications & AI agent interaction, these will definitely hold up a full day on a full battery.

I won’t spend too much time on the display since it’s the same green JBD panel we’ve now seen across multiple devices – still a green peak of ~529nm with a respectable FWHM of 31nm.The big new thing that I started measuring here is the actual field of view (FoV) versus the claimed FoV. I actually show the comparison of the MemoMind Ones to the Even Realities G2s below, with as full of a projected screen as I can get using my block test pattern – if you count the 1×1 degree squares that the image covers (text and small scroll lines count, since those are technically rendered content…), MemoMind has a ~25° horizontal FoV and ~14° vertical FoV, meaning the diagonal FoV is just shy of 30°. This is actually the first glass that undersells their FoV (claims only 25° diagonal) as most will claim the upper boundary of what they realistically can achieve – see the Even Realities calculated 26° diagonal FoV, compared to their claims of 27.5°.

We’ll end this TED talk on the optical diagram which…remains unchanged from the last 4 monochrome green smart-glasses we covered. At this point I’m just recycling the diagrams (though the #2 aspheric field lens is actually an oversimplification for some of the units), but the principle of operation hasn’t changed here. Are we finally starting to see a stable hardware configuration that the brands will be competing on?

As always, all raw data collected and used here is available for all tiers of paid Patreon subscribers here: www.patreon.com/c/DisplayTrainingCenter