The Basis Peak is a vast improvement over the Basis B1, that oh-so-promising but still quite imperfect activity-tracking wristband from early 2013. The Peak is more stylish than the B1, has a better display than the B1, and is easier to use than the B1.

But best of all, the new $200 Basis Peak has a more accurate heart-rate monitor than the B1. I've tested the Peak for about a week, and while its heart-rate feature isn't perfect, it comes close to what Basis claims it can do. And that's a big deal, because Basis is claiming a level of heart-rate sensor accuracy that has so far eluded activity-tracking wristbands.

The chase for real-time heart rate monitoring

The Basis story has always leaned on science--pure, reliable, incontrovertible science. Where wristbands from Jawbone and Fitbit have emphasized ease of use, pretty charts, and feelgood goals, the Basis bands have thrown a full battery of sensors in your face, practically daring you to dispute the veracity of their data.

But the problem is, the old B1 couldn't really deliver when it mattered the most. For all its sensors--an accelerometer, an optical heart-rate monitor, a skin temperature sensor and a galvanic skin response sensor--the B1 never promised real-time, continuous heart-rate data, like what you'd get from a chest strap monitor. Instead, the B1's heart-rate data was simply fed into the Basis algorithms that calculate calorie burn and sleep quality.

You were encouraged to do "spot checks" of your heart rate throughout the day, but Basis never claimed its original heart-rate sensor could match the performance of hardcore workout equipment.

Why? Because the sensor couldn't lock onto a user's heart rate, and track that proverbial ping through the lurching and jostling of physical exercise. The sensor was useful in painting a picture of a user's overall exertion levels, and it helped a lot in what Basis calls its Advanced Sleep Analysis, which (surprise) kicks in when you're sleeping and relatively motionless. But the old heart-rate monitor still had to conform to the laws of 2013 technology. It wasn't accurate enough for, say, zone training during a rollicking cross-country run.

New tech, new possibilities

But now we have the Basis Peak. Basis says a brighter LED in the spectroscopic sensor helps cut through the noise of "channel interference" (in lay terms: light pollution). The sensor also has an improved photoreceptor. This is the component that absorbs the LED light to capture a footprint of your blood flow, and thus heart rate.

Then there's the new sensor chassis. It sits on a raised berm that supposedly forms a stronger connection to your skin, almost like a sealed gasket. The Peak is also lighter than the B1, and is therefore less likely to break contact with your skin during spirited exercise. Finally, the Peak has stretchy silicone straps that help keep the sensor snug against your skin.

You're probably getting the idea by now: Before Basis felt comfortable claiming real-time, continuous heart-rate tracking, it needed a platform that could track blood flow through thick and thin, hell and highwater. So does the new Peak deliver? That's what I intended to find out. My control device was LG's Heart Rate Earphones set, which I've found to be accurate during A/B testing with a chest-strap heart-rate monitor.

The LG earphones match my chest-strap monitor beat for beat. They also keep heart rates locked during physical motion, and reliably respond to increasing and decreasing exertion levels. This is a very important point: During bursts of increased exertion, the LG earphones show faster heart rates in a logical, evenly scaled manner. And when I back off, the earphone's numbers scale down logically, as well. These LG earphones will never match the accuracy of a hospital ECG, but they're much more portable than an ECG, especially during spirited hikes up and down Mount Davidson, my testing playground.

Close, but not perfect tracking

There's no need to keep anyone in suspense: The Basis Peak did a reasonable job in matching the numbers reported by the earphones' smartphone app. Most of the time, the Peak was within 2 to 3 beats per minute of the earphones, and I found the two devices scaled pretty closely as I chugged my way up Mount Davidson as quickly as possible (and at a pace that still allowed me to record measurements on a notepad).

Ten minutes into the hike, the Peak displayed 133 bpm; the earphones, 131 bpm. Another 10 minutes later, the Peak showed 141 bpm; the earphones, 142 bpm. So far, so good--but then the Peak threw up some implausible readings as I kicked into overdrive getting close to the summit. The earphones were scaling evenly and logically with my increasing exertion levels, but for one reading the Peak actually slowed down, showing 145 bpm to the earphones 154 bpm.

It got worse when the Peak showed 129 bpm a brief time later. At this point, I was working even harder, and the earphones were reporting 165 bpm. But the glitch only lasted for a few moments. Finally, as I was steps away from the mountain's highest elevation, the Peak rapidly shot up, reporting 162 bpm to the earphones' 165 bpm. During my descent, the Peak returned to logical bpm levels, scaling evenly with the earphones.

Basis Peak versus Microsoft Band

So that was actually my second Mount Davidson hiking test. My first test, two days earlier, revealed similar performance. Every once in a while--twice during the first hike, once during the second--the Peak would show unusual (and untenable) heart-rate numbers. Then the band would correct itself, and return to reliable reporting.

It's not heartening to see any anomalies, but at least the Peak was notably more accurate than the Microsoft Band, another wearable that claims real-time, continuous heart-rate monitoring. I tested the two wristbands together for both hikes, and the Microsoft Band's untenable heart-rate swings were much more frequent (for example: 11 times during the second hike) and much more egregious (for example: showing 109 bpm when the Basis and LG wearables were showing 156 and 157 bpm, respectively). It's worth noting that Basis says the Peak samples heart rate 32 times a second, whereas Microsoft says its band samples heart rate once every second.

But let's not give the Peak a gold star quite yet. Besides the glitches described above, the Peak also suffered a number of dropouts where it wouldn't report any heart rate at all. I didn't keep count on my first hike, but it happened three times during the second hike. Granted, if the device can't get a lock, it shouldn't outright lie by giving the user an inaccurate reading. Still, when you're depending on your wristband for feedback, you want a reading when you want it.

I also tested the Peak on an elliptical machine. It's a less physically jarring use case than an uphill hike, but I made sure to use the machine's arm workout, guaranteeing the Peak would be subjected to constant (albeit predictable) wrist motion. Here the Peak lasted throughout 15 minutes of intense exercise without any dropouts. For the first two minutes, it under-reported my heart rate (for example: 130 bpm to the earphones' 146 bpm). But after that, the Peak scaled evenly with the earphones, never varying by more than 4bpm all the way to my max of 162 bpm.

Sadly, the Microsoft Band consistently under-reported heart rates during the very same ellipitical machine test. One typical reading: 115 bpm when the Basis and LG devices were reporting 156 bpm in unison.

Best sleep tracking around

I've penetrated into heart-rate monitoring because that's the Peak's marquee feature, but the band has a few other tricks that may be more important to many users. Obviously, the Peak records step counts. Because all these devices do. The Peak's step counts looked perfectly reasonable, and well within margins of error. Example: After I completed my first Mount Davidson hike, my Jawbone UP24 reported 4,702 steps while the Peak reported 4,689. I experienced the same levels of consistency throughout five days of testing.

Step-tracking is great, but I'm much more interested in sleep reporting. This is where my quantified-self instincts kick in, and this is where Basis' technology kicks into overdrive. One of the best things about the Peak is you don't have to touch any buttons or interface menus to begin logging an activity. Using what Basis calls Body IQ, the Peak can automatically determine when you've started a walk, run, or outdoor cycling session. And it can also use Body IQ to begin and end a sleep-logging cycle.

You can't view sleep data on the hardware, but the Basis mobile app reveals an industry-leading amount of information, including not just light sleep, deep sleep, and interruptions (you know, the standards), but also "toss and turns" and periods of REM sleep, the sleep cycle associated with dreaming and improved mental health. If you're more interested in what's happening to your body when you're not active, the Basis Peak delivers the best wristband experience available.

The mobile app also reveals a scads of data on daytime activity--certainly more than I need, as I'm interested in general fitness, not serious athletic training. But I do have a quibble with the way exercise events are logged. Remember those two hikes up Mount Davidson? The Basis app broke each hike into two separate exercise events because I briefly stopped to enjoy the view at the summit. Body IQ sensed my "workout" was over, and started a new session once I headed back down the hill. It's an annoying quirk, and there's no way to splice two related events together.

Better display, better design

All of Basis' quantified-selfie schtick would be undercut if the company's hardware didn't evolve as well. Well, finally, we have a Basis band that doesn't look something halfway between an engineering sample and a medical device. Relative to the original Basis B1 and Basis Steel update, the Peak is a much more stylish accoutrement.

The Peak also offers up to four days of battery life, which is less than what you'll get from a pure activity tracker (Jawbone delivers a luxurious 14 days), but vastly better than today's smartwatches. Kudos to Basis for being sensible and sticking with a power-efficient black-and-white display

Instead of plastic, the new case is made of forged aluminum. The horribly dim screen and backlight of earlier models has been replaced with a high-contrast Gorilla Glass 3 display. It's easier to read all around, and instead of navigating menus with a clunky side button, the new display is a touchscreen--and perfectly responsive at that. There are two color schemes, light and dark. My review unit was the Matte Black model that comes with a black strap with red accents.

There's only one watchface available (a rather meh digital readout), and this device will never take the place of a stylish watch, either luxury or mainstream. But you know what? I sort of like the design. If you fancy yourself the sporty type, this wearable will help you look the part. The texturing on the silicone strap looks cool. Ditto the red accents. And if you don't like strap, you can replace it for any other 23mm aftermarket option.

Comfort? It's great. Even with the Peak battened down tightly on my wrist, I never felt like I was wearing anything, either during the day or while sleeping.

Notifications will make this great band better

With its current software build, the Basis Peak is still just an activity tracker, however advanced it may be. Smartwatch-style notifications for texts, phone calls, email and more will arrive by the end of the year as a free update. It's nice to have something to look forward to, but keeping these features dormant at launch puts the Peak behind the Microsoft Band and numerous other wearables in terms of smartwatch-fitness band convergence.

Who knows, maybe the notifications just weren't ready. Or maybe Basis wants to schedule one more news blitz for the height of the holiday buying season.

Regardless, for now we have a very competent fit-tech device. The Peak costs more than trackers from Fitbit, Jawbone, Withings and others, but with its automatic activity logging, improved aesthetics, and good-but-not-perfect heart-rate monitoring, it's certainly one of the top fitness wristbands around.

Would I depend on its real-time heart-rate monitoring if I were a serious athlete who needed indisputable data anytime, anywhere during any type of workout? No, not based on my testing experiences. But the Peak's data is certainly good enough for less exacting scenarios--like recreational hikes and cardio training in the gym. And when Basis' algorithms marry its new heart-rate data to data gathered by its other sensors--measuring steps, body temperature and perspiration--you should get a get a very good idea of just how hard you've been working out.