However, in a recent video, i showed that the heart rate tracking of the very reasonably priced mi band 6 seemed to have improved significantly in 2022 due to firmware updates. This is really good news, given that not everyone can afford or want to spend hundreds of dollars on a smartwatch. However, how about all the other health and sports tracking features? Are these any good on the mi band 6? Well thats? What well test in this video? Specifically well test the sleep tracking oxygen, saturation, measurements and step counting and well also summarize the heart rate testing from last week to provide a complete scientific overview of the mi band 6. and to summarize overall im quite encouraged by the improvements of the mi band. 6. Though not all features are that accurate, with rumors of the mi band 7 on the horizon. Some of you might be waiting for this to be released after the performance. Changes of the mi band 6 actually shows that firmware updates months after release are sometimes needed to get proper performance out of a smartwatch which might make the mi band 6 a more attractive purchase for some people. Now, for those of you that are new to the channel, my name is rob and im a postdoctoral scientist specializing in biological data analysis. To give this video some structure, i want to start off by showing you the things the mi band 6 is bad at then. Move on to the things it does decently and close off the things it does well.

So, to start with the thing the mi band 6 was worth that out of all the health related features is sleep tracking. I tested that in several ways. The first is by checking if two mi band 6 is worn at the same time, detect the same sleep stages. Now, if the sleep tracking algorithm works well, it should give roughly the same results for the two mi band 6s. However, if the two watches are not consistent, that means that either the raw measurements themselves or the algorithm underneath is not very stable. For three knights, i wore one mi band 6 on my left arm and one on my right arm both running the same firmware. Now the fact that i wore one on my left and one on my right arm could introduce some differences in the movement data that both watches record during sleep. However, this is exactly the thing that a good sleep tracking algorithm should be able to deal with. If not, this means that its inherently unreliable at estimating your sleep stages lets see if thats the case on the left here is the mi band 6. I wore on my left arm and on top is the one i wore on my right arm: ill refer to them as the left and right mi band throughout this video for convenience. But there are both me ben sixes running the same firmware now in this table. I took the right mi band 6 as the reference, and i checked what the left me band 6 predicted for each of its sleep stages, which means each column here sums to 100 percent now i know this might be a bit confusing but ill.

Give an example. Take this first column here, for instance, this shows that, out of all of the deep sleep, the right mi band, 6 predicted 21.4 percent was also predicted as deep sleep by the left me ben 6.. However, much more was predicted as light sleep by the left. Mi band 6 about 67 percent indicating a poor match between both devices light sleep matched a bit better between the two mi band 6s, with about 72 percent of what the right mi band 6 predicted. As light sleep also predicted slightly by the left one. If they did disagree, this was mostly with deep sleep and ram sleep now ram sleep had a really poor match, with only about 30 of the ram sleep predicted by the right. Mi ben 6 also predicted his rem sleep by the left mi band 6.. Most of it was actually predicted as being light sleep by the left. Mi band 6. now in terms of awake detection, disagreed pretty well between the two mi band 6s, with close to 90 of what was awake time according to the right mi band. Six also predicted as awake time by the left me band six. If they did disagree, this was mostly because of the left knee bend, six detecting light sleep, whilst the right one was detecting me as being awake. So what does all of this mean? Well, it indicates that the algorithm used by the mi band 6 is not very stable and it will give vastly different sleep tracking, depending on which arm you wear it on now, while i was sleeping, it could of course be that the movement of my right arm Was different from the movement of my left arm, however, there is no excuse because in reality i went through certain sleep stages and if a device claims to be able to approximate this, it should give the same results no matter what arm i wear it on at Least thats my take on things, but we can do some more testing to confirm this as a secondary test.

I also wore this eeg headband that can actually measure my brain waves and was specifically created to track my sleep stages here. I display a similar plot to before, but now using the dream, 2 eeg headband as a reference, and i will focus on the left me band 6, for which i had 6 nights of data for getting an overall impression of how well the watches perform the dream. 2 should likely be good enough as a reference, however, its not perfect and the gold standard would be polysonography, which you would also like to try in amiban 6 in the future. First of all, deep sleep detection did not agree well with the eeg headband, with an agreement of about 40, and this is not very good. In fact, almost 55 percent of what was deep sleep according to the eeg device was predicted as light sleep by the mi band 6. light sleep showed a somewhat better agreement with 67 of what was light sleep according to the eeg headband also predicted. As light sleep by the mi band 6., if it was predicted differently, this was mostly as deep sleep and rem sleep, similar to what we saw with the consistency between the two watches ram sleeve detection agrees very poorly with the eeg headband only about 25 percent of What was ram sleep according to the eeg headband was also detected as rem sleep by the mi band 6.. Instead, almost 2 3 of what was ram sleep according to the eeg headband was detected as light sleep by the mi band 6.

. Now awake detection agreed a little bit better at about 54 and if it did disagree, this was mostly with light sleep which we see often since light sleep is the closest sleep stage to being awake now, these results are eerily similar to the results we saw when Comparing the two mi band 6s to each other, both results show especially poor rem, sleep agreement and also quite bad deep sleep agreement, and we can put it into further perspective by comparing the mi band 6 against many of the other watches and sleep trackers ive tested Over the last two years, specifically, we can compare the mi band 6s agreement with the eeg device, with the agreement of many other watches along the horizontal axis. We have the average agreement over the four individual sleep stages and on the vertical axis we have the agreement of the worst sleep stage, the better the agreement overwatch with the eeg device, the more to the top right. It is, and, as you can see, the best degreeing devices include different fitbits whoop straps and a widthing sleep analyzer. If we now plot the results for both me band 6s. In the same plot, we can see that theyre, really amongst the worst sleep stage, trackers out there with an average agreement of about 50 percent. What this means for me is that i would not rely on their sleep stage tracking to make any statements about my sleep quality, so the sleep stage tracking of the mi band 6 was not very good.

However, it was pretty decent at detecting the moment. I fell asleep and attacking the moment i woke up, so it can be used to track your total time spent in bad with that extenuating circumstance in mind id like to give the sleep stage tracking of the mi band 6 2 out of 5 stars. I should say that sleep tracking was truly the worst feature of the mi band 6 and all other features appear to perform significantly better one feature enemy band 6 appears to do better, is measuring your oxygen saturation or spo2. Over the last week, i measured my oxygen saturation at ground level in the morning and evening using the mi band 6.. At the same time, i also recorded my oxygen saturation with a dedicated finger pills. Exhibitor oxygen saturation basically indicates the percentage of red blood cells in the bloodstream that contain oxygen at ground level. My oxygen saturation should be in my normal range, which is generally between 97 and 100, and it should not fall below roughly 95 percent on the left. Here are 35 measurements taken with me band 6 and on the right matching measurements taken with the fingerprint exhibitor. As you can see, the mi band 6 is mostly within my normal range of spo2 values. The values do tend to be a bit lower than those taking with the finger pulse exhibitor, but overall theyre, mostly in the expected range. So this is looking pretty good. However, can the mi band 6 detector lowered oxygen saturation to test that i took the mi band 6 with me on two airplane flights in the plane, the pressure in the cabin is decreased, which effectively lowers the oxygen concentration before moving to the results.

If this video is proving interesting to you, a sub to the channel and a like, or a comment on, this video would be amazing. Now to the results in pink here, i plotted how my oxygen saturation changed during a flight as measured using a dedicated fingerprint exhibitor. As you can see, my spo2 value started out normal and as the plane ascended my oxygen saturation decreased and its state load during the flight. Then, as we descended it increased again and got back to normal levels here you can see the measurements taken by the mi band 6 in green dots, and indeed it seems that during the flight, the values tend to be lower than before and after the flight, which Is really good, though there is still definitely some variation or noise now to put that into perspective here on the left are the values we looked at before that i took at ground level and, as you can see, the values in air do tend to be lower Compared to those at ground level – and i also took a second flight – which you can see right here – we again see lower values during the flight and more or less normal values before takeoff and after landing in a similar range as those taken at ground level. In the morning and evening, to me, this indicates that the spo2 monitor of the mi band 6 is able to at least more or less detect a lowered oxygen saturation level with some variation overall.

This is pretty good, though i would recommend taking several measurements. If you get a low value and later confirm it with a dedicated oximeter still based on all this data id give the spo2 sensor of the mi band 6 4 out of 5 stars. The next thing the mi band 6 was ok or even pretty good. At is step counting to test the step counting accuracy. I went out and took exactly 4 000 steps, while wearing 2 mi band 6s, one on my left and one on my right arm now, since i do not like counting 4 000 steps in my head, i manually counted each step using this telecounter. I actually counted my steps in four segments of 1000 steps, switching the telecounter between my left and right hand, which is what the left and right labels refer to here, and i wore one mi band 6 on my left and one on my right arm. Now. These numbers right here are the actual steps counted for each of the four segments by the mi band sixes. As you can see, the number of steps counted is pretty close to the actual 1 000 steps i took. They were a maximum of 33 steps off, which is not bad at all. Most of the time they were even within 20 steps of the correct 1000 steps, or even less than 3 steps away. Now. This shows us how good the mi band 6 is at counting steps when it is supposed to count steps, but doesnt count any steps when its not supposed to count steps with that i mean.

Does it count steps when doing other activities that do not involve walking like cycling indoors or when weight lifting or when typing at my computer? Well, it also appears to be pretty decent on that front. First of all, lets look at one two three times i was cycling indoors for half an hour here. Those steps are displayed in these yellow boxes right here and, as you can see, it generally counted about 300 steps per 30 minutes while cycling indoors, which means about 10 steps per minute now. This is not that bad, given that, while walking, would normally count about 100 steps for a minute and fitbits, for instance, are notorious for counting as many steps per minute, while indoor cycling as well actually walking and to some degree at least it would make sense as Step counters confuse the motions made while cycling indoors with walking, but it shouldnt be too much. Next lets take a look at strength, training or specifically, weightlifting, and two examples are displayed here, and we see that for this half an hour of training, it counted 33 steps and for this half an hour 79 steps now this is not bad at all, since this Means roughly one to three steps per minute. Finally, while i was working on the script for this video, i checked how many steps were counted whilst typing for 30 minutes when wearing one me band 6 on the left and one on my right arm.

One watch counted 22 steps and the other 24 steps so a little on the one step per minute, which is not bad at all. Overall, based on the data ive collected so far, id get the step counting of the mi band 6 4 out of 5 stars. Now, as we saw last week, the mi band 6 improved quite a lot over the last year when it comes to heart rate tracking, now, im not going to show you all the results again, but i do want to summarize the main findings. First of all, we saw quite an improvement in the heart rate tracking, while cycling indoors ill, compare the heart rate measurements of the mi band 6 against the polar h10 hg chest strap, which can generally record my heart rate very accurately. Cycling indoors is one of the easiest types of exercises for a watch to track since its an exercise relatively little movement, and here we can see an overview of the heart rate accuracy for cycling indoors. Each dot here has a single heart rate measurement, with long horizontal axis, the value according to the polar h and ecg chest strap and on the vertical axis, the value. According to the mi band 6., the darker black, the color, the more dots there are in a certain area and, as you can see, theres a good agreement between the ecg chest strap and the mi band 6. As most points are along the blue line.

The correlation. This r value up here is 0.9, which is quite good. We want this r value to be as close to 1 as possible. Cycling outdoors is more tricky for watches because of the bumpiness and the increased tension on my arm. Still, though, as this is played here, the mi band 6 performed reasonably well with most points along the blue line. However, there are now also quite a few points below the blue line, indicating it sometimes detected the two low heart rate. That is because, for some cycling sessions, like the one displayed here, it was not able to keep up with the increases in my heart rate in blue. Here is my heart rate. According to the chest, strap and in red my heart rate, according to the mi band 6., as you can see, it did okay for part of the session, but it struggled for quite a bit as well. Now. There were also plenty of good sessions like this one right here, where it followed along nicely with the chest strap, but it definitely struggled during some of the rides now weight. Lifting is something that the mi band 6 really struggle with, and this is something i would definitely not recommend it for the problem is that, each time i do a set of an exercise, the tension on my wrist increases and it cannot detect my increased heart rate. You can see that in this example right here, where, for each time i do a set theres a peak in my heart rate in blue, but the mi band 6 in red cannot detect this for many more tests on the heart rate tracking check out the full Heart rate review of the mi band 6 linked up here overall id normally give this heart rate performance 3 out of 5 stars, but taking into account the low price of the mi band 6 ill bump that up to 3 and a half out of 5 stars.

Where i was quite negative about the mi band 6, when i tested it last year, it has won a small place back in my heart. It isnt one of the best health or fitness trackers out there, but for the price, its not bad at all. For about 35 dollars, you get a smart band that is pretty decent at most health tracking related functionalities, except, of course, for tracking your sleep stages. The spo2 tracking seems pretty good, and so does the step counting the heart rate tracking is okay, though not the best. Now, with the event 7 rumored to be released soon, it might be worth waiting for this. On the other hand, the mi band 6 is likely cheaper and has a more mature and developed firmware its up to you to decide what to do and, of course, ill be reviewing the mi band 7. As soon as i can get my hands on it. If you care about good heart rate tracking check out the videos i made on the huawei watcher t3 and huawei watchity runner that have amazing heart rate tracking right here, if youre looking for a smartwatch for a more premium brand check out. My reviews of the recently released garmin watches right here now. I hope this video informed you about the health and sport tracking capabilities of the mi band 6..

https://www.youtube.com/watch?v=49XAmrSzR7s