Amazfit Balance : Full Scientific Review (Best Amazfit Yet!)

The Quantified Scientist
30 Dec 202326:19

Summary

TLDRThe Amazfit Balance watch, featuring a new BioTracker 5.0 PPG sensor, shows improved heart rate tracking compared to previous Amazfit models. Reviewer Rob, a postdoctoral scientist, tests its performance during various exercises and finds it reliable for indoor activities but less so for outdoor cycling and weightlifting. Sleep tracking is mediocre, and GPS performance is decent. The watch offers good battery life and is a step up for Amazfit, but there's room for improvement.

Takeaways

  • 😀 The Amazfit Balance watch features a new BioTracker 5.0 PPG sensor, which improves heart rate measurement reliability compared to previous Amazfit models.
  • 🏃‍♂️ The heart rate tracking during indoor cycling shows good agreement with the Polar H10 ECG chest strap, with a correlation value of 0.95.
  • 🏃‍♀️ For outdoor running, the Amazfit Balance maintains a similar level of accuracy in heart rate tracking as seen with indoor cycling.
  • 🚴‍♂️ During outdoor cycling, the watch struggles more with heart rate tracking due to increased arm movement and tension, resulting in a lower correlation value of 0.65.
  • 💤 Sleep stage tracking on the Amazfit Balance is not very accurate, with significant discrepancies when compared to zMAX EEG headband data.
  • 📍 The GPS tracking of the Amazfit Balance is decent and better than expected, although not perfect, with some deviation from the route.
  • 📊 The watch's performance in heart rate tracking during weightlifting is mediocre, often failing to capture heart rate peaks accurately.
  • 🌟 The Amazfit Balance represents a significant improvement over previous Amazfit devices in terms of heart rate tracking.
  • 🔋 The watch offers good battery life, lasting about two weeks on a single charge.
  • 💬 The reviewer finds the Amazfit Balance design bulky compared to other Amazfit watches, affecting its aesthetic appeal.

Q & A

  • What is the main feature of the Amazfit Balance watch discussed in the script?

    -The Amazfit Balance watch features a new BioTracker 5.0 PPG sensor, which improves its heart rate tracking capabilities, making it the first Amazfit device tested that measures heart rate somewhat reliably.

  • How does the Amazfit Balance perform in heart rate tracking during indoor cycling?

    -The Amazfit Balance shows a good agreement with the Polar H10 ECG chest strap during indoor cycling, with a correlation of 0.95, indicating that it can track heart rate reliably with only minor deviations.

  • What is the correlation value between the Amazfit Balance and the Polar H10 during indoor cycling?

    -The correlation value between the Amazfit Balance and the Polar H10 during indoor cycling is 0.95, which is considered really good.

  • How does the Amazfit Balance compare to other watches in terms of heart rate tracking during running?

    -The Amazfit Balance has a similar performance to other devices during running, with a correlation of 0.95, indicating that it can track heart rate accurately even during outdoor activities.

  • What challenges does the Amazfit Balance face during outdoor cycling?

    -During outdoor cycling, the Amazfit Balance struggles with increased arm tension and movement, resulting in a lower correlation of 0.65 and missing several heart rate peaks.

  • How does the Amazfit Balance perform in heart rate tracking during weightlifting?

    -The Amazfit Balance performs only moderately well during weightlifting, with a correlation of 0.74, often detecting lower heart rates and missing many peaks.

  • What is the reviewer's overall rating for the Amazfit Balance's heart rate tracking performance?

    -The reviewer gives the Amazfit Balance a rating of three to maybe three and a half out of five stars for heart rate tracking performance.

  • How accurate is the Amazfit Balance in sleep stage tracking?

    -The Amazfit Balance's sleep stage tracking is not very accurate, with only about 40% of deep sleep correctly identified and a significant amount of light sleep confused with REM sleep.

  • What is the reviewer's rating for the Amazfit Balance's sleep stage tracking?

    -The reviewer rates the Amazfit Balance's sleep stage tracking at two out of five stars, indicating it's not reliable for detailed sleep analysis.

  • How does the Amazfit Balance perform in GPS tracking?

    -The Amazfit Balance's GPS tracking performance is decent, with signals being relatively consistent and close together, although not perfect.

  • What is the reviewer's overall rating for the Amazfit Balance's GPS tracking performance?

    -The reviewer gives the Amazfit Balance's GPS tracking performance a rating of about three to three and a half out of five stars.

  • What does the reviewer think about the design of the Amazfit Balance?

    -The reviewer finds the Amazfit Balance to be chunky and not as good-looking compared to other Amazfit watches, which are generally known for their nice design and screens.

Outlines

00:00

📌 Introduction to Amazfit Balance Review

The script introduces the Amazfit Balance smartwatch, highlighting its new biot tracker 5.0 PPG sensor that offers improved heart rate tracking. The reviewer, Rob, a postdoctoral scientist specializing in biological data analysis, shares his initial impressions and plans to test the watch's heart rate tracking during various exercises, sleep stage tracking, and GPS accuracy. Historically, Amazfit devices have not performed well in Rob's tests, but the Balance seems to break this trend. Despite this, Rob finds the Balance's design chunky and less appealing compared to other Amazfit models. The focus is set on objective testing rather than subjective design opinions.

05:00

🚴‍♂️ Heart Rate Tracking During Indoor Cycling

Rob tests the Amazfit Balance's heart rate tracking capabilities during indoor cycling, comparing its performance against the Polar H10 ECG chest strap, known for accurate readings. The results show a generally good agreement with a correlation of 0.95, indicating a reliable performance. However, there are occasional discrepancies, particularly when the heart rate decreases, where the Balance overestimates. Despite these minor issues, Rob is satisfied with the Balance's performance for indoor cycling, considering it a significant improvement over previous Amazfit devices.

10:01

🏃‍♂️ Heart Rate Tracking During Running

The script details Rob's testing of the Amazfit Balance's heart rate tracking during outdoor running. Despite the increased physical movement and potential for less accurate readings due to arm tension and bumpiness, the Balance maintains a strong correlation of 0.95 with the Polar H10, suggesting it does not suffer from cadence lock and can accurately track heart rate. Individual run analyses show only minor deviations, confirming the Balance's reliability during running.

15:02

🚴‍♀️ Challenges in Outdoor Cycling and Weightlifting

Rob discusses the increased difficulty in tracking heart rate during outdoor cycling and weightlifting. Outdoor cycling introduces more arm tension and movement, which typically hinders accurate heart rate readings. The Balance struggles with this, showing a lower correlation and missing several heart rate peaks. Weightlifting, another challenging exercise for heart rate tracking, also presents issues for the Balance, with many peaks missed and a correlation of 0.74, indicating less reliable performance. Comparatively, the Balance performs better for indoor cycling and running than for these more dynamic activities.

20:03

💤 Sleep Stage Tracking Evaluation

The script describes Rob's evaluation of the Amazfit Balance's sleep stage tracking capabilities, compared to a zMAX EEG headband known for measuring brain waves. The Balance shows only about 40% agreement for deep sleep, with more time incorrectly categorized as light sleep. Light sleep agreement is about two-thirds, and REM sleep agreement is roughly 53%, with significant confusion with light sleep and awake times. Individual night analyses reveal a mix of accurate and inaccurate sleep stage detections, suggesting the Balance is not reliable for detailed sleep analysis but may provide a rough estimate of total sleep time.

25:03

🚴‍♂️ GPS Tracking and Final Thoughts

Rob tests the Amazfit Balance's GPS tracking by cycling the same route multiple times and observes the signal overlap. The Balance shows some noise in signal acquisition but generally acquires the signal quickly. Throughout the routes, the GPS signals are relatively close, with occasional deviations. While not perfect, the GPS performance is better than expected. In conclusion, the Balance offers improved heart rate tracking over previous Amazfit models, especially for static exercises, but struggles with more dynamic activities and sleep tracking. Rob gives the watch a tentative rating of three out of five stars, acknowledging its的进步 but also its limitations.

Mindmap

Keywords

💡Amazfit Balance

Amazfit Balance is a smartwatch model by Amazfit, which is a brand known for its wearable devices. In the context of the video, it is the central product being reviewed. The reviewer discusses its performance in various metrics such as heart rate tracking, sleep stage tracking, and GPS accuracy, comparing it to other devices. It is noted as an improvement over previous Amazfit models.

💡Biot tracker 5.0 PPG sensor

Biot tracker 5.0 PPG sensor refers to the photoplethysmography (PPG) sensor version 5.0 used in the Amazfit Balance for tracking biometrics. PPG sensors are optical sensors that measure heart rate and other vital signs. The video mentions it as a new addition to the Amazfit device, which improves heart rate measurement reliability compared to previous models.

💡Heart rate tracking

Heart rate tracking is a feature in smartwatches and fitness devices that measures the wearer's heart rate continuously or at set intervals. In the video, the reviewer tests the Amazfit Balance's heart rate tracking during different exercises like indoor cycling and running to evaluate its accuracy and reliability.

💡Sleep stage tracking

Sleep stage tracking is the ability of a device to monitor different phases of sleep, such as deep sleep, light sleep, and REM sleep. The video script discusses the Amazfit Balance's performance in tracking sleep stages and compares it with an EEG headband, which is a device that measures brainwaves to determine sleep stages.

💡GPS tracking accuracy

GPS tracking accuracy refers to how precisely a device can track and record the user's location and movements using the Global Positioning System. The reviewer assesses the Amazfit Balance's GPS tracking by cycling the same route multiple times and observing the overlap and deviation of the recorded paths.

💡Correlation

In the context of the video, correlation measures how well the Amazfit Balance's data aligns with that of other devices or reference methods. A high correlation (close to 1.0) indicates a strong agreement between the Amazfit Balance and the device it is being compared to, suggesting reliable tracking performance.

💡Polar H10

Polar H10 is a heart rate monitor chest strap used as a reference device in the video for comparing the accuracy of the Amazfit Balance's heart rate tracking. It is noted for providing very accurate heart rate measurements, which are then compared with the data from the Amazfit Balance.

💡Outdoor cycling

Outdoor cycling is a type of exercise that the reviewer uses to test the Amazfit Balance's heart rate tracking capabilities. It is considered more challenging for watches to accurately track heart rate during outdoor cycling due to increased arm movement and tension, which can affect the quality of the PPG signal.

💡Weightlifting

Weightlifting is another exercise used to test the Amazfit Balance's heart rate tracking. It is highlighted as a difficult activity for watches to track accurately due to the tension and movement involved, which may cause the watch to miss heart rate peaks or record inaccurate data.

💡zMAX EEG headband

The zMAX EEG headband is a device that measures brainwaves and is used as a reference to test the Amazfit Balance's sleep stage tracking. The headband provides a standard for comparison, allowing the reviewer to evaluate how accurately the Amazfit Balance can detect different sleep stages.

💡Affiliate links

Affiliate links are a form of performance-based advertising where an affiliate (the reviewer, in this case) earns a commission for marketing another company's products. In the video script, the reviewer mentions that some links in the description are affiliate links, meaning they may earn a commission if a viewer purchases an Amazfit Balance or other items through those links.

Highlights

Amazfit Balance features a new BioTracker 5.0 PPG sensor for improved heart rate measurement.

The Amazfit Balance's heart rate tracking is more reliable compared to previous Amazfit models.

Heart rate tracking during indoor cycling shows good agreement with the Polar H10 ECG chest strap.

The device's heart rate readings have a high correlation of 0.95 with the Polar H10.

Amazfit Balance's design is considered chunky and less aesthetically pleasing by the reviewer.

The watch's heart rate tracking during outdoor running also shows a strong correlation of 0.95.

Outdoor cycling presents a challenge for the Amazfit Balance, with a lower heart rate correlation of 0.65.

The device struggles with heart rate tracking during weightlifting, with a correlation of 0.74.

Sleep stage tracking on the Amazfit Balance is not highly reliable, with significant inaccuracies.

GPS tracking on the Amazfit Balance is decent but not perfect, with some deviation from the route.

The reviewer rates the Amazfit Balance's heart rate tracking at three to three and a half stars out of five.

Sleep stage tracking receives a rating of two out of five stars.

GPS tracking performance is estimated to receive a rating between three to three and a half stars.

Overall, the Amazfit Balance is an improvement over previous Amazfit devices but still has room for improvement.

The reviewer personally finds the watch's design bulky and not as attractive as other Amazfit models.

The Amazfit Balance is priced at around $200, offering good battery life but with a mixed performance review.

Transcripts

play00:00

amazfit finally made a watch that

play00:02

doesn't totally suck this is the amazfit

play00:04

balance which has the new biot tracker

play00:06

5.0 PPG sensor which makes this the

play00:09

first amazfit device I tested that

play00:11

actually measures my heart rate at least

play00:13

somewhat reliably it definitely isn't

play00:15

perfect but it might just be good enough

play00:17

in this video we're going to test the

play00:19

hard rate tracking performance of the

play00:20

amaz Fit balance during different types

play00:22

of exercises and we'll additionally test

play00:24

the Sleep stage tracking performance and

play00:26

the GPS tracking accuracy of this new

play00:28

amazfit device as always I don't want to

play00:31

waste your time so there are time stamps

play00:33

on the timeline let's get

play00:38

started hello everyone for those of you

play00:41

that are new to the channel my name is

play00:42

Rob and I'm a postdoctoral scientist

play00:44

specializing in biological data analysis

play00:47

now amazfit devices have historically

play00:49

performed terribly in my testing and

play00:51

like I said the amazfit balance is the

play00:54

first watch to break this trend sadly

play00:56

the one thing that a maze fit generally

play00:58

does do very well does not flag back in

play01:00

this watch namely the nice design of

play01:02

most of their watches amazfit devices

play01:04

generally have quite beautiful screens

play01:06

and a good design but I personally find

play01:08

the am F balance quite chunky and not so

play01:11

goodlooking however that's a subjective

play01:13

opinion and that's not what you're here

play01:14

for let's focus on some objective

play01:16

testing instead and let's start with the

play01:18

heart rate tracking performance I'll

play01:21

start off by looking at the performance

play01:22

of the Amit balance during indoor

play01:24

cycling which is a relatively easy

play01:26

exercise for watch the track given the

play01:28

limited movement and of tension on my

play01:30

arm and here you can see an overview of

play01:32

those results now to test the

play01:34

performance I'll compare the heart rate

play01:36

measurements of the amaz with balance

play01:38

against the polar h10 ECG Jaz strap

play01:40

which can generally record my heart rate

play01:42

very accurately now each dot in this

play01:45

plot is a single heart rate measurement

play01:47

with along the horizontal axis the value

play01:49

according to the polar h10 ECG chest

play01:51

strap and on the vertical axis the value

play01:53

according to the amaz Fit balance now

play01:55

the closer the points are to this blue

play01:58

line right here the better the agreement

play01:59

and and the darker black the color mod

play02:01

that there are and as you can see

play02:03

there's generally a pretty good

play02:04

agreement between the polar h10 and the

play02:06

amaz Fit balance most points are on or

play02:08

at least close to the blue line and the

play02:11

correlation which is this r value up

play02:13

here is also quite high at

play02:15

0.95 the maximum value of this

play02:18

correlation is 1.0 so a correlation of

play02:21

0.95 is really good there are a few

play02:24

points away from the Blue Line

play02:26

especially below it right here but still

play02:29

this doesn't look too bad but we really

play02:31

need to look at those individual bike

play02:32

rides to see what's going on and here

play02:35

you can see that first example interval

play02:36

spinning session where we see a pretty

play02:39

decent agreement between the amaz Fit

play02:40

balance and the ECG chest strap now

play02:42

along the horizontal axis we have the

play02:44

time and my heart rate is on the

play02:46

vertical axis with in blue green my

play02:48

heart rate according to the polar h10

play02:50

ECG chest strap and in red my heart rate

play02:52

according to the amaz with balance and

play02:54

for this ride most of the time the heart

play02:56

rate was correctly detected only right

play02:58

here where I had a DE decrease in my

play03:00

heart rate did the M balance detect an

play03:02

increase in my heart rate so detected

play03:04

about double my heart rate right here so

play03:06

it detected a heart rate of about 140

play03:09

BPM I would say whereas in reality my

play03:12

heart was about 70 BPM so it probably

play03:14

had a bit of a noisy signal but this is

play03:16

the exception rather than the rule

play03:18

looking at my next spinning session we

play03:19

see this looks quite good most of the

play03:22

time my heart R agrees between the polar

play03:23

h10 and the amaz with balance and that's

play03:26

also what we see for this bike ride

play03:28

right here though again there's one art

play03:29

effect right here where it all of a

play03:31

sudden detected an increase in my heart

play03:32

rate but as I said this is the exception

play03:34

rather than the rule because for most

play03:36

indoor cycling sessions it looks rather

play03:38

good like for instance right here and

play03:40

also right here there is some minor

play03:42

deviation for instance right here a

play03:43

slight delay in it picking up an

play03:45

increase in my heart rate but overall

play03:47

I'm quite happy with the performance of

play03:49

the mid balance during indoor cycling so

play03:52

that's not looking too bad there are

play03:53

some small artifacts but nothing much to

play03:56

complain about compared to other older

play03:58

MV devices is this is actually great

play04:00

performance but let's take a look at how

play04:02

this watch Compares and we will do that

play04:04

by comparing the performance of the M

play04:06

with balance to many of the other

play04:07

devices I've tested in the past and that

play04:09

overie is displayed right here the

play04:11

correlation value I was talking about

play04:13

before is the metric we will use for

play04:14

this which is displayed along the

play04:16

horizontal axis here we want that value

play04:18

to be as close to one as possible on the

play04:21

vertical axis actually order to watch us

play04:23

from worst to best so the further to the

play04:25

right and the higher devices the better

play04:28

is its correlation with the ref device

play04:30

and I marked the Amit balance in red

play04:32

right here and as you can see compared

play04:34

to all other watches out there it's in

play04:36

the upper middle class of watches it

play04:39

isn't amazing as the Apple watches up

play04:42

here are for instance but it's

play04:43

definitely decent and probably good

play04:45

enough what is important to know is that

play04:47

many of the other amazfit devices are on

play04:49

the bottom right here and right here so

play04:51

out of the bottom eight watches right

play04:53

here one two three four five six were

play04:58

amazed with devices so this is really a

play05:00

significant Improvement all the way up

play05:02

here but let's zoom in a bit so we can

play05:04

actually read those labels and that

play05:05

zoomed infus is played right here so

play05:07

these are just the watches with a

play05:09

correlation of 0.9 or higher and as you

play05:11

can see even amongst these better

play05:13

watches the maze Fit balance is doing

play05:14

quite decently it's quite close to the

play05:17

Fitbit Lux the Huawei watch 3 and the

play05:20

polar Varity sense for instance so this

play05:22

doesn't look too bad there are

play05:24

definitely better devices out there

play05:26

especially Apple watches some Huawei

play05:28

watches but also Google pixel watches

play05:30

and Samsung Galaxy watches aren't bad at

play05:32

all but I also suspect that this

play05:34

performance of the amaz with balance is

play05:35

good enough for most people so that's

play05:37

not too bad at all but we haven't really

play05:39

challenged the amaz with balance so

play05:40

let's do that let's now take a look at

play05:42

the next most difficult exercise for

play05:44

watch the track running outside now

play05:46

running outside increases the bumpiness

play05:48

during the exercise however there's

play05:50

still little tension on my arm and on my

play05:51

wrist so this is only slightly harder

play05:53

for the watch you get a clean heart rate

play05:55

signal let's see how the amaz with

play05:56

balance did and you can see that overiew

play05:58

right here but now for running and as

play06:00

you can see the agreement is still very

play06:02

good the correlation is very similar at

play06:05

0.95 and all values are on or close to

play06:08

the blue line or most measurements at

play06:10

least there are a few below it for

play06:12

instance right here and also right here

play06:14

and a few above it but the deviation is

play06:16

quite minor so this looks good for the

play06:18

ma with balance it doesn't seem to

play06:20

suffer from Cadence lock where it meses

play06:22

your Cadence instead of your heart rate

play06:24

but let's take a look at the individual

play06:25

runs to see where there's still some

play06:27

points below and above the blue line and

play06:29

here you can see the results for my

play06:31

first run with again the polar h10 in

play06:33

blue and the amaz Fit balance in red and

play06:35

there's generally a good agreement

play06:36

between the amaz set balance and the

play06:38

polar h10 there are some minor deviation

play06:41

for instance right here and also right

play06:42

here but generally this looks very good

play06:45

and that's also we see for the second

play06:46

run right here though in the beginning

play06:48

it missed my heart rate for a while

play06:49

before it caught on but after it

play06:51

acquired the heart rate signal it

play06:52

actually looks really good and that's

play06:54

also what we more or less see for this

play06:57

example run right here though here it

play06:58

did struggle a bit in the beginning

play07:00

right here but later on it looks really

play07:02

good and that's also true for this final

play07:04

example I wanted to show you here it

play07:05

struggled a bit near the end but overall

play07:08

not a bad performance so here we have a

play07:10

similar overie to before but now for

play07:12

running so again the further to the top

play07:13

right the better device and as you can

play07:16

see there has far fewer devices in this

play07:17

overiew because I only recently started

play07:20

running so I don't have data for that

play07:21

many watches but as you can see the amaz

play07:24

with balance is actually doing quite

play07:26

good again Apple watches are some of the

play07:28

top performers out there but the amaz it

play07:30

balance isn't half bad I would say it's

play07:32

good enough for most people it's really

play07:34

close to for instance some Huawei

play07:36

watches and also some pixel watches so

play07:39

I'm not disappointed by this so we saw

play07:41

some small artifacts again but the

play07:43

correlation is still quite high I would

play07:44

say at roughly 0.95 however let's now

play07:47

make things even more difficult by

play07:49

looking at the performance of the amazit

play07:50

balance for outdoor cycling now cycling

play07:53

outside increases the tension on my arms

play07:55

because I have to all on to the

play07:56

handlebars there's also much more

play07:58

movement and bumpiness making it much

play08:00

harder for a watch to get a clean heart

play08:01

rate signal I tested the MH balance for

play08:04

a total of 21 bike rides I have for

play08:06

first a quick S note if you're

play08:08

interested in the latest updates on the

play08:09

wearables I'm testing I'm planning to

play08:11

start back up with my newsletter and

play08:12

posting more Optical things on my

play08:14

Instagram and my YouTube shorts channel

play08:16

so if you're interested in any of those

play08:17

those are linked below also I'm trying

play08:19

to become part of the first people who

play08:21

receive watches to review from companies

play08:23

like Garmin Samsung and mafit and if you

play08:26

want to help me make faster and better

play08:27

reviews it would really help if you like

play08:30

subscribe and or comment but of course

play08:32

this is totally up to you now enough

play08:34

self- promotion let's get back to

play08:35

testing the amaz with balance and that

play08:38

overview is displayed right here so this

play08:39

is a similar overview to before but now

play08:41

we're biking outside and as you can see

play08:43

this is where the amaz with balance

play08:45

starts a struggle there are still some

play08:47

points close to Blue Line or at least

play08:48

the majority is still close to the blue

play08:50

line but there are many points below the

play08:53

blue line right here so there's this big

play08:55

triangle right here of points below the

play08:57

Blue Line indicating that in these

play08:59

moments the watch detected a two low

play09:01

heart rate the correlation is also much

play09:03

lower at 0.65 where before the

play09:06

correlation was 0.95 for both running

play09:08

and cycling inside cycling Outdoors

play09:11

really makes the watch struggle but

play09:13

let's take a look at the individual bike

play09:14

rides to see what's going on here and

play09:16

here we have the first example bike

play09:18

right with again the reference in blue

play09:19

green and the mafit balance in red and

play09:22

as you can see the agreement is not

play09:24

amazing there's generally quite a delay

play09:26

in it picking up an increase in my heart

play09:28

rate so it does seem to catch up over

play09:30

time but it really needs a bit more time

play09:32

to get those peaks in my heart rate you

play09:34

can see that right here but also right

play09:36

here right here and right here so each

play09:38

time I have a sudden increase in my

play09:40

heart rate it really struggles to keep

play09:41

up that's what we see to a higher or

play09:43

lower degree in many of the bike rides

play09:45

so for this bike ride right here it

play09:47

didn't struggle as much during the

play09:48

second part but during the first part it

play09:50

had some issues for this bike ride it

play09:52

actually looks quite good so there's not

play09:54

much to complain about also this one

play09:56

doesn't look that bad then for this bike

play09:59

right where it did detect my heart rate

play10:00

it was quite good though it missed my

play10:02

heart rate for a large part of the ride

play10:03

so somehow it lost the signal right here

play10:06

and then for other bike rides we again

play10:08

see a delay in it picking up an increase

play10:10

in my heart rate you can see that

play10:11

especially near the end of this bike

play10:13

ride right here and also for this bike

play10:15

ride it either had the delay or it

play10:17

completely missed a peak in my heart

play10:19

rate as you can see right here so it's a

play10:21

bit of a mixbag some bike rides look

play10:24

okayish I would say this one looks

play10:26

okayish though it still has its issues

play10:28

the other are quite a bit worse where

play10:30

for instance right here it misses a

play10:31

large part of the peak right here so be

play10:33

aware that at least on medium as balance

play10:35

is struggling a bit and again this is a

play10:37

similar overview to before but now for

play10:39

biking outside and as you can see the

play10:42

amazfit balance marked in red isn't

play10:44

amazing it's not the worst device out

play10:47

there which again are other amazit

play10:49

devices so the top five worst performing

play10:52

watches here are all amazit devices so

play10:54

for instance the amaz with bipu Pro but

play10:57

also the amazfit T-Rex Pro and GTS S2

play10:59

mini and also some newer ones like the

play11:01

GTR 4 aren't doing great amazfit devices

play11:04

are really some of the worst devices out

play11:06

there at least up until the Amit balance

play11:09

which is doing quite okay for most

play11:11

exercises but it's not amazing for

play11:13

biking outside as you can see right here

play11:16

most devices are doing better than the

play11:17

amaz with balance for biking outside so

play11:19

I wouldn't really recommend it for that

play11:21

type of exercise and again Huawei

play11:23

watches Google pixel watches and

play11:25

especially Apple watches are doing a lot

play11:27

better so while biking out outside the

play11:29

amaz with balance didn't perform as well

play11:31

as it did for instance for indoor

play11:32

cycling and running it missed a

play11:34

significant number of peaks in heart

play11:36

rate and also has some delay in picking

play11:38

up on changes in my heart rate so if you

play11:40

want to use it for outdoor cycling I

play11:42

wouldn't recommend it now before looking

play11:44

at the Sleep stage tracking test let's

play11:46

take a look at the performance of the

play11:47

amaz with balance while weightlifting

play11:49

this is generally also a very hard

play11:51

exercise for many watches to track and

play11:53

here we have that overview similar to

play11:54

before but now for weightlifting and as

play11:57

you can see it looks okay okay it's not

play11:59

as bad as what we saw for outdoor

play12:01

cycling so the correlation at 0.74 is a

play12:04

bit higher than what we saw for outdoor

play12:06

cycling which I believe was 0.64 0.65

play12:10

but it's definitely not as good as what

play12:12

we saw for indoor cycling and running

play12:14

which has a correlation of

play12:15

0.95 and that's probably due to it

play12:18

sometimes detecting a to low heart rate

play12:20

especially here in the higher heart rate

play12:22

range it almost always attacks it to low

play12:25

heart rate this indicates to me that

play12:27

most likely the individual sets that I

play12:30

do it cannot pick up on those peaks in

play12:31

my heart rate but let's take a look at

play12:33

the individual weightlifting sessions to

play12:35

see what's going on here so here we have

play12:37

the first example weightlifting session

play12:38

I wanted to show you with again the

play12:40

reference in blue green and you may Fit

play12:42

balance in red with the peaks in my

play12:44

heart who representing each set that I

play12:45

did and as you can see the peaks in blue

play12:48

aren't completely followed by the

play12:50

amazfit balance in red some of the Peaks

play12:52

are somewhat detected for instance right

play12:54

here and right here but many of them are

play12:57

missed and this is what we see for many

play12:58

of the weightlifting sessions some of

play13:00

the Peaks are caught but many or even

play13:03

most are missed and I would say the amaz

play13:05

with balance isn't that reliable for

play13:07

weightlifting and again here we have a

play13:09

similar overview to before but now for

play13:11

weightlifting and here the amaz FD

play13:14

balance is somewhere in the middle I

play13:16

would say you can see it marked here in

play13:18

red it's not amazing and also not

play13:20

terrible though generally I wouldn't

play13:23

recommend using any device with a

play13:24

correlation of 0.9 or preferably even

play13:27

0.95 or high for weightlifting you

play13:30

really want decent performance and you

play13:32

also want it to pick up on those peaks

play13:33

in heart rate and the amaz with balance

play13:35

just doesn't do it but that's true for

play13:37

most watches out there it's just really

play13:39

difficult to track your hard during

play13:40

weightlifting if we zoom into the best

play13:42

performing watches we can see that only

play13:44

some selected Huawei watches Apple

play13:46

watches and also some Google products

play13:48

like the pixel Watch 2 and the Fibber

play13:49

charge 6 are decent during weightlifting

play13:52

other watches I would really not

play13:54

recommend and if you want accurate heart

play13:56

rate tracking during weightlifting my

play13:57

general recommendation is to use an ECG

play13:59

chest strap like for instance the polar

play14:01

h10 this doesn't suffer from that

play14:02

tension on your wrist in your arm so

play14:04

it's just much more reliable but if you

play14:06

don't find it comfortable one of these

play14:07

devices is probably your best bet so for

play14:10

weightlifting the amaz with balances Med

play14:12

the Yer I would say it isn't as bad as

play14:15

some of the other devices out there but

play14:16

it definitely isn't good it misses many

play14:18

of my peaks in heart rate so I wouldn't

play14:20

rely on it for weightlifting still

play14:22

overall it showed a significant

play14:24

Improvement compared to basically all

play14:26

other amaz devices out there and it did

play14:28

quite well for two easier exercises

play14:30

cycling indoors and running but it

play14:32

struggled during weightlifting and

play14:33

outdoor cycling therefore overall i'

play14:36

give the hard rate tracking of the amaz

play14:38

balance three to maybe even three and a

play14:40

half out of five stars next let's take a

play14:43

look at the Sleep stage tracking of the

play14:44

amaz F balance which I tested during a

play14:46

total of seven nights now to test the

play14:48

Sleep stage tracking performance I'll

play14:50

compare the amaz F balance to the zMAX

play14:52

EG headband which can actually measure

play14:54

my brain waves now this device also has

play14:56

its limitations especially when it comes

play14:58

to to detecting awake time I discussed

play15:00

this in several of my videos so I won't

play15:02

go into details here but what it boils

play15:03

down to is it'll mostly ignore the awake

play15:06

time detection in the analysis and here

play15:08

I show an overview of the sleep test

play15:10

results now on top are the sleep stages

play15:12

as recorded by the zMAX eug device and

play15:14

on the left are the sleep stages as

play15:16

recorded by the amaz with balance now

play15:18

each column here sums to 100% meaning

play15:21

that we can see what percentage of each

play15:23

of the sleep stages according to the

play15:24

zMAX was predicted as each sleep stage

play15:26

body M with balance and if they

play15:28

perfectly agree all values on the

play15:30

diagonal should be 100% now first of all

play15:33

we see that only about 40% of what was

play15:35

deep sleep according to the EEG device

play15:37

was also deep sleep according to the

play15:39

amfit balance so that doesn't look very

play15:40

good more was actually predicted as

play15:42

being light sleep instead now light

play15:45

sleep agreement is okayish I would say

play15:47

about 2third of what was light sleep

play15:49

according to the EG device was also

play15:51

detected as light sleep by the am F

play15:53

balance however a significant amount is

play15:55

actually being confused as RAM sleep in

play15:57

sted about 25% now Ram sleep agreement

play16:00

also isn't that great though not as bad

play16:02

as I initially feared at about 53% and

play16:04

most confusion was with light sleep at

play16:06

about 27% but a lot of it was also being

play16:09

confused as awake time instead now as I

play16:12

said before we're not going to focus on

play16:13

the awake time detection but just for

play16:16

completeness only about onethird of what

play16:18

the zMAX detected as a wake time was

play16:20

also detected as a wake time by the

play16:22

amazit balance but the zMAX actually

play16:24

tends to detect a lot of short awake

play16:25

moments that I'm not sure are really

play16:27

relevant so we're not going to focus on

play16:28

this too much luckily most confusion was

play16:30

with light sleep which makes sense

play16:32

actually as light sleep is the closest

play16:34

Sleep stage to being awake but again

play16:36

we're not going to focus on the awake

play16:37

moments we're mostly going to focus on

play16:39

this part right here so with the Deep

play16:40

Sleep light sleep and R sleep and for

play16:42

the amaz Fit balance it's actually

play16:44

really interesting to look at the

play16:45

individual Knights to help us better

play16:47

understand what's going on here and here

play16:48

we have the first Knight I wanted to

play16:50

share with you on top we have the sleep

play16:52

stages according to the zMAX EG headband

play16:55

with along the horizontal AIS the clock

play16:57

time and my sleep St stages on the

play16:58

vertical axis and on the bottom we have

play17:01

a similar plot but now for the amid

play17:03

balance now in purple right here I

play17:05

highlight the deep sleep as recorded by

play17:07

the zMAX EG headband and as you can see

play17:10

there's a marginal agreement between the

play17:12

zMAX and the amid balance some of the

play17:15

deep sleep that the zMAX detected was

play17:17

also detected by the amaz with balance

play17:18

about half of it which matches that 50%

play17:21

that we saw before but the amaz with

play17:23

balance also detects some extra deep

play17:25

sleep and for this second night we see

play17:27

more or the same thing some of the Deep

play17:30

Sleep detected by the zMAX is also

play17:32

detected by the amid balance again

play17:34

roughly half I would say but also some

play17:37

extra deep sleep is detected and that's

play17:39

more or less the general pattern that we

play17:41

see though there are some better kns

play17:43

like this one right here for instance

play17:44

now as I mentioned before Ram sleep

play17:46

agreement is okay though not amazing you

play17:49

can see that right here for instance

play17:51

where some of the ram sleep segments

play17:52

detected by the EG device were also

play17:55

detected by the amid balance for

play17:56

instance right here right here and right

play17:58

here but the am balance also detects

play18:00

some extra Ram sleep for instance right

play18:02

here and right here and it also misses

play18:03

some Ram sleep that's more or less the

play18:05

general Trend that we see though some

play18:07

Knights are better like this one right

play18:09

here and some are not as great like this

play18:11

one right here so it's really a bit of a

play18:13

mix so that doesn't look terrible but

play18:15

also not great so I wouldn't really rely

play18:17

on it as I said I'm not going to focus

play18:20

on the awake moments marked in green

play18:21

here according to the EEG device because

play18:23

there are just so many of these short

play18:25

awake moments that the EEG device

play18:26

detects that I'm not sure that

play18:27

interesting but we can generally see

play18:29

that the amaz balance roughly detects

play18:31

when you fall asleep and when you wake

play18:33

up we see that for most nights also

play18:35

right here it doesn't look too bad for

play18:37

my falling asleep and waking up moment

play18:39

though I had a bit of a slumbering face

play18:41

right here where the amaz with balance

play18:42

detecting me as being awake all of the

play18:44

time still overall it looks pretty okay

play18:47

and we see something similar right here

play18:49

where again I had a slumbering face and

play18:51

the amaz with balance just decks me as

play18:53

being awake but still generally for

play18:55

getting an estimation of your total time

play18:57

of sleep this doesn't look too bad okay

play18:59

that's honestly not very encouraging it

play19:01

doesn't seem like a totally random

play19:03

result but it's definitely not good and

play19:05

there are plenty of better devices out

play19:07

there believe me and to prove that to

play19:09

you let's see how the am withit balance

play19:10

performs compared to other watches that

play19:12

will help us put these results into

play19:14

perspective now this graph right here

play19:15

shows you an overview of the agreement

play19:17

of different watches with different

play19:18

reference devices along the horizontal

play19:21

axis we have the average agreement over

play19:22

the individual sleep stages and on the

play19:25

vertical axis we have the agreement of

play19:26

the worst sleep stage now a better de

play19:28

agreement the more to top right to

play19:30

devices and this overview is slightly

play19:33

complicated because I use different

play19:34

reference devices so the devices not

play19:36

marked in any color were tested against

play19:38

the dream 2 EG had been my general

play19:40

reference but dream went bankrupt so I

play19:42

cannot use it anymore the devices marked

play19:44

in blue purple were tested against

play19:46

polography so the gold standed in sleep

play19:49

stage tracking and the devices marked in

play19:51

green were tested against the zMAX EG

play19:53

headband so the same device we're using

play19:55

in this test today and you can see the

play19:57

amaz fit balance marked in green right

play19:59

here so it's really amongst the lower

play20:02

performing watches I would say it's not

play20:05

as bad as for instance the Mi band but

play20:07

it's also not nearly as good as many of

play20:09

the other devices out there especially

play20:12

Apple watches are very good sleep stage

play20:14

trackers but also the eight Sleep p 3

play20:16

aura rings the pixel watch and even the

play20:19

whoop strap are pretty decent sleep

play20:21

stage trackers and the amazit balance

play20:23

just doesn't come close so I really

play20:25

recommend just using the amazit balance

play20:27

for tracking your total in bed maybe

play20:29

your total sleep time but not more than

play20:31

that so the Sleep stage tracking

play20:32

performance of the amaz with balance is

play20:35

basically what I expected it to be like

play20:37

it's not something I would rely on I

play20:39

think it can be used to get an estimate

play20:41

of your total time spent in bed but

play20:43

that's about it I wouldn't use it for

play20:44

anything more detailed therefore I'd

play20:46

give the Sleep stage tracking

play20:48

performance of the amaz with balance two

play20:50

out of five stars next we're going to

play20:52

take a look at the GPS tracking

play20:54

performance of the amaz with balance I

play20:56

tested the GPS tracking performance by

play20:58

by cycling the same route several times

play21:00

and seeing how well the signals overlap

play21:02

let's take a look and here you can see

play21:03

the results for six times I cycled back

play21:06

from work now I always started my ride

play21:09

either right here or right here and I

play21:11

didn't give the watch any extra time to

play21:13

acquire the signal and as you can see

play21:15

there is some noise in the moment it

play21:17

acquired the signal so this ride right

play21:19

here for instance I probably started all

play21:21

the way down here and it took some time

play21:23

for it to acquire the signal and if we

play21:25

zoom in a bit to these four rides right

play21:27

here all of them are supposed to start

play21:29

roughly in front of this building right

play21:31

here and you can see definitely there's

play21:32

a bit of noise in when it acquired a

play21:34

signal one time the signals all the way

play21:36

down here and I'm not sure what it is

play21:38

I'm not sure if this is a very late

play21:39

acquisition and I actually started it

play21:41

all the way back there where also this

play21:42

one was started or if it was actually

play21:44

started down here and it's noisy still

play21:47

the signal acquisition isn't perfect but

play21:49

it definitely doesn't take a lot of time

play21:51

for it to acquire the signal so might be

play21:53

good enough for most people you might

play21:54

just have to wait a little bit for it to

play21:56

get the signal and if we then look down

play21:58

the route we can see it's a bit noisy

play22:00

though not terrible there's one signal

play22:03

which is very deviant right here others

play22:04

are actually relatively close to each

play22:06

other and this one keeps deviating quite

play22:08

a bit also one more signal is deviating

play22:11

a bit right here and if you go down the

play22:13

route there's always one or two signals

play22:16

that are deviating a bit and the others

play22:18

are relatively close it's not terrible

play22:20

actually many of the signals here are

play22:22

super close and this looks better than

play22:23

some other watches let's zoom out a

play22:26

little bit if we go down the rout we can

play22:28

see the signals are actually not that

play22:30

far away from each other it actually

play22:32

looks quite decent here I took different

play22:34

routes you can see one time I went along

play22:36

here and one time along here but the

play22:38

signals stay quite close together

play22:40

usually here near the radio tower

play22:42

there's a bit more deviation and we can

play22:43

also see that right here there might

play22:45

actually be some signals coming from

play22:47

right here that are interfering still

play22:48

overall this doesn't look too bad

play22:50

honestly the signals are relatively

play22:53

close together and there are many

play22:54

watches that are doing worse there are

play22:56

some better devices out there like for

play22:58

instance the Apple watch and especially

play22:59

Garmin watches but this doesn't look

play23:01

that bad to me but let's take a look at

play23:03

some rides to work to see what these

play23:05

look like and these are displayed right

play23:07

here I only have three rides for this so

play23:09

these are three times I cycled to work

play23:12

here we can see that the signals always

play23:14

acquired very quickly and very

play23:15

consistently I always started it on the

play23:17

corner right here so there's a bit of

play23:19

deviation right here but this doesn't

play23:21

look too bad and if we zoom out a bit we

play23:23

can actually see that the signals are

play23:25

relatively consistent now we're only

play23:26

looking at three signals right here so

play23:29

there isn't as big of a chance of a

play23:30

deviating signal but generally if we go

play23:33

along the Route it actually looks quite

play23:35

decent it's not amazing there is some

play23:37

deviation also right here and right here

play23:39

but it generally doesn't look too bad in

play23:41

my opinion if we go along the Route even

play23:44

further there's generally a relatively

play23:46

small deviation between the signals but

play23:48

you can judge for yourself compared to

play23:50

some of the other watches I've tested in

play23:52

the past where I take a similar route

play23:54

overall I'm actually not that

play23:55

disappointed yes there is some deviation

play23:58

it goes a bit away from the road but

play24:01

overall this is relatively consistent

play24:03

and for a device like this I actually

play24:05

expected worse performance so overall

play24:08

not bad so the GPS tracking performance

play24:10

of the amaz Fit balance isn't half bad

play24:13

it isn't amazing but it's actually

play24:15

better than I expected I always find it

play24:17

difficult to put an actual number on it

play24:18

without actually quantifying it but it

play24:20

probably give the GPS tracking

play24:22

performance around three to maybe even

play24:24

three and a half out of five stars

play24:26

before getting to my final conclusions

play24:28

of course there's many limitations to my

play24:30

testing the main one being that I just

play24:32

tested it on myself with whatever

play24:34

firmware was available at the time now I

play24:36

tried to summarize my final judgment on

play24:38

the amazfit balance and I find myself

play24:40

struggling somewhat I saw you can now

play24:42

get it for about $200 which isn't

play24:44

terribly high it's not cheap but it's

play24:46

also not the most expensive watch out

play24:48

there it also has a good battery life at

play24:50

about 2 weeks and the watch is

play24:52

significantly better than anything amaza

play24:54

that's produced so far however it still

play24:56

isn't that great it has quite good heart

play24:59

rate tracking for more static exercises

play25:01

but it really struggled for cycling

play25:02

outside and weightlifting the Sleep

play25:04

stage tracking also wasn't that great

play25:06

the GPS tracking performance on the

play25:08

other hand wasn't too bad also and I

play25:10

know this is more subjective I really

play25:12

don't like the way it looks a mafit

play25:14

generally does such a good job making

play25:16

watches with a really nice screen and a

play25:17

relatively slim design and this thing is

play25:20

just a bit too bulky for me but of

play25:21

course that last part is very subjective

play25:24

all in all I'd probably give this watch

play25:26

three out of five star bars mafit has

play25:29

definitely improved their heart rate

play25:30

tracking which is good but they've got a

play25:32

bit to go before I would strongly

play25:33

recommend any of their watches of course

play25:35

also check out some other reviewers if

play25:37

they reach different conclusions if you

play25:39

do decide to get an amazfit balance an h

play25:41

s pot 3 a whoop strap an order ring

play25:44

another device or anything at else on

play25:46

Amazon for that matter even something as

play25:48

small as toilet paper want to

play25:49

potentially save some money and at the

play25:51

same time support the channel their

play25:53

different affiliate and non-affiliate

play25:54

links in description below they do not

play25:55

cost you any extra and some even

play25:58

discount also given that you made it to

play25:59

the end of this video on the amazfit

play26:01

balance check out my other reviews of

play26:03

amazfit devices as well and also my top

play26:05

General recommendations in the meantime

play26:08

thank you so much for watching and catch

play26:10

you in the next

play26:13

[Music]

play26:18

video

Rate This

5.0 / 5 (0 votes)

Ähnliche Tags
Amazfit BalanceHealth TrackingWatch ReviewHeart RateSleep StageGPS AccuracyTech GadgetsWorkout TechProduct TestWearable Tech
Benötigen Sie eine Zusammenfassung auf Englisch?