10-24-2017 04:31
10-24-2017 04:31
I just got myself an Ionic!
However, from the moment I started wearing it I felt like the heart rate wasn't quite right. When checking my wrist, the Ionic was about 8 beats per minute ahead of what I counted myself. To be sure I wore both my old charge 2 (on my right wrist) and the Ionic (on my left wrist) for a day.
Here's a screenshot from me waking up that morning, Ionic in blue, Charge 2 in red. Every horizontal line is 5 beats per minute.
Now, what's going on here? How can these two trackers have such different results? I've been wearing them both according to the through fitbit recommended instructions.
Sleep also isn't entirely similar (note that they aren't the same length so are a bit skewed at the end):
So, well, who's lying here? And how can I be sure that I've got some thrustworthy data to work with? I really looked forward going to the Ionic from the Charge 2 but now I'm doubting both.
Any info will be appreciated,
Kind regards,
David
10-26-2017 13:57
10-26-2017 13:57
@Davidbit wrote:So, well, who's lying here? And how can I be sure that I've got some thrustworthy data to work with? I really looked forward going to the Ionic from the Charge 2 but now I'm doubting both.
@Davidbit Your guess is as good as the trackers! As a healthy middle aged guy I haven't found a real reason to track 24x7 HR or sleep. In particular with sleep tracking, its not possible to verify the estimated sleep states (unless its outright wrong, or you have access to a sleep lab). If you are like most busy people with a job and family, you sleep as much as you can fit into a busy day, so having sleep data (assuming its accurate) isn't something I can do anything about. Others feel differently, or are just fascinated looking at the pretty graphs. I think sleep data is a "feel good" stat, makes you feel like you know something about yourself whether true or not. They are fun to look at from time to time.
The Fitbit HRM is generally good at tracking HR during the day, when you are sitting on the couch or at your work desk. Was a bit surprised at the differences in your 1 hour 50 minute timeline.
Nice work on merging the sleep graphs, great visual!!
Not sure you will get more definitive answers. Enjoy your Ionic!
Aria, Fitbit MobileTrack on iOS. Previous: Flex, Force, Surge, Blaze
11-01-2017 03:16
11-01-2017 03:16
Thank you for your reply! I have to admit I am one of the people taking it a bit too seriously, especially for the accuracy of these devices.
That said, I'm still surprised by the inaccuracy of, well, either of these devices.
This is taken while I was working at my computer at work, again with the vertical lines being 5 bpm each. Again the Ionic (in blue) is about 10 bpm higher than the charge 2 (in red), except for one bit.
And then there is this, same day, same computer.
I've ordered a Polar H7 to for once and for all find out who's lying to me here, but I have the feeling I'm probably going to give up on the Ionic.
I'll keep you up to date!
11-01-2017 07:04
11-01-2017 07:04
Upfront let me clarify that I am not an apologist for Fitbit. I am only trying to make sense of the graphs presented here constructively. If I examine the graph very closely it is apparent that sometimes Charge had a higher reading while ionic had a lower reading. I am wondering we are comparing apples to Oranges by wearing the watches in left and right hands. Clearly every one has a dominant hand. It is also true that limb movements will interfere with HRM reading. Why not wear them both at the same hand, two finger width from the wrist.
Perhaps I should do something similar too.
With reference to Polar, I have a Polar H7, H10 and myzone. I have checked instantaneous heart rate at rest and extremely low intensity ambulatory conditions; the heart rates do appear to match. However it never matches once intensity begins to increase whcih is understandable ;I don't have a problem with it and hence I don't worry about it. I have never compared the devices for a longer duration as I assumed it will be more or less same.
How did you produce this graph so I may borrow the idea? I assume you exported the data to spreadsheet.?
11-01-2017 07:17
11-01-2017 07:17
If the wrist you wear it on makes a 10 bpm difference I think there's a significant flaw somewhere either way, but I will give it a try.
I used both trackers with different phones and different accounts and exported the data of a single day of both through https://www.squashleagues.org/fitbit/fitbitdatadownload. Then I imported the data into two scatter plots in excel, formatted the axis to ensure they'd match, made one transparent, and just layed it on top of the other.
I'm curious to see any data you come up with, please let me know. 🙂
11-01-2017 08:10
11-01-2017 08:10
@bbarrera wrote:
@Davidbit wrote:So, well, who's lying here? And how can I be sure that I've got some thrustworthy data to work with? I really looked forward going to the Ionic from the Charge 2 but now I'm doubting both.
@Davidbit Your guess is as good as the trackers! As a healthy middle aged guy I haven't found a real reason to track 24x7 HR or sleep.
Couldn't we apply that statement to steps, HR, as well?
11-01-2017 08:24
11-01-2017 08:24
@bobo03 wrote:
@bbarrera wrote:
@Davidbit wrote:So, well, who's lying here? And how can I be sure that I've got some thrustworthy data to work with? I really looked forward going to the Ionic from the Charge 2 but now I'm doubting both.
@Davidbit Your guess is as good as the trackers! As a healthy middle aged guy I haven't found a real reason to track 24x7 HR or sleep.
Couldn't we apply that statement to steps, HR, as well?
Why would a healthy middle aged guy want a ferrari instead of a ford? We can apply that statement to everything in life, but that's besides the point of this topic.
11-01-2017 08:26
11-01-2017 08:26
@Davidbit wrote:
@bobo03 wrote:
@bbarrera wrote:
@Davidbit wrote:So, well, who's lying here? And how can I be sure that I've got some thrustworthy data to work with? I really looked forward going to the Ionic from the Charge 2 but now I'm doubting both.
@Davidbit Your guess is as good as the trackers! As a healthy middle aged guy I haven't found a real reason to track 24x7 HR or sleep.
Couldn't we apply that statement to steps, HR, as well?
Why would a healthy middle aged guy want a ferrari instead of a ford? We can apply that statement to everything in life, but that's besides the point of this topic.
That was your point..........
11-01-2017 09:08
11-01-2017 09:08
I will offer my personal unbiased opinion to this context. I am a couch potato turned crossfit athlete. Before embarking on a life style change, I was on bp medication for nearly 12 years. I was on substantially higher statin dose for 17 years. I had a ST elevation abnormality on ECG. Every Friday, legs would automatically lead me to a cigar bar. I smoked, drank fine wine. I used to enjoy fine spirits money could buy. One fine day I had two choices. Go buy a 34 inch pant which would make me appear as if wearing a pant meant for an elephant or do something to reduce the waist. I am genetically small framed. I went to a local Globo gym where I accidentally ended up with one of the finest coaches. He asked me what my goal was. I replied that I want to fit in my suits and reduce a few inches off my waist. There goes my transformation story. I am now an athlete competing at local level events. I am completely off statin and bp medication; much to the surprise of my physician, I no longer have ST elevation. At 48 I am able to run a mile in six minutes after doing 100 pullups, 200 pushups and 300 squats. Two years ago, I could not run more than 2 minutes at a 4.5 mph pace on a treadmill. Today I am the best runner in the whole of my gym. I can do a full tabata at 8mph 12% incline. I will challenge anyone to try it and see how hard it is on the body. It only takes 4 minutes but yet it is a total destruction on our nervous system. From nearly 27% body fat I am now at 9%. I am now slightly less than 27 inches at my waist. I had to spend 900 dollars in altering my pants. I am impressed with myself and my performance when I realize that I can compete and outrun young active duty military personnel who are in great shape. My cardio respiratory endurance has remarkably improved. I don't have time to drink now a days as it messes with recovery. I don't smoke. From training once a day I now train twice a day like every other garden variety full time athlete except that I am also an executive. I was not so into swimming but I want to give triathlon a try and I am actively swimming now-a-days. I have reached a point where sometimes I am forced to take time offs only to train. This is not to be misconstrued as an addiction. I don't have time and I have to make up time.
Although fitbit had nothing to do with this kind of transformation, I am sure it would not have been possible without tracking calories, heart rate, heart rate variability, religiously documenting what I eat and adjusting what I eat etc.
Today it is all about data science and how you can get the data what you can do with it can be a life changer. Predictive analytics is a big part of life. It is not only limited to stock markets. It can be applied in everything we do.
Tracking sleep is crucial as it tells me and my coach about how to adjust my training, especially the percentage of deep sleep. On days I have had a larger chunk of deep sleep, I can see I recover well. Although sleep itself is a brain oriented activity, research shows that heart rate variability (HRV) is also something that could be used to glean how we sleep. Midway during my past local event schedule, my resting heart rate breached a red zone. Because I had no option other than to compete I competed and as you would expect I strained quad muscle. Although it is fairly mild, the case in point is that it would not have happened normally as I would have voulantarily backed off by adjusting intensity or rest sans competition. That is why it is crucial to me and those who know how to use the data.
You said tracking HR 24 x 7 is not needed. I say you can get insights about body's inflammatory response in response to training intensity. Long story short, use of Fitbits and polars helped me a lot in making proper decisions about training. Could I have done the same thing with Apple watch or Garmins?. May be. But I went fitbit route. Moderate loss of accuracy is fine with me as I can compensate however I can't sustain grossly inaccurate data. I am not an apologist for fitbit and thanks much for reading this long reply.
11-01-2017 10:04
11-01-2017 10:04
I would never compare 2 optical HR devices. The optical HR technology (sensors, algorithms, software) that is currently present in our watches is not there yet, it still struggles to achieve perfection. If I were you I would lend or buy a watch with a chest strap from a reputable brand (Polar, Garmin, Suunto) for accurately readings the HR and compare those readings with the Ionic and Charge 2.
The chest strap readings are one of the best you can possibly have... except maybe from some fancy and expensive medical equipment.
My experience : I have a Garmin Fenix 3 HR that worked pretty decent for me when I used the optical HR and when I compared it with the Ionic I realized that the Ionic HR does not work too well when my HR is past 130BPM. The Ionic tracks my heart very good as long as it is from 55bpm to 125bpm. When I do a sprint and my heart goes past 140bpm it does not show correctly the spike in bpm as the Garmin does even if it has a certain lag specific to the optical HR technology.
Just to be clear : The Fenix 3 is an old mature product and received numerous firmware updates, some related to the HR so I never expected the Ionic to perform at the same level simply because it was just launched and I know first hand how this industry runs these days... we need patience and updates 🙂 . I only did the tests to see how it stands and to know what to expect from it at it's current stage. I have high hopes for Ionic because it's HR reader it's faster and more precise than my old Fenix 3.
11-01-2017 10:26
11-01-2017 10:26
Agreed that chest heart rate monitors using EKG signals are more accurate. However aren't you contradicting your statement. Why compare fenxi 3 and ionic?
The point in comparisons is to understand whether we should continue to use ionic (until its issues are fixed) or not when we have an older model which we had validated to be accurate. When someone pays 300$ for an item he is rightfully correct to validate accuracy. For example to see its effectiveness I tracked my weights session the other day something I don't usually do. IONIC was on hiper and it was grossly inaccurate. I records a max heart rate of 160 on a dumbbell bench press. I think I know how I feel when my heart works at that rate.
although I didn't wear blaze on that day, I am sure it wouldn't have done that. I have also attached the polar recording for the same weights session.
Now here is what Polar shows for the same 58.20 min duration.
I haven't found any use case for ionic until its issues are fixed.
11-01-2017 10:37
11-01-2017 10:37
@Venkats wrote:Agreed that chest heart rate monitors using EKG signals are more accurate. However aren't you contradicting your statement. Why compare fenxi 3 and ionic?
The point in comparisons is to understand whether we should continue to use ionic (until its issues are fixed) or not when we have an older model which we had validated to be accurate. When someone pays 300$ for an item he is rightfully correct to validate accuracy.
Have you read most of the replies on this forum? The kool-aid is in abundance. People are defending this beta watch and software at all costs.
11-01-2017 10:52
11-01-2017 10:52
Yes but I think they are all considering mostly the ability to listening music, store music, paying at starbucks, looking at weather etc. I expect a fitness oriented watch to do better on those line items oriented towards fitness. Across the board I have seen most people see heart rate aberrations. Since everything eventually is tied to the accurate monitoring of HR, hopefully development will fix the inaccuracies. Inaccuracies at higher intensity can be tolerated since it is comprehensible. There is some whining with respect to SpO2 but that is something I would say it is their own fault of not reading fine prints.
In a user forum it is normal to see agreements and disagreements. When I am not overtly pleased with IONIC as released, I am now able to see my exertion during shower which I never had insight into it. Can you believe I get 14 active minutes from showering confirming it is also an exertion (isometric) of some sort.
There is literally no public review that highlights the gross inaccuracies. Almost every review praises the watch and some highlight even SpO2 sensor which is something not available.
11-01-2017 10:53
11-01-2017 10:53
I've used my Fenix 3 HR with Optical HR and chest strap for a long time and I learned that the optical HR is pretty accurate (for me) with some exceptions of course (i am not getting into details). I am also using sports watches with chest straps for 5 years now and I know too well where my heart is in therms of bpm to know if the readings are correct or way off.
I did not contradict myself, I just gave you my results comparing 2 optical devices, 1 that I know for a long time, it's been tested in many scenarios, many times and I know it works fine, and 1 new, untested... I compared them both on optical just because I was curious to see how is the lag on both and how accurate (roughly) are.... Just to be clear, I am not praising the Garming here, the Fenix 3 HR is not perfect, the HR technology in general is far for perfect but for some things it works fine.... like running, cycling.
I did not compared the Ionic with a chest strap because I could tell from the start that it does not work quite right... when you know your body and when your HR is at 160-165bmp and you look at the watch and wait to catch up and you don't see it getting past 142 bpm you realize that it needs more work....
At least this is my experience and I am sharing it.
As I said the Optical HR technology is still not where all we love to be to give us good and consistent reading in various scenarios...Also what works for me it does not work for the next guy.....because it does have a lot more hair on his skin, or the skin tone is different, or how is wearing the watch....
We need patience and updates 🙂
11-01-2017 10:59
11-01-2017 10:59
Actually we are making the same point, is it not?. In my case I know I can trust my Blaze as well as Surge as long as I comprehend neither will be accurate at higher intensities. I have done several verification to that effect. However with IONIC it is not so accurate.The originator of the thread compared Charge believing it was accurate. You compare fenix believing it was accurate. I compare blaze believing it was accurate. Least common denominator is ionic which we all seem to agree being inaccurate.
Right or wrong?
Hopefully all of this goes into development team's ears.
11-01-2017 11:03
11-01-2017 11:03
@Venkats thanks for sharing, and congrats on changing the course of your life! My wakeup call was five years ago, about the time I turned 50 and had three surgeries including major shoulder repair. In the last two years I've transformed myself into a cyclist. That said I'm genetically predisposed to be a slightly above-average endurance athlete. There is a large and active cycling community here, with many former racers, so its always fun trying to hang onto the wheel of someone in a different league (genetically and experience wise).
I started training with Fitbit and quickly realized that steps were meaningless and sometimes counter productive. Don't get me wrong, competing in step challenges is fun, but often led to unwise decisions like not going to bed on time and instead pacing around the house after 10pm to win a challenge. Lately I've been training 10 hours a week, enough that its difficult staying awake after 9:30pm given I must wakeup at 5:30am. I still track sleep using an app that doubles as alarm clock. The most interesting data in it is the amount of minutes snoring at night.
Agree with you about predictive analytics, I've come to rely on TrainingPeaks to monitor, plan, and predict impact of overload and recovery (adaptation). I also cross-check recovery predictions using HRV, with the recovery feature on my bike computer. In addition I also cross-check recovery level with an HRV phone app and chest strap in the morning when I wake up. Here is an example of my rising and falling fitness level from TrainingPeaks:
Higher fitness levels on that chart correlate to an increased performance on both short (weekly group ride) and long events (10,000' of climbing). As a mini guided tour, on the 28 and 90 day charts you can see that I'm ramping up my cycling fitness level for a big event on November 11th. On the 365 day chart you can see two previous fitness peaks, one for a February event and the other for the May double century. For reference, back in May my sustained 1-hour power output was about 10% higher than it is today.
By entering my future training sessions:
TrainingPeaks will predict fitness levels in the future (this is the simplified graph):
I also use another fitness platform that predicts my ability to sustain long hard efforts (1 hour power), like the signature 3000+ foot climb in 9+ miles of the upcoming Mt Fig event.
So yeah, I bought into predictive analysis two years ago and none of those tools use steps or sleep or 24x7 HRM. I do monitor morning RHR and HRV, but HRV monitoring isn't something Fitbit offers and Fitbit overestimates RHR. TrainingPeaks was designed for professional endurance athletes and their coaches. I'm self coached, and once I understood how my body reacts to training and my individual tolerance for ramping fitness level, I've found the tools to be very accurate in planning for events because it does a good job at predicting my future fitness level and overload/recovery requirements.
Aria, Fitbit MobileTrack on iOS. Previous: Flex, Force, Surge, Blaze
11-01-2017 11:28
11-01-2017 11:28
I've seen a lot of people on this forum outraged of the Ionic problems but I guess the people from this community don't read the Garmin forums, Polar, Suunto to see what problems and bugs have those watches that cost a lot more ( 600-800$).
I am not saying that this is normal but for those who felt the pulse of the tech fitness industry know that this is how things were done in the last 3-5 years. They (all major brands) put the watch out with some features and lots of bugs and while they were selling it they developed the software and added the features that were promised and turned the watch from a beta watch with some features in a full featured watch. The problem is that usually it takes time, a lot of time... and people these days don't really have patience and/or time to wait.
I am not a fan of Fitness but I bought it because I like to test first hand the new devices. I don't trust the so called reviews... most of them are full of empty descriptions, nice words but none of the real problems.
I understand that for some heaving a watch with issues/bugs it's a bit drawback but there isn't much you can do about it besides waiting for fixes if you are loyal to the brand or you have many friends in the ecosystem or you just really love the watch (design, accessories).
I usually have patience if I like the hardware but if the software development is really slow I can always return it... buy it later when and if it matures... or move to another product, there are alternatives.
It's true that I have some good old watches that I can always fall back to 🙂
11-01-2017 11:40
11-01-2017 11:40
@Venkats wrote:Actually we are making the same point, is it not?. In my case I know I can trust my Blaze as well as Surge as long as I comprehend neither will be accurate at higher intensities. I have done several verification to that effect. However with IONIC it is not so accurate.
Interesting, I have found both my Ionic and Surge yield virtually identical results, the initial few minutes of a run (where the Ionic often shows a spike in my heart rate way above reality) being the lone exception.
11-01-2017 12:18
11-01-2017 12:18
@Davidbit wrote:
@bobo03 wrote:
@bbarrera wrote:
@Davidbit wrote:So, well, who's lying here? And how can I be sure that I've got some thrustworthy data to work with? I really looked forward going to the Ionic from the Charge 2 but now I'm doubting both.
@Davidbit Your guess is as good as the trackers! As a healthy middle aged guy I haven't found a real reason to track 24x7 HR or sleep.
Couldn't we apply that statement to steps, HR, as well?
Why would a healthy middle aged guy want a ferrari instead of a ford? We can apply that statement to everything in life, but that's besides the point of this topic.
I'm a BMW guy 😉
Going back to the original topic of HR and sleep accuracy. The point I'm making is that unlike testing step counting accuracy with a 100 or 1000 step count test, it isn't possible to test sleep tracking accuracy. What's the point of tracking sleep states if you have no clue about its accuracy? And I'd argue that sleep states it isn't inherently actionable for most people, my goal is 7-8 hours of sleep a night and beyond that I can't actively do something tonight to increase deep sleep. And when Ionic and Charge 2 don't agree, what can you do? Other than make you aware that sleep state estimation is hit or miss, is it really going to change anything? To me its like stair climbing data, the data is hit or miss, the graphs are fun to look at, but the data itself is not of much significance to how I live my life.
Steps, well, that is relatively easy to verify the accuracy at least under certain somewhat ideal conditions (walk test). And the key point is that for people with a sedentary lifestyle, monitoring step count is something that can drive behavioral changes to improve activity level and lead a more healthy lifestyle. For example if I only have 5000 steps at 7pm, and haven't done any cycling/swimming/weightlifting, its either a rest/recovery day or a reason to take the dog for a walk. Either way its something actionable.
HR accuracy is a more interesting discussion. HR is actually part of the "input" along with fuel and muscle strength and metabolic efficiency that drives the "output" - running, cycling, swimming, etc. For healthy humans the question is really "are you getting enough moderate to vigorous cardio exercise per week." The 24x7 HRM graphs are fun to look at, but at the end of the day its about getting 30 minutes of moderate to vigorous cardio. If you are looking to increase performance, there is some interesting trending data for endurance athletes by correlating output (pace or power) to input (HR), but that isn't tracked by Fitbit and more importantly requires an level of accuracy I'm unable to obtain with any optical wrist HRM. Fitbit pushes the concept of optimizing workouts for HR zones, and better calorie estimation. If it works for you, great, I haven't found much value myself (accuracy issues while cycling). Beyond that I think it brings us to a discussion of how HRM accuracy is highly individualized, the merit of using oversimplified zones, Fitbit calorie estimation algorithms, etc.
So in summary, I believe that sleep state accuracy isn't something to get worked up over. You can't independently verify its accuracy and if you sleep on a regular schedule the data can't be used to drive any changes in behavior. As for HR, well, thats a bit of a judgement call for you to decide 😉
Aria, Fitbit MobileTrack on iOS. Previous: Flex, Force, Surge, Blaze
11-01-2017 12:41
11-01-2017 12:41
Sleep stage accuracy has been independently verified by John Hopkins university and there is a published document. fitbit technology is 96% accurate with respect to a EEG based system. One would assume they adopt the same (if not better ) proven technology / algorithm across all device eco system. 96% accuracy is more than sufficient for a non clinical grade device, is it not?. There are several studies in evaluating accuracy of various sleep trackers and all comparisons were made against EEG based measurement.