Floor count accuracy

[HHAYNIE]

On a 200 stair set, my Charge HR seems to count the first 100 or so accurately, but then begins to undercount significantly. Example I manually counted from 139 to 153, Fitbit showed floor count of only 143, yet it counted the first 100 or so accurately.

[HeydyF]

Hi @HHAYNIE, welcome to the Community!  Fitbit trackers are very accurate, so if you manually count the steps or floors you are walking, or like on this case, climbing, you'll see that the tracker will give you a very close amount of them. Based on that, I see that the amount of floors remaining is not that high, so your tracker still pretty accurate. 

Let me explain you, your Charge HR detects floors using an altimeter, which is a sensor that calculates altitude based on atmospheric pressure. Though your tracker is designed to look for pressure changes based on elevation gains, pressure changes due to other causes—such as a gust of wind, a weather change, or opening a door—can occasionally cause your tracker to register an extra floor or two. Another factor is floor height. Your tracker registers one floor when you’ve gone up about ten feet, which is the average between residential and commercial floor heights. If you climb long staircases you may find that the tracker’s floor count doesn’t match how many floors you’ve gone up since the staircase was taller than ten feet.

If you havve any other question, let me know! 

[HHAYNIE]

Thanks for the explanatory response. Interestingly, my floor count accuracy
problem is always one of undercounting and seems to get worse after I've
developed a good sweat. Two examples: first in my house going up and down a
single flight of stairs, my Charge HR counts the first 75 or 100 fairly
accurately, advancing one floor per flight of stairs, then, it begins to
miss flights, and won't advance for for the next 4 or 5 or 6 flights, then,
it will intermittently start counting again. The undercounting send to
correlate most with profuse sweating; I can't relate it to doors opening or
furnaces clicking on. Second, in a fairly steep loop hike in the Columbia
River Gorge, say 3000 get elevation gain, it significantly understated the
gain. Again, these are high sweat hikes.

I'm now on my 3rd Charge HR, as they seem to malfunction - won't charge
anymore, display goes dark - with heavy daily use.

Initially, steps/ distance was highly inaccurate also, always low, but, I
adjusted it with the stride length setting and now, it's reasonably
accurate.

I wonder if there is some configuration setting for pressure sensing
sensitivity that could me made available in a future software update.

I really love the minimalist design of the Charge HR complemented by the
comprehensive easy to use website, but as a heavy daily user for both
health and fitness reasons, I am disappointed in the mean time to failure
of the device and frustrated with the floor undercount issue, and, to a
league extent, the heart rate insensitivity during exercise.

My sense is that water proofing, particularly due to large amounts of
sweat, is a significant issue. Personally, I would pay more for a water
proof model. I've read about the Charge 2, but it appears to have no better
water resistance than my HR. I don't know if the pressure sensing altimeter
technology has iterated for the Charge 2 or not.

On my hikes, including 5 different Caminos across Spain, I also use Garmin
GPS devices, beginning with the Dakota, then the Oregon, and now the
Montana. I would love to do down to a single device, maybe a future Fitbit
will include better GPS support with way point abilities.

Harlan

[HHAYNIE]

Here are today's test results. I live in a 2 story house. The stairway is just over 10' in height, maybe about 10 feet 2 inches. Today again, I did 200 stairs, this time tediously manually counting them. Here are the results of my Fitbit count versus my manual count:
40 - 40
67 - 80
69 - 100
69 - 120
69 - 140
77 - 160
81 - 180
82 - 200

Interestingly, on successive days, although the floor count is uselessly inaccurate low, the steep and distance counts after each 200 set are consistent, at about 7 kilometers.

I have a Charge HR. If a moderator read this, I would like to know if the new Charge + had the same or an improved mechanism for counting floors.

At this point, although I love my Charge HR, I am researching the internet for alternative minimalist wrist devices that provide more accurate altimeter results. Maybe devices with built in GPS will be more accurate.

[HHAYNIE]

Today, anticipating a long hike tomorrow. I did a shorter stair climb workout, but the results were the same as always, accurate floor counts at first followed by increasingly reduced counts. The first column is the Fitbit count the second my manual count:
20/20
39/40
59/60
78/80
88/100
95/120
97/125

[HeydyF]

Hi @HHAYNIE, thanks for sharing your results here.  Yeah, as you mentioned, the difference seems to getting higher when you climb more stairs. Interesting. Although, as I mentioned before, all of the Fitbit trackers have this little error range so not all of your floors will be counted, 'cause they are not 100% accurate. And after checking all the numbers you posted, only at the end is when the amount changes more. So let me suggest you something, please try to restart your Charge HR by following the instructions on this post. 

Let me know how it goes! 

[HHAYNIE]

Hi HeydyFitbit,

I tried your reset suggestion this morning - my Charge HR reported v122. I
then immediately did a stair set with the following results: Fitbit/manual
count.

20/20
41/40
60/60
80/80
97/100
103/120
105/140
112/160
113/180
120/200

Disappointingly, the pattern of undercounting persists after a reset, with
significant undercounting beginning, as before, about halfway through the
workout. Distance was 6.38 kilometers, a little lower than usual, but
generally in the ballpark of the last several workouts.

I'm wondering if the pressure based floor measurement is temperature
sensitive as around halfway through my Fitbit is wet with a lot of sweat.
As this water evaporates, like in a swamp cooler, it could lower the
operating temperature inside the Fitbit slightly.

One thing I would like to know, if you have time, is: does the new Charge 2
have the same floor measurement hardware as my Charge HR?

I'm also wondering if devices with built in GPS generate more accurate
elevation measurements.

Thanks for your help and let me know if you have any more ideas or
suggestions. The good news is that all these 200 stair sets are really
getting me in shape (I'm 67)...testing ideas is great motivation. Thanks
again for your help and attention.

Harlan

[HHAYNIE]

 Sunday, Dec 4, 2016 Floor Count Report:

Km/∆km/floors/∆floors/manual/time/∆time/sweat level

0.67 /          / 20 /  20 / 8:08 /.     / +

1.31 / 0.64 / 39 /  40 / 8:18 / 10 / ++

1.93 / 0.62 / 40 /  60 / 8:29 / 11 / +++

2.55 / 0.62 / 40 /  80 / 8:40 / 11 / +++

3.18 / 0.63 / 40 / 100 / 8:51 / 11 / ++++

3.81 / 0.63 / 42 / 120 / 9:02 / 11 / ++++

4.45 / 0.64 / 46 / 140 / 9:12 / 11 / ++++

5.12 / 0.67 / 52 / 160 / 9:24 / 12 / ++++

5.77 / 0.65 / 59 / 180 / 9:36 / 12 / ++++

6.43 / 0.66 / 59 / 200 / 9:48 / 12 / ++++

7.07 / 0.64 / 61 / 220 / 9:59 / 11 / ++++

7.78 / 0.71 / 68 / 240 / 10:10 / 11 / ++++

Sweat level +++ is full sweat, ++++ signifies Fitbit is covered in sweat

The kilometer distance per 20 stairs is quite consistent, as is the floor count accuracy for the first 40 or so stairs, which effectively negates the hypothesis that the Fitbit is broken, and instead suggests that floor count accuracy errors low due to some change introduced over time. My working hypothesis is that it's related to sweat, probably more likely to the evaporation of sweat and it's associated heat of evaporation loss.

At any rate, my Charge HR is clearly dramatically inaccurate low in documenting a single stair climbing session of more than about 20 minutes, disappointingly rendering it ineffective in documenting workout history to the associated website over time perons of weeks and months.

I would really like to know if the new Charge 2 suffers from this same problem

[Rictoo]

My fitbit HR seems to go out after about 70 floors. With steps there's at the end of the day if the fitbit  is in the range of 15,000 then my phone will be about 3,000 steps less.

[G1XOZ]

My ionic was underestimating floors and my new versa 2 take a complete stabin the dark, in fact it crap at it.

obviously where ever they decided 10 feet between floors live in large building and surely there should/could be a factor value  us brits could add to lower the height of our buildings?

i went up and down the stairs yesterday at home at least 50 time and it counted a measley handful of floors