Fitsense (in)accuracy at fixed cadences

Quoted message said:

At this new cadence and over my nominal 2 mile flat paved
course, the watch now indicates 2.05 miles when running at
a 9.6 min/mi pace, and
1.91 miles when running at a 7.4 min/mi pace (both
maintaining a 180+-3 ff/min cadence).

The Fitsense loses accuracy as your stride changes, and a
quicker stride plus varying speeds may throw it off. You may
need to recalibrate it at your new cadence, at both speeds.

I think this is a limitation of accelerometer
technology, and other brands would show similar results.
(But I'm no expert.)

--
Brian P. Baresch Fort Worth, Texas, USA Professional editing
and proofreading

If you're going through hell, keep going. --Winston
Churchill

At the faster pace even with 180 steps/minute, you are
covering more ground with each step than at a slower pace
with a 180s/m. You now realize that if the distance each
step measured by the speedometer watch is constant, when you
increase your speed you distance covered each step
increases.

What I have people do is run through water that I pour on
the ground. They start about 100 feet away and get up to
the speed they estimate is a 7 or 9 minute pace by the time
t hey run through the water. We then measure the distance
of one step from the heel imprint of the back foot to the
back of the heel of the front foot. They begin to realize
they're covering more ground even though the cadence
remains the same.

In health and on the run, Ozzie Gontang Director, San Diego
Marathon Clinic, est. 1975 Maintainer - rec.running FAQ faqs.orgrunningOpen ↗
faq/ Mindful Running: mindfulness.commr.aspOpen ↗

From: [email hidden] (Ronald H. Nicholson Jr.)
Newsgroups: rec.running Subject: Fitsense (in)accuracy at
fixed cadences Date: Sat, 5 Jun 2004 02:04:27 +0000 (UTC)

I bought a Fitsense FS-1 speedometer watch a couple years
ago and have been quite satisfied with it's usefulness and
accuracy until a few weeks ago.

I had been running at a cadence that varied over a range of
about 10% from slow to fast pace (160 ff/min to 175
ff/min); and the watch had been consistantly indicating my
regular course distance to about +- 1% repeatability at any
and all paces.

The last few weeks I've been working on changing my cadence
from variable to a constant 180 ff/min for all paces
between 7 min/mi to 10 min/mi. At this higher cadence, my
knees are happier and I can run farther, but my stride
seems a pinch choppier.

At this new cadence and over my nominal 2 mile flat paved
course, the watch now indicates 2.05 miles when running at a
9.6 min/mi pace, and
1.91 miles when running at a 7.4 min/mi pace (both
maintaining a 180+-3 ff/min cadence). This is an 7%
distance measurement variance over a 29% pace
difference. The batteries are fairly fresh; and the pod
location is taped down to the laces of a single running
shoe. But the watch is now only acting like 3/4rds of an
accelerometer and 1/4th of a pedometer
(deltaDist/deltaPace at a fixed cadence).

I wonder if the Nike or Polar would be any more accurate
over a wide range of running speeds given a fixed cadence.
Of could the Fitsense just be worn out of old age?

Ronald H. Nicholson Jr. said:

The last few weeks I've been working on changing my
cadence from variable to a constant 180 ff/min for all
paces between 7 min/mi to 10 min/mi. At this higher
cadence, my knees are happier and I can run farther, but
my stride seems a pinch choppier.

You've got to recalibrate your unit. As you observed, a
different cadence will throw it off, and it does have some
drift with pace. I don't think it has anything to do with
wear and tear on the unit. Expect any accelerometer to have
similar problems. If you wish, you could get a GPS unit, and
deal with its problems/limitations instead.

I wouldn't count on any of these doodads to give an accurate
enough measurement that you could use it to measure your
*pace*, though both the fitsense and the GPS units should be
adequate for logging milage (and you should be able to
reliably get within 1% when you're close to your usual
training pace)

Cheers,
--
Donovan Rebbechi pegasus.rutgers.edu~elflordOpen ↗

Quoted message said:

I wouldn't count on any of these doodads to give an
accurate enough measurement that you could use it to
measure your *pace*, though both the fitsense and the GPS
units should be adequate for logging milage (and you should
be able to reliably get within 1% when you're close to your
usual training pace)

I've been using my Fitsense to estimate my pace on long
tempo and MP runs, and it seems to be doing OK (though I
don't have any other thing to measure it against). I think
it beats wild guesses, which is my other option.

--
Brian P. Baresch Fort Worth, Texas, USA Professional editing
and proofreading

If you're going through hell, keep going. --Winston
Churchill

In article <[email hidden]>,

Donovan Rebbechi said:

You've got to recalibrate your unit.

It doesn't seem possible to calibrate it, given that it
reads high at one running pace and low at another
running pace, over an identical course. Any calibration
would make one situation more accurate, but the other
even less correct.

IMHO. YMMV.
--
Ron Nicholson rhn AT nicholson DOT com
nicholson.comrhnOpen ↗ #include
<canonical.disclaimer> // only my own opinions, etc.

In article <[email hidden]>,

Ozzie Gontang said:

At the faster pace even with 180 steps/minute, you are
covering more ground with each step than at a slower pace
with a 180s/m. You now realize that if the distance each
step measured by the speedometer watch is constant, when
you increase your speed you distance covered each step
increases.

Not quite. It is true that (given no foot slippage):

speed = cadence * stride_distance * kConstant

So if the speedometer watch was a pedometer (a pedomenter
doesn't measure stride length, only the count), the error
would be 100% correlated with changes in speed. But the data
below indicates that my FS-1's error is only 25% correlated
with changes in pace. So the accelerometer must be measuring
my stride length changes about 75% correctly. Not too
accurate, but still better than nothing, assuming the data
is at least repeatable at a given constant pace and cadence
(seems to be).

Quoted message said:

Quoted message said:

At this new cadence and over my nominal 2 mile flat paved
course, the watch now indicates 2.05 miles when running at
a 9.6 min/mi pace, and
1.91 miles when running at a 7.4 min/mi pace (both
maintaining a 180+-3 ff/min cadence). This is an 7%
distance measurement variance over a 29% pace
difference. The batteries are fairly fresh; and the pod
location is taped down to the laces of a single running
shoe. But the watch is now only acting like 3/4rds of an
accelerometer and 1/4th of a pedometer
(deltaDist/deltaPace at a fixed cadence).

To calculate stride distance at a given cadence and pace:

inches/ff = (inches/mile) / ( (min/mile) *
(ff/min) )

So:

stride_distance = 63360 / ( pace * cadence )

Or a 44 inch stride running 8 min/miles at a 180
ff/min cadence.

IMHO. YMMV.
--
Ron Nicholson rhn AT nicholson DOT com
nicholson.comrhnOpen ↗ #include
<canonical.disclaimer> // only my own opinions, etc.

Ronald H. Nicholson Jr. said:

In article <[email hidden]>,

Donovan Rebbechi said:

You've got to recalibrate your unit.

It doesn't seem possible to calibrate it, given that it
reads high at one running pace and low at another running
pace, over an identical course. Any calibration would make
one situation more accurate, but the other even less
correct.

Right, so you'll need to make a choice. I do most of my
milage at an aerobic pace, so I use the same pace to
calibrate the unit. Needless to say, I don't use the unit to
pace myself during speed workouts.

Cheers,
--
Donovan Rebbechi pegasus.rutgers.edu~elflordOpen ↗

Quoted message said:

It doesn't seem possible to calibrate it, given that it
reads high at one running pace and low at another
running pace, over an identical course. Any calibration
would make one situation more accurate, but the other
even less correct.

I've heard of folks who adjust the cal vals depending on
whether they'll be going at a "normal" training pace or
a tempo pace, after calibrating it for each. So if
you're going to be doing 9 minutes per mile you'd set it
at one value, and if you're doing a tempo run you'd set
it at another.

--
Brian P. Baresch Fort Worth, Texas, USA Professional editing
and proofreading

If you're going through hell, keep going. --Winston
Churchill

In article <[email hidden]>,

Brian Baresch said:

I've heard of folks who adjust the cal vals depending on
whether they'll be going at a "normal" training pace or a
tempo pace, after calibrating it for each.

Thanks! That's a great idea. I never thought of using
multiple calibration settings, depending on intended pace.

I guess I'll need to go to a track, run some
fast/medium/slow miles and develop a (re)calibration table
or formula for my FS-1.

IMHO. YMMV.
--
Ron Nicholson rhn AT nicholson DOT com
nicholson.comrhnOpen ↗ #include
<canonical.disclaimer> // only my own opinions, etc.