You mentioned measuring the distance between the white stickers down to the MM of accuracy... I understood it as that distance was completely irrelevant to how RUNN calculates the pace. All that matters is how long it takes for one sticker to pass under the two sensors so there is nothing gained by having the distance between the stickers the same. Do you understand it to work differently?

---

______________________________________________
Team Zoot

That was my understanding as well, @Gregtay. The manual seemed a bit ambiguous in terms of how many to actually use. I put two on the treadmill, have done one run and it was pretty damn close all in all. Much closer than my Stryd (either iteration).

---

IG: NCGregory8778

That is also how I read it in the manual: put some stickers on the treadmill.


So I did that in the beginning, placing stickers a foot apart with the very last one farther than the others. From my understanding, the sensor will look for patterns to figure out if the full rotation of the belt has been completed. But when I was calibrating the sensor, it looked like it would "calibrate" itself too quickly, before the full cycle completed. That was my perception, at least.


In the review, I gave an example of a 13.1-mile run, which was reported by the treadmill as 13.4. That was with 5 stickers placed 2 feet apart (last one taking up the difference). I was already pretty happy with that result.


But then I decided to go down to 4 stickers at 33 inches (838 mm) apart evenly spaced. Precise placement would eliminate any guesswork from the sensor (my opinion). I then ran a 10K interval run (which is the most error-prone) right after, and it reported 6.3 Zwift vs 6.27 Treadmill miles. That is almost identical.


Extrapolating that reading to a half marathon, it would be a difference of 0.07 miles, whereas before it was 0.3 miles. That is a huge difference for not much extra work of figuring out the exact placement of stickers.

The Runn measures how long it takes for one of the white stickers to pass two sensors of known distance. The distance between white stickers is irrelevant. I see no reason to calibrate a Runn against the speed of a treadmill. Treadmills are all over the place in terms of accuracy. Provided the Runn is not defective, it should provide a close measurement of the speed, and most likely will be far more accurate than the treadmill it is used on.

I agree on the finnicky-ness of the installation, I had to redo it once, trying to level my bracket and get the height of the sensor from the belt. In terms of stickers, the distance can't be completely random, I did that at first, and my Runn was registering double my speed (I had 4 on at roughly 12" distances. I settled on two stickers, about 2feet part, and I have had great success and accuracy with mine, once I recalibrated it. After that, I just leave mine plugged in to the wall all of the time and I get consistent readings

That makes sense and explains why having two many stickers on the treadmill overwhelms the sensor. How about intervals? Wouldn't rapid acceleration and deceleration affect the readings between randomly spaced stickers? Aesthetically it is also nice to have them aligned ;)


I wonder why NPE put a white button on the Runn sensor to calibrate it and wrote a step-by-step guide on how to do it using a treadmill at 5 mph. My Runn is not defective and after all the recalibration is a lot more accurate.


True in most cases, but highly dependent on the build quality of a treadmill. I run on a ProForm Boston Marathon 4.0, when I bought it, it was on the higher end of the spectrum. Solid motor with up to 15mph and almost zero belt slippage. So I trust what it reports.

Sorry to say this, but your logic and review is just riddled with methodology errors.

No problem, I don’t take it personally!

Could you describe in more detail and specifics what methodology errors and logic I made? This way we will all learn something new and it will help me write better reviews in the future.

Thank you in advance!

1) If you have only 2 devices on which to compare, it is impossible to know which one (if either) is accurate.
2) You don't even understand how the Runn works (ie, spacing of the strips doesn't matter as the space between strips is not measured)
3) The device comes pre-calibrated and NPE advises against calibration against your current treadmill because if your treadmill is inaccurate (likely) now your Runn is too
4) Treadmill accuracy can vary with speed, so comparing runs against eachother is pointless unless they are at the exact same speed, and even then may vary (your 10k v 13.1 mile runs)
5) The error rate in each measurement means a .07 mile difference v a .3 mile different in a run of 10k and 21.1k may or may not be meaningful. Depends on the measurement system and its precision.

Since the belt slightly accelerates/decelerates with each footstrike, the time it takes the sticker to pass between the two sensors may vary slightly depending where you are at in the gait cycle. It seems like having more stickers would provide a better chance of getting a more accurate overall reading (assuming, of course, that there aren't so many that a new sticker passes under the Runn before the previous one exits).

---

ECMGN Therapy Silicon Valley:
Depression, Neurocognitive problems, Dementias (Testing and Evaluation), Trauma and PTSD, Traumatic Brain Injury (TBI)

I really appreciate your detail response and this info is fascinating.

First of all, my review is that of an "end-user" not an engineer. It is not meant to be a scientific side-by-side test measuring accuracy between treadmill, running power meter pods and treadmill trackers. The baseline for me is a treadmill and I'm absolutely happy to see so close to each other. I care about precision, not accuracy.

Lets take it point by point:

1. This review only compares Runn against a treadmill, but I also ran with Stryd and Garmin foot pods on it too. So I have an idea of how close the results should be to what the treadmill reports. I found Stryd to be very accurate and reliable, I just don't like charging it all the time.

2. Thank you for the explanation on how Runn acutally works! It is true, I don't understand how Runn tracks speed because I can't find any information in the User Guide or on their website explaining it. Here is the best info I've found so far on NPE's website:

Quote:

Q: How does Runn track speed?
A: Included with your Runn purchase are sensor strips that are placed on the treadmill belt, and as they pass under your Runn, they are picked up by speed sensors.

3. I can't find anywhere that says that NPE is advising against recalibrating the Runn. In fact, they've created an entire page dedicated to recalibrate Runn. Here is the first paragraph:

Quote:

When should I calibrate my Runn speed & incline?
If your speed or incline is off in your fitness app, you may need to calibrate your Runn. Runn is factory calibrated to a verified speed of 5 mph (8 kph) and 0% incline.

4. I totally agree about comparing runs with speed for testing purposes. I almost always run at the same 7:30 min/mile (8 miles an hour) on the treadmill, with the only exception being intervals. I've done many 5k and 10k runs at constant 7:30 pace with the Runn, and the results were similar.

5. The goal is not to scientifically prove that Runn is much more accurate than the treadmill. After recalibrating Runn and readjusting belt strips I was able to get more precise readings, which can be inaccurate. An extra third of a mile to run feels a whole lot harder when running a half marathon rather than a 5K.

I'd love to get into more details on how Runn works and the best ways to measure pace and distance on the treadmill. Again, thank you for these insights, @marklemcd!

btw, NPE just introduced, i believe today, its updated firmware. you'll find it in the app store, it's called "configurez", and it's pending in google play. it's the same as the beta 3.0.2, so, if you have that you don't need to upgrade.

---

Dan Empfield
aka Slowman

Thank you the tip, Dan!

I'll get on it and will download and update my Runn. Here is a link to an official YouTube video by NPE on how to update the Runn:

https://www.youtube.com/watch?v=Q5SkJc2OnQU

Well, that didn't take long at all to update the firmware.

Installed the app, brought the phone near the Runn sensor and it recognized it. Pressed a few confirm buttons in the app and voila, the firmware update finished in under a minute. I'll have to do a run to confirm, but so far, everything seems to work properly.

I share a lot of the same thoughts that you described early on. thought this was going to be a 5/5 when i ordered it. but was very disappointed at first. the stickers would stay on, whiteout wouldnt pick up...etc.....speeds were all over the place. then i did something stupid. i calibrated it like they described. then i realized 'well this is stupid' i just calibrated it to say exactly what my treadmill is reading, which is what it does now. and my treadmill isnt not close to spot on. i feel like by now i know the difference between an 8:15 mile and a 7:30 mile.
anyways, is there a way to go back to factory setting? i wonder if the new update does that.

why would anyone ever want to calibrate to 5mph based on what the treadmill is saying? seems counter intuitive

My Runn sensor just arrived and I am ready to play fetch!

I just used 1 of those sensors. It seems accurate enough for me, for now.

My bigger concern is that my Samsung phone is too small to see well enough.

I went out & bought a new Samsung tablet that runs Android 9.

However, Zwift is telling me this machine is incompatible. Tell me it ain't so!

What's my next move?

Thanks.

---

http://www.fitspeek.com the Fraser Valley's fitness, wellness, and endurance sports podcast

I'm not an engineer, but this does sound plausible. I'd love to have more stickers on the treadmill if it does improve the accuracy. Maybe the future versions of Runn will have more processing power to handle extra stickers?


From an end-user perspective, there is almost not much extra effort involved in placing 4 vs. 8 eight stickers.

Just figure out the treadmill setting that you think is as close to 5mph actual speed as possible and run the calibration procedure with the treadmill running at that speed.

Do you have an old TV or monitor laying around? Get a USB-C to HDMI cable to project your phone to it. Your phone may have a mini HDMI port as well (depends on the model).

---

ECMGN Therapy Silicon Valley:
Depression, Neurocognitive problems, Dementias (Testing and Evaluation), Trauma and PTSD, Traumatic Brain Injury (TBI)

The only thing I've noticed with having 1-2 stickers is that the response to speed changes is a bit slower. Definitely noticed is last night during a Fartlek Pyramid set I did last night. But all in all it was pretty damn accurate.

---

IG: NCGregory8778

to reset to the factory settings, hold the white button down until the leds start to forward sequence, about 10 seconds.

https://npe-inc.zendesk.com/...0038799731-Runn-LEDs

---

salmon - not because I'm a fish

thanks for the advice, i tried this with my samsung phone and an old monitor - it works! resolution is not good. may have to shell out 140 $ to move into 1990's quality resolution.

---

http://www.fitspeek.com the Fraser Valley's fitness, wellness, and endurance sports podcast

How does it work after the firmware update? I got mine last weekend. First run was ok, got it to connect to Zwift on ipad, but could not get it to work with my Garmin 620. Start communicating with support. Do second run on Tuesday, drops out a ton, still no connection to Garmin. Support tells me to update firmware. Do that and now it won't recognize speed at all when on Zwift. When I use the GymTraker app they have it shows incline, no speed. I'm totally at a loss and have not heard from them since Thursday afternoon.

Since installing the latest firmware update I only ran once: 6.2 miles (Zwift) vs 6.14 (Treadmill). The big difference I saw right away was the cadence. On Zwift I got used to seeing it about 50-60% of the time because of drops. Since the firmware update the cadence is now showing 95% of the time and is much more accurate. I haven't noticed any other changes yet. Needless to say, my Runn sensor is fine.

I'd love to help. Is the problem now Zwift not picking up Runn after the firmware update?