CardioTrainer: Droid vs. Eris

[bilbravo]

First this isn't a "should I get a Droid or Eris" thread. I have a Droid and my wife has an Eris.

Recently we found the CardioTrainer app and are using it to track our workouts. Today was the first day we both used our phones, and found that we got wildly different speeds and distances travelled, even though we both walked at the same pace and did the same workout. The time was the same on both phones and the map that was drawn also was the same.

From what I've read it seems this would indicate a GPS signal issue. By driving the area we walked I've come to the conclusion that my phone was the more accurate of the two (actually it was very accurate). Has anyone noticed the GPS hardware in the Eris to be inferior? We have a silicone case on her phone, could this be the culprit?

A few items to note:

1) My phone was in my pocket, hers was in her hand (behind moved back and forth on a much more varied arc than mine--against my leg)
2) Her phone is in a silicone case, mine is "naked"
3) Her stride is smaller than mine, but we think we may have measured them incorrectly.

Another thing we thought we may try to tomorrow (along with removing the case) is to adjust the GPS filter from Good to Fair. The polling is still set to 1 second on both devices.

So let me know if you've used this app and have experience with it being accurate or not.

 

[jwm2]

I've not noticed any inferiority with the gps in the phone i had. I used it with google navigation and everything was right on.

 

[Sytar]

Movement of the phone (like being in someones hand) could greatly effect the distance of any application. It is going to see you as speeding up and slowing down over and over.

I would recommend in a pocket, or a pack. just to see how it goes, have both of you do the same thing next time, and then swap for the next (you carry yours, and she has hers in a pocket.)

I think you will find the most consistent info will be from your pocket.

 

[bikeral]

I am not convinced the program uses the GPS to measure speed and distance. If it did, why would our stride length be important? In the stride length description, it says, "stride length that is used to compute distance from step counting." I shortened my stride length a couple of times and distance are pretty close now.

Al

 

[Sytar]

if it is using the stride length, then it is trying to use the accelerometer to judge steps. If you have a smooth hand, and carry your phone, then it may not detect most of the steps, or if it does it may count them wrong.

I have used My Tracks to judge a walk, and it uses the GPS for distance, and the accelerometer to judge when you are moving, or if you are standing still.

 

[jlrosell]

I am not convinced the program uses the GPS to measure speed and distance. If it did, why would our stride length be important? In the stride length description, it says, "stride length that is used to compute distance from step counting." I shortened my stride length a couple of times and distance are pretty close now.

Al

Stride length would still be important for measuring calories burned. It knows your distance, but it needs to know how much work you did to get there in order to estimate your calories burned.

 

[jlrosell]

First this isn't a "should I get a Droid or Eris" thread. I have a Droid and my wife has an Eris.

Recently we found the CardioTrainer app and are using it to track our workouts. Today was the first day we both used our phones, and found that we got wildly different speeds and distances travelled, even though we both walked at the same pace and did the same workout. The time was the same on both phones and the map that was drawn also was the same.

From what I've read it seems this would indicate a GPS signal issue. By driving the area we walked I've come to the conclusion that my phone was the more accurate of the two (actually it was very accurate). Has anyone noticed the GPS hardware in the Eris to be inferior? We have a silicone case on her phone, could this be the culprit?

A few items to note:

1) My phone was in my pocket, hers was in her hand (behind moved back and forth on a much more varied arc than mine--against my leg)
2) Her phone is in a silicone case, mine is "naked"
3) Her stride is smaller than mine, but we think we may have measured them incorrectly.

Another thing we thought we may try to tomorrow (along with removing the case) is to adjust the GPS filter from Good to Fair. The polling is still set to 1 second on both devices.

So let me know if you've used this app and have experience with it being accurate or not.

Not that I know anything about how exactly the app works. I thought it used GPS for everything, but after reading your post it seems like the phone uses GPS only to show you your route on google maps. Otherwise everything else works like a pedometer where your calories, and distance are estimated by your stride length and number of steps.

 

[bilbravo]

Thanks for the feedback everyone. Based on comments I've read and what we've discussed, we've come to the conclusion that the GPS is mainly used to track your route and elevation, but plays a small role in your speed/distance. We are going to put the Eris in my pocket tomorrow along with the Droid and see if things are better.

My wife is about a foot shorter than me, and we also think her stride may have been calculated wrong when we did that setup. Her holding it in her hand and swinging it plus the inaccurate stride setup could be causing the problem.

Will let everyone know what we find!

 

[slickpick]

go into the settings in the cardio trainer app and set the gps to "exact" and to refresh every second. This made a difference for me.

 

[WorkSmart Labs]

Hey bilbravo,

This is something we see a lot of, and I would definitely recommend dropping down your GPS filter settings to Fair, or even Poor. This is a pretty confusing setting that we're considering tweaking to make it easier to deal with, but the lower the filter, the better the chance of getting a signal (setting to Exact means CardioTrainer will ignore everything that doesn't have "exact" accuracy).

Improving the overall accuracy of the app is a huge goal for us, in large part to make up for the problems caused by the GPS reception and hardware. Also, CardioTrainer will only use the pedometer and stride length setting when you're doing a Treadmill Walk or Treadmill Run workout.

Also, our FAQ covers lots of these questions: CardioTrainer Frequently Asked Questions (Help Center)

Let us know if this helps, we want you to have accurate tracking!

Charlie

 

[eghill1125]

I have to agree with Slickpick here even though WorkSmart has said differently on the app as well as here. My wife and I have had differences of a half a mile in calculations from each other... every day using Cardio Trainer. We are doing the same exact routine together at the same time. I had read the help section and had both of our Eris phones set to Fair and 1 second on the GPS.

Yesterday, I followed Slickpicks advice and changed both of ours to exact and we were within .06 difference and I did take a wider turn on our route than she did and would definitely be the difference there. I was extremely happy with our results yesterday by going to exact on the settings. Now it is perfect for our runs.

Our run yesterday showed her at 3.28 Miles and mine at 3.34. Very, very good on accuracy and never before has it been that close to each other. As far as battery, it went from around 6-8% used to about 13%. So basically double, which is about what I would expect and is perfectly acceptable to me.

Cardio Trainer now will become my default app for all of my exercising. Could not be happier now with it showing the accuracy it did yesterday.

 

[bilbravo]

lots of useful stuff

Thanks for your input, and we'll try this (as well as the opposite, as some others have had success with that) to find what works best. As I said, my Droid seemed to be more accurate (judging by what we know to be our actual speed and distance) but the Eris was off by a factor of 1.5. However, the track mapped out on google maps was identical. I think most of the problem may be her wild swinging motion so we will also put the phone in her pocket this time.

That said, this is a great app and I also did not expect to receive support on a "random" forum from the people who make/work on it. Customer Service++.

Thanks to everyone else for suggestions too, namely Slickpick.

 

[Seagull84]

I bike the exact same path to/from work every day and I've noticed the distance can be off by a little less than 1% of the average distance that CardioTrainer records using GPS tracking.

Example:
Biked 2.56 miles to work this morning.
Biked 2.54 miles to work yesterday morning.

The distance averages out to about 2.54 miles, so it has remained consistent over two months of daily commuting.

Maybe I swerved a few times this morning, I try to avoid the pot holes, and sometimes I'm forced onto a sidewalk by a car. I also sometimes turn it on outside my apartment, and sometimes inside.

GPS technology is extremely precise with the ability to locate the device's position within a few feet, so swinging your arm back and forth isn't a big deal. I can't imagine what could cause the wildly different distances you both walked, unless one of you was using GPS and the other was using Wi-Fi to track position. Wi-Fi positioning is horribly inaccurate (can be up to half a mile off), and it requires that it gains a signal from wireless nodes to approximate your position.

 

[bilbravo]

Ok an update. Today I put both phones in my pockets (1 on each side) and we walked. The eris speed was much closer to what the Droid said, but the distance was still off by .5 miles (2.5 vs 2.93). I had her GPS set to "Poor" or "Fair" the whole way. I will try putting it on "Exact" tomorrow.

I noticed that both phones had a "-" for "steps". Perhaps it isn't counting steps and that will make it more accurate?

edit:

The really odd part about this whole thing is that her phone has the exact same track mapped out on Google maps. So the GPS is working well. I just can't figure out what is messing up the distance measurement.

edit2: Seagull... thanks for your input. Both phones are using GPS and show a good signal. Wifi is off on both of them. I suppose she could be getting a location signal from the VZW network, but I think the GPS has 3/4 bars in the corner *and* the map looks right so I think something else is messing with the distance. Your thoughts?

 

[hybridowner]

Try zooming in on the map on each phone. I think you may notice that as you zoom in to a very fine detail, the phone that has the longer distance (Eris?) may have the line jumping back and forth in a zig-zag pattern. A position (the "dots") is being recorded every second, the distance is figured by connecting the "dots" so to speak. If the "dots" are wondering and off by 10 or 15 meters each time, then the distance will be much greater than if each dot was more accurate. Hope that makes sense, but it should be obvious if this is the case if you zoom way in on a portion of the line then pan to follow the line and see what it looks like.

 

[bilbravo]

Try zooming in on the map on each phone. I think you may notice that as you zoom in to a very fine detail, the phone that has the longer distance (Eris?) may have the line jumping back and forth in a zig-zag pattern. A position (the "dots") is being recorded every second, the distance is figured by connecting the "dots" so to speak. If the "dots" are wondering and off by 10 or 15 meters each time, then the distance will be much greater than if each dot was more accurate. Hope that makes sense, but it should be obvious if this is the case if you zoom way in on a portion of the line then pan to follow the line and see what it looks like.

Ok so I did this and I see the Eris does have a more wild line. Much crazier than the Droid. Both are pretty crazy though I guess that's just the nature of the beast. So this appears to be part of the issue, so how do I fix it? I don't like that the next interval is 10 seconds. When you take a 90 degree turn onto another road (we walk in a neighborhood) it could potentially have you cutting across the corner--which could be *less* accurate than I'm getting now? Should I mess with the GPS filter?