Outdoor Cycle & Battery Endurance Test

[exxxviii]

I did my first outdoor cycle today. Overall, the watch did very well. Its distance was only 0.6% different from my Garmin and my bike computer. That is definitely a win. Average HR matched exactly over almost 2 hours. The only thing suspect is calorie calculation. The AW calculated 1,500 calories total, or about 814 calories per hour. That seems high, given my HR only averaged 128 and it was not an A-level effort. The Garmin calculated about 1,100 calories total, or 580 calories per hour. I am much more inclined to believe the Garmin's metrics here.

Where the AW totally loses it, and I think this is the root of the calorie calculation algorithm problem, is that the AW had 209 BMR calories (total calories - active calories). This extrapolates to 2,722 BMR calories per day. That is beyond bogus.

This also concluded a battery demand test today. I ran for about 90 minutes this morning and then biked for about 120 minutes in the afternoon, following a full day of normal use in between. I started at 5:00 AM off charger, and I put it back on the charger at 6:30 PM at about 10% battery. So, the watch easily supported 3-1/2 hours of workout plus another 10 hours of regular use. That is more than enough power for any endurance workout or race I will ever do with the watch.

 

[BlueMoon63]

I did my first outdoor cycle today. Overall, the watch did very well. Its distance was only 0.6% different from my Garmin and my bike computer. That is definitely a win. Average HR matched exactly over almost 2 hours. The only thing suspect is calorie calculation. The AW calculated 1,500 calories total, or about 814 calories per hour. That seems high, given my HR only averaged 128 and it was not an A-level effort. The Garmin calculated about 1,100 calories total, or 580 calories per hour. I am much more inclined to believe the Garmin's metrics here.

Where the AW totally loses it, and I think this is the root of the calorie calculation algorithm problem, is that the AW had 209 BMR calories (total calories - active calories). This extrapolates to 2,722 BMR calories per day. That is beyond bogus.

This also concluded a battery demand test today. I ran for about 90 minutes this morning and then biked for about 120 minutes in the afternoon, following a full day of normal use in between. I started at 5:00 AM off charger, and I put it back on the charger at 6:30 PM at about 10% battery. So, the watch easily supported 3-1/2 hours of workout plus another 10 hours of regular use. That is more than enough power for any endurance workout or race I will ever do with the watch.

I really appreciate all your testing and I agree that 814 calories seems high with what you said. The Garmin seems right. It's interesting that we are taking about too high when some of your other tests have the watch too low. Def says the algorithms are all over the place.

You are also very active!! I've also had great success with the battery. Used it for 4+ hours golfing starting at 95% and finished with 40% with the workout other running the whole time. App said I burned 1500+ calories or roughly 380 per hour at an average heart rate of 115. In all they time, the heart rate was working about 98% of the time. Had about 20 times where it dropped to 50 but always came right back.

I'm not following your BMR information and that is just my ignorance. Today I only have 650 active calories and my total is close to 2600 total calories with 4 hours to go. That puts me right on track to hit my BMR 2300 plus 650 active for 2950 at midnight.

I just don't really follow too closely and don't understand your numbers.

 

[exxxviii]

I calculated the BMR calories from just the workout, not the day's totals. For example, open your golf Woorkout Other from the phone and look up the active and total calories for that workout. Subtract, and you get the BMR calories for that workout. Then, extrapolate that to a day (workout BMR calories / workout duration hours * 24) to see what you get.

 

[BlueMoon63]

I calculated the BMR calories from just the workout, not the day's totals. For example, open your golf Woorkout Other from the phone and look up the active and total calories for that workout. Subtract, and you get the BMR calories for that workout. Then, extrapolate that to a day (workout BMR calories / workout duration hours * 24) to see what you get.

I'll take my simple workout today that was 59 minutes at an average heart rate of 105. Just wasn't feeling it today.

Active calories was 321 and total was 436 which equals 115 BMR calories for the workout. So basically for one hour it is 115*24 = 2760 which is a little low.

Yesterday, I worked out for 60 minutes with active calories of 433 and total of 560. Avg heart rate of 120.

127*24 = approx 3148 so I get your math

What you are saying is the gap between active calories and total calories should never change if all the total calories is doing is taking your total daily BMR/24 * length of workout. Correct? You are right.

But where is the problem? Do they calculate a higher BMR average during workouts based on heart rate average? The only difference in my workouts was avg heart rate and type of workout. They could formulate a difference on workout type that raises your BMR numbers because of calories burned in subsequent hours after working out?

Who knows, I'm just guessing.

 

[iPhysicist]

Post your Apple Watch work out/ Activity rings!

 

[exxxviii]

What you are saying is the gap between active calories and total calories should never change if all the total calories is doing is taking your total daily BMR/24 * length of workout.

Yes, I expected that the BMR calories for a workout should always be a constant, affected only by the length of time of the workout.

But where is the problem? Do they calculate a higher BMR average during workouts based on heart rate average?

The definition of BMR is the calories you would burn anyway, independent of the activity. So I would not have expected the AW to change that number during an activity, because it seems that those calories should be credited as active calories. Then, to make the day's total work out, the AW must be subtracting the activity BMR calories from the rest of the day when we are not active. (The AW will have the same total BMR calories for a day, independent of how much activity we do.)

I'm just curious about the calculations and algorithms. For most men, the daily BMR calories ought to be around 1,800 calories, plus or minus probably 200 to 400, depending on age, weight, activity, genetics, etc. And, I would expect that things that calculate calorie would treat BMR as a constant, once set. But for the AW, the BMR seems to be a variable number. Yet it sums to a constant at the end of the day, but during a workout, it seems to change to a different number. I would not have expected that.

 

[iPhysicist]

Yes, I expected that the BMR calories for a workout should always be a constant, affected only by the length of time of the workout.

The definition of BMR is the calories you would burn anyway, independent of the activity. So I would not have expected the AW to change that number during an activity, because it seems that those calories should be credited as active calories. Then, to make the day's total work out, the AW must be subtracting the activity BMR calories from the rest of the day when we are not active. (The AW will have the same total BMR calories for a day, independent of how much activity we do.)

I'm just curious about the calculations and algorithms. For most men, the daily BMR calories ought to be around 1,800 calories, plus or minus probably 200 to 400, depending on age, weight, activity, genetics, etc. And, I would expect that things that calculate calorie would treat BMR as a constant, once set. But for the AW, the BMR seems to be a variable number. Yet it sums to a constant at the end of the day, but during a workout, it seems to change to a different number. I would not have expected that.

I posted a link to an other thread above your last post. It was no race but a solo tour. My calories should be at least close to correct.
I took this link to check http://www.sparkpeople.com/resource/calories_burned.asp?exercise=4
But this was watchOS 2.0 not 2.1

 

[BlueMoon63]

Yes, I expected that the BMR calories for a workout should always be a constant, affected only by the length of time of the workout.

The definition of BMR is the calories you would burn anyway, independent of the activity. So I would not have expected the AW to change that number during an activity, because it seems that those calories should be credited as active calories. Then, to make the day's total work out, the AW must be subtracting the activity BMR calories from the rest of the day when we are not active. (The AW will have the same total BMR calories for a day, independent of how much activity we do.)

I'm just curious about the calculations and algorithms. For most men, the daily BMR calories ought to be around 1,800 calories, plus or minus probably 200 to 400, depending on age, weight, activity, genetics, etc. And, I would expect that things that calculate calorie would treat BMR as a constant, once set. But for the AW, the BMR seems to be a variable number. Yet it sums to a constant at the end of the day, but during a workout, it seems to change to a different number. I would not have expected that.

I have to agree with you. I went back and looked at the last couple of months and the more active calories I had, the larger the gap between the numbers. It would always work out to be close to my expected number at the end. They never seem to match up day to day even if I don't change my weight. I do change my weight a lot but nothing big.

Some days I have 1000 active calories and it ends up being about 2900-3000 calories and another day I have 600 active calories and it comes in with a total of 2900.

I was hoping Apple with all their "fitness" hires, they knew something we didn't.
[doublepost=1454250120][/doublepost]

I posted a link to an other thread above your last post. It was no race but a solo tour. My calories should be at least close to correct.
I took this link to check http://www.sparkpeople.com/resource/calories_burned.asp?exercise=4
But this was watchOS 2.0 not 2.1

I am ignorant to fitness calorie burns and what is correct, but I tried that link and it can't be right or I did t see everything you can set.

It asked for my weight and length of time working out and it gave me the calorie burn. It seemed way too high. That assumes that all people weighing 215 would burn the same amount of calories going at one speed. It can vary by 50%. I just don't know what is right.

I assume I didn't do something right.

 

[iPhysicist]

I am ignorant to fitness calorie burns and what is correct, but I tried that link and it can't be right or I did t see everything you can set.

It asked for my weight and length of time working out and it gave me the calorie burn. It seemed way too high. That assumes that all people weighing 215 would burn the same amount of calories going at one speed. It can vary by 50%. I just don't know what is right.

I assume I didn't do something right.

This particular link only works if you were bicycling with an average of 16 to 19 mph (wich is quite a wide spread, i know) and have some data backing up the calculation. My weight is 158 pound and my exercise was roughly 260 minutes. The value is close enough to the apple watch calculation. So I assume the watch is right (in my specific scenario). Generally calculating the calorie burn just from the heart rate, height and weight alone is prone to errors.

 

[exxxviii]

I posted a link to an other thread above your last post. It was no race but a solo tour. My calories should be at least close to correct.
I took this link to check http://www.sparkpeople.com/resource/calories_burned.asp?exercise=4
But this was watchOS 2.0 not 2.1

This actually looks like it could be in the ballpark. And everybody's body is very different. You were averaging 18 MPH, which is a pretty stout pace. Additionally, your BMR calories from that ride works out to 92 calories/hour (2,210/day) which is also in the right ballpark. I cannot remember when Apple played with the calorie calculations, if it was 2.0 or 2.1.

The AW problem is that the calorie calculation seems to be OK for some people. But how do you know if you are one of the "some people?" You have to analyze the watch's output across multiple type of events, and compare that to other sources, like online calculators, Garmin devices (if you have one), treadmills and ellipticals etc. (if they collect enough data from you), and then give the AW the sniff test.

 

[BlueMoon63]

This actually looks like it could be in the ballpark. And everybody's body is very different. You were averaging 18 MPH, which is a pretty stout pace. Additionally, your BMR calories from that ride works out to 92 calories/hour (2,210/day) which is also in the right ballpark. I cannot remember when Apple played with the calorie calculations, if it was 2.0 or 2.1.

The AW problem is that the calorie calculation seems to be OK for some people. But how do you know if you are one of the "some people?" You have to analyze the watch's output across multiple type of events, and compare that to other sources, like online calculators, Garmin devices (if you have one), treadmills and ellipticals etc. (if they collect enough data from you), and then give the AW the sniff test.

I for one really appreciate your efforts and everyone's efforts to try and figure this out. You would think with all the "hires" Apple made in the health and fitness environment, they would have this down. Unfortunately, that doesn't seem to be the case.

 

[exxxviii]

My hope is that as we call out some of the odd or inconsistent behaviors, the AW community amplifies it and then eventually Apple acknowledges the issues and fixes them. If anyone else, like Fitbit, Nike+, Garmin, Microsoft, etc. had issues like these in one of their activity trackers, they would (had have been) roasted. Apple has somehow escaped a lot of that, and the AW is a lesser product because of it. Apple has the resources to make the AW great, but they chose not to with this first release.

 

[nicho]

remember when apple fixed resting calories?

I think they forgot to change the equivalent calculation in the "active/total" calories per of workouts.

Before watch os 2 my resting calories were around 2800 a day. Now they are 2000 a day or so. I'm only 1kg different in weight compared to the period I just checked.

Every workout I've checked since the beginning pegs my resting calories for that workout at somewhere between 1.9 and 2 per minute. Extrapolate that and it's 2800 a day...

 

[BlueMoon63]

My hope is that as we call out some of the odd or inconsistent behaviors, the AW community amplifies it and then eventually Apple acknowledges the issues and fixes them. If anyone else, like Fitbit, Nike+, Garmin, Microsoft, etc. had issues like these in one of their activity trackers, they would (had have been) roasted. Apple has somehow escaped a lot of that, and the AW is a lesser product because of it. Apple has the resources to make the AW great, but they chose not to with this first release.

You are so right about the competitors. Fitbit is in a huge lawsuit over the inaccuracy of their products heart rate monitoring when used during heavy exercise. Apparently, if you believe the suiters, the Fitbit's accuracy is horrible when exercisers are working out in the high range.
[doublepost=1454348115][/doublepost]

remember when apple fixed resting calories?

I think they forgot to change the equivalent calculation in the "active/total" calories per of workouts.

Before watch os 2 my resting calories were around 2800 a day. Now they are 2000 a day or so. I'm only 1kg different in weight compared to the period I just checked.

Every workout I've checked since the beginning pegs my resting calories for that workout at somewhere between 1.9 and 2 per minute. Extrapolate that and it's 2800 a day...

I've gone back several months and my numbers vary from accurate to off by 300 or more. I've seen it before and just ignored. Now we have a reason and I hope this gets fixed. I can tell you that it isn't fixed on 2.2 beta 2.

 

[exxxviii]

You are so right about the competitors. Fitbit is in a huge lawsuit over the inaccuracy of their products heart rate monitoring when used during heavy exercise. Apparently, if you believe the suiters, the Fitbit's accuracy is horrible when exercisers are working out in the high range.

Funny, I was just talking about that with a friend at lunch. In my anecdotal experience, if there truly is a problem, Apple's HR problem is far worse than Fitbit's. I run with a friend who has a FB Charge HR, and his has been more trustworthy than my AW for HR readings-- on runs, especially in the cold, I my AW will either have no reading or something way out of reality while his Charge HR is dead on. And others I know with Charge HRs are consistent with my friend's. Apple's only advantage here is that they have fewer HR-enabled units in the market compared to FB, so the class action opportunity is smaller with Apple. But, I would not be surprised at all if Apple gets hit in a year or two.

 

[exxxviii]

I did another mega workout battery stress test today. I had 4 hours and 45 minutes of workout time in two workouts. Watch on the wrist at 8:00 AM. After I was done at about 3:00 PM, the watch was at 13% battery. I can live with that. (The watch is back on the charger so I can finish the day with it.)

 

[128KMac]

Your experiences tend to mirror mine. Good for about 4 hours of activity, HR all over the place, and calories a bit optimistic. At least once a year I do day long bike rides, so the watch doesn't survive the day. With my Garmin Edge 800 I can tape a power tube to the bike frame and plug in the Garmin for a day's worth of power. No way to do that with the AW :-(

 

[exxxviii]

HR all over the place, and calories a bit optimistic.

You are right; what a surprise. I think this is the first post I have ever read that says the AW calories are optimistic. My AW calories have come in low to very low for me in running, walking, and indoor cycling on a trainer with power. I simply quit looking at calories a while ago and assumed they were low for every activity. But, I just looked back at my three outdoor rides I have with the AW, and the AW total calories is higher than my Garmin 910XT calories on every ride. (I do not have power on the bike, just the trainer.)

That totally begs the question-- how did Apple screw up the calorie algorithms so badly that indoor cycling is almost 50% low, indoor running is around 25% low, outdoor running is about 75% low, but outdoor cycling is 10% high? I just hope they fix it.

 

[zhenya]

I have been using an external HR strap for the last month or so, and I'm happy to report that it drastically improves the battery life while running. I get a pretty consistent 10% per hour drain this way with lots of runs in the 2-3 hour range. And of course my HR info is spot-on.

 

[BarracksSi]

I have been using an external HR strap for the last month or so, and I'm happy to report that it drastically improves the battery life while running. I get a pretty consistent 10% per hour drain this way with lots of runs in the 2-3 hour range. And of course my HR info is spot-on.

Which strap? I may consider one if I ever get back into heavy training again.

 

[exxxviii]

Which strap?

I can give a huge shout out to the Scosche Rhythm+. After nearly a decade of running with various Garmin chest straps, the Rhythm+ is a breath of fresh air.

 

[zhenya]

Which strap? I may consider one if I ever get back into heavy training again.

I am using the Wahoo Tickr. Inexpensive, bluetooth and ANT+ means it can work with either my iPhone or my Garmin or even both at the same time.

 

[128KMac]

I am using the Wahoo Tickr. Inexpensive, bluetooth and ANT+ means it can work with either my iPhone or my Garmin or even both at the same time.

So - the AW and/or iPhone pick up chest strap info instead of the HR off of the AW?

 

[BarracksSi]

I am using the Wahoo Tickr. Inexpensive, bluetooth and ANT+ means it can work with either my iPhone or my Garmin or even both at the same time.

Good to hear.

So many reviews and comments about the AW and workout tracking only talk about its optical sensor, it's rare to hear how it does with a chest strap.

Being able to check HR without a chest strap is a selling point for any LED-based sensor. It's a feature my wife finds very attractive, too. But I've never put too much trust into any optical HR device; there are too many variables which can change the readability of blood flow in the wrist.

 

[BarracksSi]

So - the AW and/or iPhone pick up chest strap info instead of the HR off of the AW?

Yup.