Hacker News Reader

Comment by HorizonXP | Hacker News Reader

Great summary of your findings.

I've been wearing a Fitbit Charge HR since last June, and I definitely do not reach 10k steps daily. However, I've noticed that on days I hit the gym and do weight lifting, my caloric burn easily tops 3300, while on days where I'm lazy and just working on my laptop, I barely break 2100.

As a rule, I typically do not do any cardiovascular exercise. I think, I do not know for sure, the Fitbit is better suited for measuring cardiovascular activities, so I'm not sure how or why it's adjusting my caloric burn on my gym days. Maybe it's looking at my heart rate and extrapolating?

Anyway, I'd be very much interested in some kind of fitness tracker that can provide better estimates on caloric burn. That coupled with calorie counting/recording of meals is what will really help me shed this weight.

View on HN

I've always thought the killer healthcare app would be some sort of way to automatically track calorie and nutritional macro intake without manual data entry. No idea how that would actually work, though.

This startup is trying to do it, at least for liquids, https://www.myvessyl.com/vessyl/

This seems the most ridiculous - internet of shit - thing; when in the video he turns the "Vessyl" and it reads beer ... just look at your can, it's written here. Tracking liquid calories is easy, you drink your soda can and the number of kcal is indicated clearly. You drink water from the tap and ... you consume 0 kcal.

If you are like me and drink a lot of water each day, you will get really really tired of measuring it and keeping track of it. Making it automatic is important - it eliminates the psychological component of trying to beat your numbers. A product like this is not just about tracking calories but also how much you are drinking.

Very few people have any need to closely track their water consumption unless they have a particular condition, though, and this product seems to be marketed to a fitness/lifestyle market, not health care.

Great idea. Perhaps an AI that recognizes what we eat from a cell phone photo and tracks nutrition.

Sounds possible at least for prepackaged and restaurant/fast food.

I doubt it would be feasible:

1) food appearance gets changed a lot when cooked and plated

2) a photo would not be enough to recover the volume of food in metric scale

Good points, but did you miss the discussion nearby? It seems someone already made a plate with cameras with object recognition to identify food and 3 scales to determine the volume.

> http://www.fastcoexist.com/3045810/world-changing-ideas/this...

Yep, missed that!

I'm pretty sure someone has just released something like this. A handheld scanner that will tell you the calories/macros for your plate of food.

It was a one off cost of circa £150, then £15 or so a month ongoing.

Apologies but can't be arsed to find the link on mobile.

Was it this miniature near-infrared spectrometer?

> The SCiO, a handheld "pocket molecular sensor" available now on Kickstarter, gives you the data on your food.

http://www.fastcoexist.com/3030342/fund-this/this-tiny-devic...

Also found this plate with 3 scales and cameras that uses object recognition algorithms to identify food:

> This Smart Plate Tells You How Many Calories You're Eating—And Whether You're Eating Too Fast

> The ultimate in portion control: a plate that yells at you.

http://www.fastcoexist.com/3045810/world-changing-ideas/this...

> Maybe it's looking at my heart rate and extrapolating?

Yes. At least that's what I hope/assume mine is doing.

> Anyway, I'd be very much interested in some kind of fitness tracker that can provide better estimates on caloric burn.

The 2nd most accurate formula involves weight and heart rate. http://fitness.stackexchange.com/questions/12435/is-it-possi...

Your Charge HR knows both of those, so I would hope it uses them. That's the whole reason I bought one in the first place.

It looks like measuring VO2max is even more accurate. But that's very annoying and invasive to measure as it involves breathing through a tube.

I do not believe Fitbit uses heart rate to track calories. If it uses heart rate to track calories, it would be able to work well with cycling exercises, but it does not. From fitbit website: "Fitbit trackers are optimized for walking, running, and general lifestyle activities but are not recommended for tracking cycling"[0]. Sounds like it uses the step counter and not heart rate to estimate calorie burn along with your weight and age etc.

The Charge HR heart rate functionality could be useful to keep my heart rate in a specific range during cardio. However, it doesn't seem like it's good for that either[1].

[0]https://help.fitbit.com/articles/en_US/Help_article/Does-the...

[1]http://arstechnica.com/tech-policy/2016/01/lawsuit-claims-fi...

When I use it for boxing, the calorie burn graph seems to very closely follow the heart rate graph. But only if the tracker was in exercise mode (longpress the button).

As for the article about inaccuracies. Their manual clearly states that during exercises that involve a lot of forearm flexing and vigorous arm movements the tracker will be inaccurate.

And I am sceptical that people wouldn't be able to tell from how they're feeling that they are at a dangerously high heart rate. But I guess it's possible.

The highest number my fitbit ever reported was about 170 (sprinting) so it's definitely cspable of measuring high heart rates.

I wonder why it's not recommended for cycling then. Cycling is my main exercise. Very little arm movement on a bike.

My old Fuelband used to register quite a bit of activity while cycling, surprisingly. It would register about half as many Nikefuel per hour as running, but still way way way way above, say, walking, let alone being seated.

I haven't cycled much since I got my fitbit charge hr.

Here is an interesting study regarding energy consumption http://www.cell.com/current-biology/abstract/S0960-9822(15)0... Tldr days of high level of exercise does not linearly correlate to high level of energy usage. The body can somewhere(brain?) spend less energi to compensate the exercise.

> As a rule, I typically do not do any cardiovascular exercise

Do you mind explaining why? Time constraints? Health issue? Genuinely curious...

In my case (not op) I don't belong to a traditional gym and go rock climbing several times a week. I don't really care to go for a run but I love climbing. Also my knees are in bad shape because I am both tall and heavy which means running/biking/anything where I regularly move my knees hurts. Climbing is nice because most of the time unless I am trying to dyno (jump) to a hold I only use my feet to keep myself against the wall or to statically take my weight while I move my hands.

I guess a better term would be, you are not doing much `aerobic' exercise. High intensity sports like rock climbing or lifting definitely train your cardiovascular system. In a different way than `cardio', of course.

If your goal is to bulk up, then you need to consume more calories than you burn. Cardio burns calories, meaning you need to eat more food to hit that caloric surplus.

Google Fit tracks the repetitive movements of gym exercises as steps. Maybe Fitbit does the same.

Probably they would also track sex as steps. Now I want to see somebody reach 10.000 steps!

I've seen sex get recognized as "Sport" by the Fitbit "auto-detect" feature. Also does count steps.

Wearing your fitbit then seems weird.

I'm pretty sure it looks at heart rate. In the app, you can see a graph of heart rate over time when you log an exercise. I have the surge, not sure how the charge works.

Anyway, you should really get some cardio. That heart is gonna keep you alive man. Treat it well.

I do not know what you do in the gym but the delta between resting and lifting seems way too high for lifting and even a moderate warm up.

100% agree, this is why I doubt the number. I'm moving more, since I'm not in the house, and I'm walking around the city and taking the bus.

No warmups, just straight lifting weights (bench press, deadlifts, squats, plus a few accessory exercises), and stretching.

There is the delayed caloric burn that occurs from weight lifting, but I doubt the Fitbit is accounting for that.

Reply on news.ycombinator.com