Australian Cycling Forums - Bicycles Network Australia

Techmoan has done a review of the Sony HDR-AS10 Action Camera, this is the version without the wi-fi.

The biggest problem with the Sony surrounds the mounts, not only the lack of choice but also the shape and design of the supplied case. The fact that it completely blocks off the microphone, can't be tilted up or down, can't be mounted to the side of things, but is also too long and tall to be mounted on the top of a helmet means that the Sony ultimately fails as an action cam. I will be experimenting with finding some way to attach the camera to my bike helmet whilst also recording sound through the mic-in socket…. but I'm really more interested in using this as a general purpose mini camera. For example when I was shooting another review over Christmas it turned out to be the most suitable camera I owned for filming inside the cramped interior of the car when I needed to demonstrate how a Car DVR mounted to the windscreen. I'm also looking forward to getting the Camcorder type mount if/when it is released. This apparently adds both an LCD screen and a much needed tripod hole to the camera. I should mention again the low light recording capabilities which are really amazing and the best I've ever experienced in any camera.

Andrew

On this review if the helmet (side) mount accessory is used then this leaves the microphone free - I think he wants a top mount, with the camera screw mount and a suited mount on the helmet it could be possible with DIY. Even with GoPro it is DIY or using one of the new pro mounts to get it done right.

Back to our regular scheduled programming ...

On my second climb up Bobbin Head this morning, this guy tried to race me. Notice how he spent a bit of effort to re-overtake me and then set up a fast pace.

I don't think he realised I never climb on full power, so I always have reserve in the tank.

I got a nice tow up to the top, and at the end when he got off the saddle to try and cut me loose, I managed to get back in front just by nudging two gears. Didn't even have to get off the saddle.

Christine Tham wrote:Back to our regular scheduled programming ...

On my second climb up Bobbin Head this morning, this guy tried to race me. Notice how he spent a bit of effort to re-overtake me and then set up a fast pace.

I don't think he realised I never climb on full power, so I always have reserve in the tank.

I got a nice tow up to the top, and at the end when he got off the saddle to try and cut me loose, I managed to get back in front just by nudging two gears. Didn't even have to get off the saddle.

What time did you do for the climb?

Christine Tham wrote:Back to our regular scheduled programming ...

On my second climb up Bobbin Head this morning, this guy tried to race me. Notice how he spent a bit of effort to re-overtake me and then set up a fast pace.

I don't think he realised I never climb on full power, so I always have reserve in the tank.

I got a nice tow up to the top, and at the end when he got off the saddle to try and cut me loose, I managed to get back in front just by nudging two gears. Didn't even have to get off the saddle.

the video reminds me of this pic

Christine Tham wrote:For what it's worth, I jumped onto the stationary bike this morning as it was drizzling outside, and reconfirmed that my peak power output (corresponding to max heart rate) is around 300-350 watts, and there is a very high correlation between average power output and heart rate.

Comparing the Strava power numbers against heart rate for my Waterfall ride, it would seem Strava is primarily using the heart rate as a gauge of power. But because the heart rate is a lagging indicator, it adjusts it based on other conditions like speed and incline. Strava also uses the heart rate to determine whether the rider is pedalling or coasting.

Based on that, I would suggest Strava's "accuracy" probably depends on it's assumptions about the rider fitness within age group and gender parameters. Also, the faster your heart responds to changing conditions the more accurate it will be.

good trick i've read about is to find a really popular strava segment - say the mad mile, and look at your placings. Find a person close to your rankings that has a power meter and walla - rough power. Then to go a few steps further filter by age and filter by weight too

jheeno wrote: good trick i've read about is to find a really popular strava segment - say the mad mile, and look at your placings. Find a person close to your rankings that has a power meter and walla - rough power. Then to go a few steps further filter by age and filter by weight too

Unlikely to be accurate unless the person you are comparing with is riding with you at the same time and neither of you are drafting each other.

The heart rate is a very good indicator of average power - that's why the Strava numbers are so heavily influenced by heart rate.

For those who are interested to see what riding on Military Road (Sydney's busiest road) is like - here's what it looks like at 7am in the morning on the stretch from Mosman to North Sydney - not bad, but it is early January during school holidays.

Apologies for the periodic head dipping, I try to limit my speed on this stretch to around 40 km/h - in the video I was doing about 40.3 I think.

This is a safe, comfortable cruising speed for me and leaves me with some reserve to change lanes and overtake buses as required.

Christine Tham wrote:The heart rate is a very good indicator of average power - that's why the Strava numbers are so heavily influenced by heart rate.

OT I know... but are you saying that Strava takes into account your heart rate when working out the power numbers?... so whats it do for people who never use heart rate?.

Christine Tham wrote:

jheeno wrote: good trick i've read about is to find a really popular strava segment - say the mad mile, and look at your placings. Find a person close to your rankings that has a power meter and walla - rough power. Then to go a few steps further filter by age and filter by weight too

Unlikely to be accurate unless the person you are comparing with is riding with you at the same time and neither of you are drafting each other.

The heart rate is a very good indicator of average power - that's why the Strava numbers are so heavily influenced by heart rate.

I'm not sure if heart rate is part of the equation -

https://strava.zendesk.com/entries/2042 ... sure-power

https://strava.zendesk.com/entries/2095 ... lculations

jheeno wrote: I'm not sure if heart rate is part of the equation -

https://strava.zendesk.com/entries/2042 ... sure-power

https://strava.zendesk.com/entries/2095 ... lculations

Those two links are actually quite deceptive, and inconsistent with each other (one says Strava does not take wind into account, the other implies it does).

I suspect Strava has changed their algorithm over time, and those two answers may not reflect the way Strava currently calculates power.

Compare Strava's heart rate and power curves for a ride - you will find striking similarities between the two.

Christine Tham wrote:

jheeno wrote: I'm not sure if heart rate is part of the equation -

https://strava.zendesk.com/entries/2042 ... sure-power

https://strava.zendesk.com/entries/2095 ... lculations

Those two links are actually quite deceptive, and inconsistent with each other (one says Strava does not take wind into account, the other implies it does).

I suspect Strava has changed their algorithm over time, and those two answers may not reflect the way Strava currently calculates power.

Compare Strava's heart rate and power curves for a ride - you will find striking similarities between the two.

ok so anyway with your recordings ... do you just leave it on all the time ?
I have been experimenting with different mounting positions - watching some of your videos -
my guess is you ride with just your hands on the hoods - is that right ? with the angle and how its mounted on your helmet does that mean the shot is crappy if you drop down to the hoods ?

Christine Tham wrote:Those two links are actually quite deceptive, and inconsistent with each other (one says Strava does not take wind into account, the other implies it does).

I suspect Strava has changed their algorithm over time, and those two answers may not reflect the way Strava currently calculates power.

Compare Strava's heart rate and power curves for a ride - you will find striking similarities between the two.

I don't see the inconsistency although I can see with the naming of the variables where there would be confusion. When you see P(wind) or wind resistance it may be better to think of P(air) or air resistance as they aren't talking of wind as such but the power required to overcome the resistance of the air you are travelling through. Wind is relevant in this as the resistance is relational to your air speed so when they say they assume no environmental wind conditions they just take wind out of the equation and use your ground speed as your air speed.

In regards to correlation between HR and power, this may work for you but there is no direct correlation identifiable in my rides. If you want to test it for yourself find a loop with a segment in it. Cruise gently around the loop then hit the segment as hard as you can. Then do the loop again but this time push fairly hard through the whole loop and again go as hard as you can for the segment. You should find that the second attempt produces a higher heart rate but lower power as the heart rate is affected by the work you've done before but you should have had a slower time so the speed used in strava's power calculation drops and along with it the power figure. If that's not enough to convince you then repeat it again as many times as you like alternating between easy recovery loops and hard effort loops. Alternatively find a nice flat bit of road to ride and ride it when there's a nice strong wind then compare downwind and upwind segments and see how the average HR and power figures compare. If the calculation used HR then the increased effort of the headwind should be reflected in those segments but my experience says this is not the case and in fact the prevailing wind has more impact on the calculated power figure than anything else.

Now, back to Action Cam related posting...... Did a bit of Macgyvering tonight and from an old reflector bracket, a screw and a block of rubber from a destroyed bike rack ended up with this..

Will test it out tomorrow but seams sturdy enough, hopefully stops enough vibration and gets a decent image.

I note they are on sale at Costco for $299 earlier this week.

jheeno wrote: ok so anyway with your recordings ... do you just leave it on all the time ?
I have been experimenting with different mounting positions - watching some of your videos -
my guess is you ride with just your hands on the hoods - is that right ? with the angle and how its mounted on your helmet does that mean the shot is crappy if you drop down to the hoods ?

Actually the riding position doesn't matter that much - what really matters is where my head is pointing.

The camera angle doesn't seem to change significantly whether I am on the drops, on the hoods, or even hands free sitting up. Glancing at the speedometer will make a noticeable dip in the video - as you can see from my latest video.

With my recordings, I try to leave it on the whole time, as I primarily bought it as a safety cam. But I've noticed just carrying the camera acts as a bit of a visual deterrent to drivers - they can see that I am recording my ride - maybe that's all I need - a prop.

sblack wrote: I don't see the inconsistency although I can see with the naming of the variables where there would be confusion.

It's actually a lot more than that. Basically, Strava simply doesn't have enough information to be able to utilise the formulae that they are stating.

For example, rolling resistance - Strava would need to know brand of tyres, type of wheel etc. in the same way that the power estimator for Sports Tracks will actually ask the user to input these things. Also to calculate air/wind resistance whatever Strava needs to know the aero profile of the rider.

Since Strava doesn't know any of these things, at best it is taking a wild guess, which means the formulae they give on their web site are deceptive.

What Strava appears to be doing (based on my observation) is modulating the power curve with the heartrate, and combining it with elevation data.

Anyway, hugely off topic - create a new thread if you are interested or PM me.

I asked a friend who is one of the boffins at Strava...

Ian TLL...
Hey Liam, quick question to settle an argument before it starts. When strava works out ( guesstimates ) your power over a segment etc when you don't use a power meter... does it take into account in anyway your heart rate?. I said it was the same calculation whether you used a hr monitor or not.
Cheers Ian.


Liam ...
Hey Ian, you're right - hr is not involved in the power estimates. We use an analytical model that involves rolling resistance, drag, gravity, and acceleration; effort based on hr is not used. The estimates can be pretty decent on long sustained climbs where work to overcome gravity dominates, assuming rider and bike weights are pretty good. Otherwise, it can be difficult to produce good estimates as we make a number of assumptions about rolling resistance, drag, etc. There's a write up here.
https://strava.zendesk.com/entries/2095 ... lculations" onclick="window.open(this.href);return false;

toolonglegs wrote:I asked a friend who is one of the boffins at Strava...

Ian TLL...
Hey Liam, quick question to settle an argument before it starts. When strava works out ( guesstimates ) your power over a segment etc when you don't use a power meter... does it take into account in anyway your heart rate?. I said it was the same calculation whether you used a hr monitor or not.
Cheers Ian.


Liam ...
Hey Ian, you're right - hr is not involved in the power estimates. We use an analytical model that involves rolling resistance, drag, gravity, and acceleration; effort based on hr is not used. The estimates can be pretty decent on long sustained climbs where work to overcome gravity dominates, assuming rider and bike weights are pretty good. Otherwise, it can be difficult to produce good estimates as we make a number of assumptions about rolling resistance, drag, etc. There's a write up here.
https://strava.zendesk.com/entries/2095 ... lculations" onclick="window.open(this.href);return false;

Excellent. Thanks TLL.

toolonglegs wrote:I asked a friend who is one of the boffins at Strava...

Hey Ian,

Can you ask your "friend" whether he was directly responsible for writing the power estimator or was he simply rephrasing the strava link the jheeno previously posted? If he was just rephrasing it's not really an authoritative answer.

Because as I mentioned before there isn't enough information for Strava to actually use the formulae they are quoting, and empirically by comparing power and heartrate curves it would seem the power estimated curve appears to be somewhat modulated by the heartrate curve, which should not happen if Strava was using a pure analytical model.

I agree with the comment that the analytical model that Strava claims to be using would have the best accuracy in climbs where air resistance is less of a factor. But in general the analytical model does not work well in a bunch ride or on a segment like the Mad Mile, which was the original point in this thread.

Anyway, this is very off topic. Can I suggest creating a separate thread, or using PM if you want to continue?

" "

toolonglegs wrote:" "

+

greyhoundtom wrote:

toolonglegs wrote:" "

+

Hey if an engineer at Strava isn't authoritative enough then nothing is ever going to be ( but then I expected nothing less )... especially seeing that it is a pretty simple question as to whether they use HR or not... the answer is NOT.
Over and Out.

toolonglegs wrote: Hey if an engineer at Strava isn't authoritative enough then nothing is ever going to be ( but then I expected nothing less )... especially seeing that it is a pretty simple question as to whether they use HR or not... the answer is NOT.
Over and Out.

It sounds like you got the answer you wanted all along, which is great for you.

Are you able to send me the email of your friend? I am interested to know how Strava can differentiate between coasting and pedalling (which they imply in their description of how they derive average power on their web site) using the analytical formulae they claim they are using. And how their power curve seem to fluctuate even when the speed and incline remains relatively constant, which according to their formulae should not happen.

Gosh Christine you're fast! Respect

arkle

arkle wrote:Gosh Christine you're fast! Respect

Actually I am not, but thanks anyway. Last year has been a bad year for me, and I'm still struggling to return to former form.

For example, last Sat on Mona Vale Rd, I was struggling to get to mid 40s when I know I used to blast it at close to 50. I did have one guy who was happy to suck my wheel though. It was so hot I could hardly breathe.