Virtual shifting


(Obi dave Kenobe) #1

How about a request for virtual shifting. A rider could virtually shift to an easier gear or harder one and this would cause incremental adjustments to the trainer intensity.

To extend to this idea, a user could add the parameters of their drivetrain into Zwift such that if a rider is could enter a compact crankset or a full sized crank set with an 11, 10, 9 , or even an 8 speed cassette — what ever is installed their real bike.


(Nigel Doyle) #2

I’m obviously missing something here. What’s the point when you have actual gears on your bike?

And heck, let’s get slowing down for corners when descending working before something a bit out there like this.


(Obi dave Kenobe) #3

True, you do have real gears on the bike. I was inially thinking for those who have electronic wireless shifters for deeper integration, but those probably work on a propitiatory protocol. Anyway, as it looks right now, Zwift adjusts your speed from your speed sensor then adjusts it according to your entered weight, power output (if available), and gradient, and a assumed wheel circumference. I suppose knowing a giving speed, wheel size, and maybe cadence, Zwift could figure out what gear your riding in. And with that, along with power data, Zwift could use that to adjust your speed when your climbing, so your not just spinning faster and faster in order to go faster.


(Nigel Doyle) #4

If you have a smart trainer or power meter Zwift doesn’t use your speed sensor. Speed sensor info is only used on dumb trainers to estimate power based on a known power curve for your specific dumb trainer.


(Obi dave Kenobe) #5

At the expense of accuracy. Understand that if you start with a known wheel circumference, the speed is simply rate = distance / time, where distance is wheel circumference. Then you can apply any adjustments such as grade, rider’s weight, or cadence to approximate virtual speed. Power (or watts) is the amount of output energy (energy generated by your effort). Though a power-to-weight ratio can be useful to an extent. By only using power, the math must be worked backwards. Thus, the lighter the rider weighs the less power it will take for that rider to travel at X mph at X% gradient.

You can find numerous comments of people saying that their Zwift speed or is off then their bike computer which is based off the wheel or satellites (which has its own issues). Understand that adjusting gears on the bike and virtually well have an impact the riders work load, thus their output wattage.


(Karissa Minn) #6

There is already a way to adjust your resistance level during workout mode (though it’s harder to find now than it used to be). It would be nice to have that option somewhere for free riding, in case the base level of resistance seems to be too low or high. :slight_smile:

Just a note that data from a speed sensor, run through Zwift’s virtual power calculations, is not going to be as reliable as data from a smart trainer or power meter. There may be times when virtual power is more accurate - I know that’s hard to believe for some people, haha - but too many variables can throw it off too easily.

You definitely should never have a speed sensor paired with Zwift when using a smart trainer, because Zwift will not have a proper power curve stored for the smart trainer, meaning the virtual power calculation will be incorrect. (I think there are very few exceptions - classic trainers which can “become” smart trainers with a hardware addition may have a power curve that is selectable within Zwift.)

This article may help with that. :slight_smile:


(Obi dave Kenobe) #7

There is nothing wrong about using wheel speed for speed and distance, really. Unless the scale of Zwift routes is off — I mean if Zwift miles do not match real miles 1-to-1.

Wheel speed is more accurate when you know the circumference of the wheel than mapping a speed to a power value (aka power curve). Once again, the virtual map would need be to made to scale. I get it, the physics get complicated fast.

I noticed last night while Zwifting along, the power Zwift displays seems low while the displayed speed seems high when compare to data crunched from the real world. I guess in the real world, we can factor in the wind for further analysis, where as Zwift has none.

Yes, many users do notice the discrepancies in the assumptions Zwift makes.

So, how is native BLE support for Windows 10 coming along?:smile:


(Danny Boyd) #8

Zwift is not a real world simulator and is not going to be the same as in real world for numerous factors.
It does however, a really good job or allowing ALL riders to ride together on a more equal footing, and uses WPK to calculate speed, gives a draft, etc.