Different power sources indoors vs outdoors

General
1

Hi, I’m trying to figure out the best practice for using Xert with two different power sources: an Elite Direto XRT indoors and Garmin Rally XC pedals outdoors.
I recently compared both sources during the same indoor workout. Xert EBC controlled the trainer and used trainer power, while Zwift was running at the same time using Rally pedals for power and cadence, with trainer control disabled.

What confuses me is that live power during intervals sometimes looked different by around 20 W, yet when I exported interval summaries, the longer steady blocks were actually very close overall, with only about 0.4 W average difference between trainer and pedals. But in Xert, the final evaluation of the same ride was very different: the SMART/Xert EBC activity came out at XSS 131, while the Zwift/Rally activity came out at XSS 167.

The ride summaries also differed a lot: SMART/Xert EBC showed 203 W average power, 210 W Equiv Power and 1109 kJ, while Zwift/Rally showed 213 W average power, 225 W Equiv Power and 1182 kJ.

My main concern is this: if I ride outside with the Rally pedals, Xert may give me higher XSS and possibly stronger breakthroughs than I would get indoors on the trainer. That makes me unsure how to interpret BTs and signature changes, especially because indoor workouts can then feel too hard if the outdoor source reads higher.

I do not really want to keep moving the pedals from the outdoor bike to the trainer setup every time. So I’d love to hear how other Xert users handle this: do you use one default power source only, ignore/flag some activities, apply scaling/offset, or just accept that indoor and outdoor XSS will differ?

2

“Power smoothing” and/or ERG mode? Both will give the illusion of steady effort where there is none. Pedals and Trainer might be more in line with both switched off.

Did you push the start of each interval? Might explain the big difference in XSSR.

3

I use Xert EBC in ERG mode for workouts, so there is always a short delay before the trainer settles at the target power.

But I would not say I miss a large part of the interval. It is usually only a short ramp-up, not a case where a significant portion of the interval is spent well below target.

What makes me think there is more going on is the size of the final difference in Xert. For the same session, the SMART/Xert EBC file came out at XSS 131, while the Zwift/Rally file came out at XSS 167.

The ride summaries were also notably different: SMART/Xert EBC showed 203 W average power, 210 W equivalent power and 1109 kJ, while Zwift/Rally showed 213 W average power, 225 W equivalent power and 1182 kJ.

4

I dont have an Elite, but ppl seem to complain about the “power smoothing” setting. Xert will need actual second to second power data, Elite will show flattened power lines instead.

https://elitesrl.zendesk.com/hc/en-us/articles/6866107194258-Power-Smoothing-function-for-interactive-trainers-with-Integrated-Power-Sensor