Direto vs Sigeyi readings

Hi, after finally getting the Sigeyi PM installed on my Spark, today i finally put it on the Direto to check how the readings are.
Direto recorded on TR android app, Sigeyi on a Garmin 530.
I was not expecting an exact match, but the readings are quite off: for example, a few minutes at 226W avg and 234W max on Direto, gave me 239W avg and 250W max on PM.
Not really nice at all… The plan was to do more outside workouts with the PM, but on higher intensity intervals, this is too much difference.
What do you think is the problem?
I can offset the PM to make it closer to the Direto, yet if it’s the Direto to be off then i’ll have a wrong dataset both inside and outside.
Thanks for any input and suggestions!

Just under 6% gap.

I’d check (or re-calibrate both) and then see what the gap is. I know the Elite is +/1.5% (according to Elite) which which is maybe 4w max at 226W.

This isn’t uncommon at all. My AXO, NGeco, Assiomas and Kickr are consistently 6-10% higher than my Neo 2T and Rally pedals. This is with lots of testing, calibration, latest firmware versions, etc. It’s frustrating.

Ultimately in your case I think it’s fine to scale the PM to match the Direto. You’ll have a consistent, precise data set indoors and out, doesn’t really matter if it’s accurate or not for training purposes.

The only real time I can think of where accuracy matters is Zwift racing (for me anyway, since I want to make sure I’m not cheating others or myself with power that’s higher or lower than it really is).

Thanks for the replies.
The PM supposedly calibrate itself and the Direto i actually calibrated it the other day, tightened the belt as to make the calibration value matching exactly the one written underneath.
I did contact the shop where i got the PM and next time i can pay a visit they will install a pair of Favero to test with them too; we’ll see how that goes.
If indeed the Direto is underestimating power, back in my Zwift days i could have gotten some better results for the sufferig i went through! :joy:

What with the Sigeyi PM measuring at the crank and the trainer measuring at the hub, a couple of % drop from the front to the back is to be expected. Add in the precision of both devices, and it’s not a crazy difference.

In these debates, you have two choices: you arbitrate with other measurement sources (another trainer, another PM, etc), or you pick one and stick with it (presumably the PM). Power readings are slightly irrelevant as long as they are repeatable.

1 Like

Indeed; on the one hand, i will check with the store’s Favero pedals in order to better determine where a correct measurement probably lies. On the other, in any event (since i can’t tweak the Direto’s reading) i will end up offsetting the PM to closely match the trainer and just be consistent between the two - what really counts.
It’s just a bit discouraging to see how easy it is for measuring devices to give or take more than 5% and how hard it is to improve half as much from a block to the next :laughing:

It’s much more common to use the PM as the reference if you also use this bike on the road.

1 Like

I did go through that when I bought Assiomas, and they (slightly) disagreed with my trainer (Flux S). Since I bought the Assiomas to bring power-based training outdoors, it stands to reason that I picked that as “the truth”. I use the pedals as the power source on the trainer.


In my years training with power meters I’ve used the following devices:

  • Stages Gen1 and 2
  • Garmin Vector 3S (2 different pairs)
  • Powertap P1S
  • Kickr Core
  • Tacx Neo 2t

Stages and Vector kinda read the same, Kickr, Powertap and Neo read lower and within the same ball park. Currently my FTP is 250ish on the Neo and 260ish per Vector (firmware up to date, battery ok, calibration done etc). When using DCrainmaker comparison tool (recorded on TR and Garmin just like you did), it shows the delta is larger on lower power numbers and the average starts to converge on higher values (250W+). Note the Vector will read all over the place and the Neo tends to vary less.
I wrote the above to illustrate that different instruments will yield different readings, what you’re experiencing is not uncommon. If you want to use the same tool for indoor and outdoor training I suggest using power match when riding the trainer. And if you’re curious about your “true power” I would just get the average between the 2 numbers haha…

1 Like

You have it all wrong: according to the numbers you give, your two power meters agree perfectly.

The power readings on your trainer should be lower than on your crank-based power meter, because your crank-based power meter does not see drive train losses whereas your trainer sees the power after drive train losses. On a clean drive train, a standard 1x or 2x setup is about 95–96 % efficient.

Let’s assume for the sake of argument your drive train is 96 % efficient. Then 96 % of 234 W is 225 W, which is spot on. Yes, 225 W vs. 226 W, but you also need to take into account the accuracy of your power meters, which is probably about 1.5 % for your crank-based power meter and 1 % for your Direto if memory serves.

Let’s try that again with your other power numbers: 96 % of 250 W is 240 W, again perfectly matching the 239 W you saw on your trainer.

Even if you repeated these computations with 95 %, you’d just end up on the lower end, and IMHO still within the accuracies of both power meters.

1 Like

Thanks; i do have it right in that the readings are indeed different. The wrong part is to think one of them might be off, hence me asking the question.
Your answer pretty much clears it up and it’s good to know both devices are well within spec.
I will then probably end up offsetting the PM to show something closer to the power at the hub, since the outside workouts are tailored to those readings from the trainer; if i leave it as it is, outside workouts will be easier than prescribed.

Glad we could help. Typically, there are two paths forward: you can use power match, i. e. your Direto will use the power numbers of your crank-based power meter. Or you only use your Direto for indoor training. Then you will have to retest and accept the lower numbers. However, then indoor and outdoor FTP will differ and your TSS aren’t computed correctly for outdoor rides.

I have a Suito and a Quark DZero. Even though I know the Suito (like all Elite trainers) has a reputation for being very accurate, I just default to my DZero (which is also known to be very reliable and accurate).

1 Like

Nice, I had forgotten about drive train losses. I always wondered if the direct trainers compensate for that, i.e. would they add a few % to the reading. I’m sure Tacx, Wahoo, Elite, etc won’t comment haha…

I think @GPLama would disagree. “However, measured losses are never greater than a few watts (<2%) of TOTAL power using a clean drivetrain when cross-chaining using 52/28.” Look at any of his tests and you’ll see good trainers matching reliable PMs within a couple of watts.

Smart trainers try to take into account drivetrain loss and match as closely to your pedals/spider/real power output as they can. Problem is they just aren’t awesome at measuring power, nor are they consistent from unit to unit. I like to think the highly-rated trainers mostly do a good job, and that’s probably true, but the fact is a significant percentage of trainers are off by a significant margin, even at that high end of the market. It’s a lottery whether you’ll get an accurate one or not.

If memory serves, @dcrainmaker and @GPLama would often comment how certain power meters read relative to another, because of where in the drive train chain they measure power.

But perhaps you are right and trainer manufacturers aim for emulating power at the crank rather than giving you the power at the cassette or rear wheel. Although it seems to me because of the inherently large and variable range of drive train efficiencies (factors I can think of: gear combos, cross chaining, cleanliness of the drive train, differences between manufacturers) would lead to much larger systematic errors and thus, I don’t see how manufacturers could achieve the claimed accuracies.

I don’t think that’s any different from regular power meters, and distinguishes a good power meter/trainer from a bad one. It is just a matter of consistency in production and proper calibration at the factory.

1 Like

Agreed for the most part, but I do think that the good PM companies are better at this (makes sense: they have one job, while trainers have to do a lot more).

Here’s an example of three devices that are getting it “right”

From the @GPLama Rally pedals video. My experience has my Rally pedals (XC200) about 6-11% lower than my Kickr and my NGeco (and Sigeyi AXO). This is with lots of test runs, updated firmware, temperature stability, etc.

I could find a similar screenshot for the Neo, but my Neo 2T is also significantly lower than my Assioma pedals (which match my Kickr, NGeco and AXO nicely).

Not to pick on Garmin or Tacx specifically, that was just my experience with lots of testing on my devices. Plenty of reports out there of Kickrs that run low or high, people buying multiple trainers until they get one that runs high for Zwift racing, etc. So, it’s a lottery when it comes to accuracy but I trust the tried and true companies like Favero and Power2Max more than I do the trainer companies. They’ll get there eventually but no surprise it’s taking time.

1 Like

Yeah I remember that too. Especially with the old hub PMs, that was often brought up. I’ve also seen @dcrainmaker mention drivetrain loss in a reply to someone commenting about their trainer reading low by a handful of percent. But I do think the trainer companies are taking the modern approach of attempting to get to a more accurate reading of actual power that the cyclist is generating. It’s an algorithm anyway for most of them, so they might as well account for drivetrain loss and make people/reviewers happy by matching their accurate power meters. And I am almost certain the companies are sending tested units that are known to be accurate to reviewers whenever possible. It could be the difference between a successful product launch that makes a killing and mediocre sales, of course they are.

1 Like

AFAIK Elite has a very good reputation when it comes to power accuracy, so I don’t think this is universal. However, in my mind, if you have a power meter on the bike, you should use that. And conversely, there are quite a few companies that are struggling with power meters, Shimano comes to mind, according to @GPLama’s quick test of their latest two-sided power meter, they still haven’t cracked it.

I’d say that producing accurate power meters (either on bike or in trainers) on a large scale is not easy, and when I bought my new road bike, I made it a point to get a Quark power meter. SRAM’s Quark power meters have a very good reputation for being reliable, and so far, mine has been perfectly boring. AFAIK together with the Assiomas and the Elite Diretos, they are the default benchmark power meters that @dcrainmaker and @GPLama rely on.

I didn’t know of that shift, I’ll take note of that, though. So in this case the OP might actually want to play with the slope of his trainer’s power meter (if he can).

1 Like

I don’t think you can, which is maybe one of the reasons high-level Zwift racing requires a trainer to be used as the power source (with a dual recording backup).

One of my racing friends mentioned the Direto being a popular choice for cheaters who go trainer shopping looking for higher watts, but that’s total hearsay. And clearly not universal if at all true, as OP is seeing with their Direto.

I think it’s insane that the companies are all claiming 1-2% power accuracy when you can so easily find two trainers of the same model that diverge wildly. It’s a mess out there!

I would assume they go for a match with a clean (like, lab clean) drivetrain. If you’re pushing tons of watts through your pedals with a gunked-up mess of a chain in your 39/11 and your trainer reads low that’s on you!