View Single Post
Old 04-04-2012, 12:31 PM   #246
Scooby Specialist
Member#: 235338
Join Date: Jan 2010
Chapter/Region: South East
Location: Georgia
2002 Faggin Wagon


Originally Posted by bbarnhill View Post
The smoothing algorithm has been modified as it was exponentially (compound) smoothing before. It would smooth more as the log progressed as it was using the value already smoothed for the average for the next point. This meant that the first value was compound smoothed before leading to exponential smoothing. This was completely wrong. This is the source of the power differences in 1.1.7 and 1.1.8 Just because your numbers dont match a previous version doesnt mean its wrong. Virtual Dyno has bugs in it as does all software (thats an hour long debate there) ... and I removed one and this caused the smoothing to change a bit.

Does that help?
it makes sense, but the older numbers were more inline with what I typically expect from certain setups. Now, the numbers are a fair bit higher and I feel uncomfortable making claims/telling customers output based on these numbers vs the old (while they're just numbers, the old numbers were repeatedly proven to be within 3% of an actual dyno)
Some of us catch enough flack as it is:
* Registered users of the site do not see these ads.
dux10 is offline   Reply With Quote