Page 1 of 1

Ostrich 2.0 hardware map tracing

Posted: Thu Jan 10, 2008 8:08 pm
by SSS
Not sure how many others apart from Matt have hardware map tracing on working for Ostrich 2.0, but what i've noticed that it is no where near as "smooth" as map tracing via consult.

Is there any reason behind this? ie just need some more time to develop it further?

I've also found it to be a little buggy, ie at idle the trace momentarily jumped to a large rpm value and dropped back to where it was supposed to be in the map.

Posted: Thu Jan 10, 2008 10:49 pm
by Matt
Consult tracing > Hardware tracing

Yeah dont expect it to be as good as the consult maptracing

The consult is a direct read of where the ECU is reading, the hardware maptrace is a sample of an area of memory addresses being accessed, which we filter with the maps you are looking at and then display the cursor position at that time, until the next cursor position gets reported

Sometimes the memory addresses being sampled seem to jump around a fair bit. You will also usually get it bouncing around four cells that the ECU uses to interpolate from

The only thing I can do to develop it further is add more sampling per second to get more addresses. But this wont help where the moates may not sample all addresses correctly (I think sometimes it doesnt pick up a valid address and this gets reported)

Posted: Fri Jan 11, 2008 8:02 pm
by SSS
No dramas, i'll stick with consult anyways.