|
@@ -0,0 +1,16 @@
|
|
|
+In this recording, 'dsp_compensated_buff_errs' had a bandpassfilter (dsp) and it's lag is compensated for by Cheetah:
|
|
|
+-DspDelayCompensation Enabled
|
|
|
+-DspFilterDelay_µs 3984
|
|
|
+
|
|
|
+while the 'no_dsp_buff_errs' had no dsp:
|
|
|
+-DspDelayCompensation Enabled
|
|
|
+-DspFilterDelay_µs 0
|
|
|
+
|
|
|
+The result is that 'dsp_compensated_buff_errs' has slightl more datapoints than 'no_dsp_buff_errs'. This gives a break in current
|
|
|
+neo as it requires the data to have the same length.
|
|
|
+
|
|
|
+The other problem with this data are RingBuffer errors (search the logfile), which are now treated by making different segments
|
|
|
+per gap. Both channels with have different gap times, also causing a break in current io.
|
|
|
+
|
|
|
+Thijs Ruikes
|
|
|
+t.r.ruikes@uva.nl
|