Appendix B: Algorithms
Appendices
AĆ20
W(t)
is the sampled waveform
W
^
(
t
)
is the continuous function obtained by linear interpolation of
W(t)
A
and
B
are numbers between 0.0 and
RecordLength
–1.0
If
A
and
B
are integers, then:
ŕ
B
A
W
^
(
t
)
dt
+
s
ȍ
B
*
1
i
+
A
W
(
i
)
)
W
(
i
)
1
)
2
where
s
is the sample interval.
Similarly,
ŕ
B
A
(
W
(
t
)
)
2
dt
is approximated by
ŕ
B
A
ǒ
W
^
(
t
)
Ǔ
2
dt
where:
W(t)
is the sampled waveform
W
^
(
t
)
is the continuous function obtained by linear interpolation of
W(t)
A
and
B
are numbers between 0.0 and
RecordLength
–1.0
If
A
and
B
are integers, then:
ŕ
B
A
ǒ
W
^
(
t
)
Ǔ
2
dt
+
s
ȍ
B
*
1
i
+
A
(
W
(
i
)
)
2
)
W
(
i
)
W
(
i
)
1
)
)
(
W
(
i
)
1
)
)
2
3
where
s
is the sample interval.
Time measurements on envelope waveforms must be treated differently from
time measurements on other waveforms, because envelope waveforms
contain so many apparent crossings. Unless otherwise noted, envelope
waveforms use either the minima or the maxima (but not both), determined in
the following manner:
1. Step through the waveform from
Start
to
End
until the sample min and
max pair
DO NOT
straddle
MidRef
.
2. If the pair >
MidRef
, use the minima, else use maxima.
If all pairs straddle
MidRef
, use maxima. See Figure A-4.
The Burst Width measurement always uses both maxima and minima to
determine crossings.
Measurements on
Envelope Waveforms
To Next Page
Loading...
Need help?
General
