Tektronix WFM8200 Technical Reference page 18

Specifications
Table 1: SDI input waveform vertical characteristics (cont.)
Characteristic
Step
Response,
Preshoot
Typical
SD
HD
Overshoot
SD
HD
Ringing
SD
HD
Pulse
Response,
Baseline
Typical
Ringing
SD
HD
Tilt Field Rate
Line Rate
Off Screen Recovery
2
Performance requirement
0.1%
0.1%
0.1% variation in baseline of a 5 MHz
modulated pulse when positioned anywhere
on screen at any gain setting.
WVR8200 and WVR8300 Specifications and Performance Verification
Reference information
Sine-squared bars
≤ 0.3% peak (2T5 bar)
≤ 0.5% peak (2T30 bar)
(2T60 bar for 148.5 MHz 1080p formats.)
≤ 0.3% peak (2T5 bar)
≤ 0.5% peak (2T30 bar)
(2T60 bar for 148.5 MHz 1080p formats.)
≤ 0.8% peak-peak (2T5 bar)
≤ 0.8% peak-peak (2T30 bar)
(2T60 pulse for 148.5 MHz 1080p formats.)
Most of the error seen on the display comes
from the inherent ringing in the digital data.
The response of the rasterizer is close
to the theoretical limit of a perfect sinx/x
reconstruction filter.
Blackman pulse
≤ 0.6% peak-peak (2T5)
≤ 0.7% peak-peak (2T30)
(2T60 pulse for 148.5 MHz 1080p formats.)
Pulse-to-bar ratio 0.995:1 to 1.005:1 on
appropriate Sine Squared or Blackman 2T
pulse.
A sine-squared pulse near Nyquist is not
band-limited and so inherently has ringing
much larger than the waveform rasterizer
filter. A three term Blackman pulse with the
same HAD has much less inherent ringing,
so it is a better choice for most testing. See
Digital to Analog Conversion, Data and Filter
Requirements, SMPTE Journal Mar 1995,
Vol. 104, Fibush, Baker, Penny.