C2f.        PARK LEVEL / ADDITIVE CONSTANT         (For D.S.P. processors only)

Theory:

In an analog Pro Logic decoder the Combiner is constructed of Voltage Controlled Amplifiers. For the VCAs to
function properly the control voltages need to stay within a specific range.  If the control voltage drops below this
range, the VCAs will not be able to provide a steering action.  

The  level of the control voltages to the Combiner are proportional to the levels of Lt and Rt.  The fact that an
analog processor goes un-steered when the total Lt/Rt signal level drops below a threshold is used by producers to
create a very defuse sound field.  An example could be that of the sound of wind during a quiet portion of a movie.

DSP processors do not have the limitation presented by analog VCAs.  It is possible for DSPs to provide steering
action below a level intended by the movie producer.  In order to prevent un-desired steering action during low level
passages in a movie the concept of a “Park Level” was derived.  The intent is to cause the DSP processor to go un-
steered below a certain threshold.  This threshold has been determined to be 45 dB below ROL or 65 dB below
digital full scale in the left or right channel.


Test Method:

In order to determine the “Park Level” it is necessary to monitor steering action while adjusting the level of Lt and
Rt.  The “Slow/Fast” steering Threshold test conditions provide an indication of steering action.  

For this test, a tone burst generator can be used to modulate the level of the 2.5 kHz signal being sent to the
decoder under test.  The “ON level” of the tone burst generator is set to match the level of the 1.0 kHz signal
generator measured at Lt/Rt.  The “OFF level” of the tone burst generator is set to –10 dB.  

In addition to the 2 stepped attenuators used for the Auto Balance tests, an adjustable stereo level control with a
40 dB range is desirable (see appendix C3).

Select the desired test setup as detailed in the “Slow/Fast Threshold” test, above.

1.  Set the tone burst generator to it’s ON state (high level).

2.  Normalize the monitored point for 0 dBr.

3.  Start the tone burst generator cycling.

4.  While monitoring the filtered steering action adjust the level of Lt/Rt until there is no synchronous change in the
level of the 1 kHz signal with the tone burst generator.

5.  Continue reducing the level of Lt/Rt an additional 10 dB to assure that there is still a signal present at the
measured output.

6.  Increase the level of Lt/Rt until the level change of the 1 kHz tone increases to 3 dB.

7.  Set the tone burst generator to it’s ON state (high level).

8.  Measure the level at the monitored output.  This value is the “Park Level”.

9.  Repeat the test for the other 3 test conditions.

For the Left channel test, vary the level of the Left Lt/Rt signal.

For the Right channel test, vary the level of the Right Lt/Rt signal.

For the Center channel test, vary the level of the Center Lt/Rt signal.

For the Surround channel test, vary the level of the Surround Lt/Rt signal.
Example plot of park Level measurement showing a “Park Level” of approximately –45 dBr.
      Below is a Test Procedure that I created.  This text is from an Implementations Level,  Pro Logic Test
Procedure Manual that I wrote for use by our licensees and our staff.   This test is to determine the lower level
threshold of the steering control path ways of a Dolby Pro Logic Surround Decoder IC.  I explained the reason
behind this test in the "Theory" section below.
Test Procedures
Functional Block Diagram of a Dolby Pro Logic Decoder.
      A partial schematic of a constant power pan,  matrix surround test signal encoder.  We
called it a "Steering Box".  I constructed more than a half dozen of these test fixtures.  The
Sine Cosine potentiometer is used to generate the constant power pan.  They cost a little
over $200 each.
Above is the test setup used for measuring the movement of the "sliding bands",  of Dolby B,C and S-Type processors,  
back-to-back with a lab reference processor,  under complex signal conditions.

The text below is from a Noise Reduction Test Procedure Manual that I put together.
c) Sliding band response

Purpose: To check for proper tracking of the sliding band by applying a high level fixed tone and a low level sweeping
tone. The high level tone will cause the sliding band to remain in one fixed position while the low level tone will indicate its
actual position.

Test Set: The test setup includes two signal generators, one which will provide the fixed tone and one which will provide
the swept tone. The swept tone generator must also be able to track a 10 Hz band pass filter in conjunction with a chart
recorder to plot the output signal. In addition, a variable attenuator is necessary to attenuate the fixed tone from 0 to -20
dB, and the usual meter and oscilloscope are used to monitor and calibrate the test setup.

Procedure:

1 ) Set the generator levels independently to provide Dolby Level at the encoder and calibrate the encoder and decoder
as before (400 Hz, NR off).

2) Attenuate the swept tone to -40 dB and sum the two signal generators (note that a passive sum is possible as long as
the proper impedances are preserved). It is important that the proper reference levels are maintained for each signal
generator.


3) Begin with the fixed tone at 0 dB and sweep the -40 dB tone from 20 Hz to 20 kHz, recording the output through the
BPF on the chart recorder.

4) Repeat with the fixed tone attenuated 5, 10, and 20 dB. The device should function within tolerance at all supply
voltages within its specified range and at any fixed tone attenuation.

Tolerance: +/-1.5 dB deviation for B-type, +/-2 dB for C-type.