Technical data

ManualsBrandsASTi ManualsMusical equipmentTelestra

ASTi ACE Studio Components Reference Guide Rev. M DOC-01-TELAS-CRG-4

2.3. AutoDred

Summary: The AutoDred component conducts a series of five tests to test the

system’s speaker setup.

Description: The AutoDred component starts the first test by sending pink noise to

all speakers in the system setup. Test 2 sends a sinewave at 100Hz out each

individual speaker in the setup. Test 3 sends a sinewave at 1,000Hz out each

individual speaker in the setup. Test 4 sends a sinewave at 10,000Hz out each

individual speaker in the setup. Test 5 plays pink noise out each individual speaker.

If any of the tests fail, an Error Code will display the channel number that failed

and the test number. If all tests pass, it confirms that all speakers are working

properly in the system’s setup.

Use the Manual Enable input to perform a specific test.

For the initial automated setup procedure, set the Test Enable and Setup Enable

inputs to True. Set Manual Enable to False. This will run the setup tests for every

channel. Setup will take approximately 2 minutes per channel, for 16 channels

approximately 32 minutes.

Important: If possible, nobody should be in the room/simulator with the speakers

during testing. If you must be in the same room during testing it is imperative that

you remain still with no movement during testing.

Setup Procedure

This procedure assumes a prebuilt AutoDRED mode exists.

1. Install layout, open the AutoDRED model in load viewer.

2. Adjust AutoDRED gains (NoiseGain and FilteredNoiseGain) down to be care-

ful at first. NoiseGain controls the level of the flat pink noise test. Filtered-

NoiseGain controls the levels of the 3 other noise tests, which involve

bandpass filtered pink noise.

3. Turn SetupEnable to TRUE, then TestEnable to TRUE to do a setup run. You

should hear it cycle through four noises on each speaker. Wait for it to finish,

then put SetupEnable and TestEnable back to FALSE.

4. SSH into the Target, view (using “less” or the editor of your choice) /tmp/auto-

dred.dat, which has a nice summary of setup levels and distances above the

noise floor for each test. The goal is for most tests to be >10 dB above the

noise floor. This won't be possible on some tests, like high frequency noise

through the subwoofers, and that is okay. Those cases will be considered “don't

cares”.

5. Adjust NoiseGain and FilteredNoiseGain, re-run the setup, and then reopen

autodred.dat, until most tests reach the 10 dB mark. If one particular test (e.g.

speaker 3, test 4) needs to be louder, the GainTable feature can be used to boost

or cut that specific volume. Otherwise, GainTable does not need to be used.

6. When you are satisfied that most tests are 10 dB above the noise floor, create

two TableXYs in the mathplan. Name one something like

“DRED_Ref_Levels”, and the other “DRED_Thresholds”.

7. Enter the levels from /tmp/autodred.dat into the DRED_Ref_Levels table. See

the example tables for how to structure the table. The test number goes along

the top and the speaker number goes down the side.

8. Fill in the DRED_Thresholds table to have the value “3” (dB) in all entries that

correspond to tests that were 10 dB or more above the noise floor. For test

cases that were less than 10 dB above the noise floor, enter in a high threshold

(for example 100) to make the case a “don't care”.

9. Click “notify Target” in the Mathplan window to push the tables to the Target.

10. Select the two tables in the AutoDRED component - filling in “LevelTable”

and “ThresholdTable”.

11. Run the test a few times and observe the level differences on each test. Gener-

ally tests should be within 2 dB of the setup run.