Audio Test

ISDCF Audio Test Content

ISDCF has created a Digital Cinema Package of audio test content to help understand the impact of non-perf screens (LED) on audio playback. The intent was for content to reveal known concerns for a non-perf screen. However, it can be used to test any cinema sound system.

PRIOR TO RUNNING THE TEST, GO TO 19MIN 43 SEC (LAST SEGMENT). THIS HAS VOICE SLATES AND SMPTE ST 2095-1 PINK NOISE FOR EACH CHANNEL. CHECK THAT THE SYSTEM IS CORRECTLY PLAYING BACK EACH CHANNEL AND IS CORRECTLY CALIBRATED.

The content has 16 segments. Each segment contains specific content to help analyze the sound system. Below are some comments on what to listen for and what others have seen for other playback environments. 

 1.    0.0 Introduction

Female voice introduces the test content.

2.   0m 18sec Center Localization Test 

This test is designed to ensure that sound emanating from the center channel is perceived directly from the center of the screen. In a traditional perf screen system, there is a center loudspeaker that reproduces the center channel and anchors the sound to the center of the screen. Usually, the center loudspeaker high frequency horn is set ⅔ of the way up the screen, so depending on the nature of the sound, it may be perceived slightly above or slightly below the exact center depending on the loudspeaker bin arrangement. This traditional arrangement is designed to more closely align with the actors’ heads when they are full screen.

With non-perf screens, there is no center loudspeaker. To compensate for this, there are generally several other loudspeakers above and below the screen and/or others that are specifically reflected off the screen that are used to give the listener the impression the sound is emanating from the center of the screen. 

Several different sounds are played one after the other to test the perceived location. 

• Male voice
• Female voice
• Claps
• Cowbell 
• Tambourine 
• Short gunshot
• Long gunshot

All sounds should be localized to the center of the screen and not have the image “smeared” up or down. Room “echos” should be minimal and not interfere with the perceived center image.

3.   1m 8sec Soundstage Localization Test

This tests that sounds panned to specific locations across the screen can be precisely localized and that their timbre stays consistent in each panned position. The sounds start center and then are panned to the left 20 degrees, 45 degrees, 60 degrees, hard left and lastly off screen, which is evenly split between left and left surround. The same is then done panning to the right, starting in the center and then right 20 degrees, 45 degrees, 60 degrees, hard right and lastly off screen, which is evenly split between right and right surround. 

A single handclap, Female Voice and Male voice sequences are employed and panned as noted.

Each sound should be perceived entirely on the screen and not above or below it. The last sound in the sequence should feel just off to the side of the screen and not on the screen or in the room. The timbre of each sound should match in each panned position.

A traditional perf screen system will reproduce the sounds in a single loudspeaker or a combination of the LCR  and surround speakers to attain the perceived location. Any changes in timbre as the sound pans is generally due to inconsistency in the loudspeaker voicings, which can then be addressed. 

A non-perf screen will utilize a number of loudspeakers above and below the screen as well as specific reflections to localize the sound, which presents a greater challenge in maintaining timbre consistency as sounds are panned. Check from various listening positions that each sound localizes correctly and maintains the same timbre, as the sweet spot may sound good but other listening positions may reveal issues.

• 1min 13 sec   Single Handclap Pan (front screen positioned from center to offscreen left, center to offscreen right)

Each clap should be perceived entirely on the screen and not above or below it. The last clap in the sequence should feel just off to the side of the screen and not on the screen or in the room. The timbre of each clap should match in each panned position.

• 1min 44sec Female Voice Pan (front screen positioned from center to offscreen left, center to offscreen right)

Her voice should be perceived entirely on the screen and not above or below it. The last line in the sequence should feel just off to the side of the screen and not on the screen or in the room. The timbre of her voice should match in each panned position.

• 3min 46sec Male Voice Pan (front screen positioned from center to full left, center to full right)

His voice should be perceived entirely on the screen and not above or below it. The last line in the sequence should feel just off to the side of the screen and not on the screen or in the room. The timbre of his voice should match in each panned position.

4.   5min 38sec Surround Localization Test

This tests errant reflections in the room, especially from the screen, that may cause sounds that emanate only in the surrounds to be incorrectly perceived as coming from the screen or elsewhere in the room, rather than from the intended surround location.

A pair of handclaps two seconds apart are played sequentially in Lss, Rss, Lrs, Rrs, between Lss and Lrs (left rear corner), between Lrs and Rrs (straight back), and between Rss and Rls (right rear corner).

Screen reflections can interfere with perceived surround sound locating. While perf screens can exhibit reflections, a non-perf screen is solid and can exhibit loud reflections that compete with the level coming out of the surround loudspeaker and interfere with location perception. Perf screens may also have a solid wall behind them, which if not acoustically treated, can create loud reflections as well.

The claps’ perceived location should be exactly where they are played (see screen diagram during the sequence). If the room has errant reflections (“slap”) off the screen or elsewhere, this will be very obvious. Check this from various listening positions in the room.

5.   6min 26sec Walkaround Center Image Test 

As noted in the Center Localization Test, while a perf screen system has a center loudspeaker to anchor the sound to the center, a non-perf screen does not. Therefore, a correctly perceived center image is often dependent on the listener’s position in the room. While it may be good in the “sweet spot”, it is a challenge to be consistent for all parts of the listening area in a non-perf system.

The sequence plays the same line three times from the center channel, spoken alternately by a Female and Male voice. Move around the room to check that the image remains in the center for both voices.

The voice image location should be perceived as coming from the center of the screen regardless of the listening position. Check that it does not shift side to side or up and down as you move around the room, and that the timbre is maintained at any listening position.

6.   7min 37sec Center Channel Timbre Test

As noted in earlier tests, a perf-screen system has a center loudspeaker but a non-perf screen system does not, and relies on several loudspeakers and/or specific reflections to create a center image for material played from the center channel. In addition to the location tests above, this segment tests the timbre consistency of the center channel reproduction, which can be challenging in a non-perf screen system.

The piano has a very specific and recognizable harmonic structure, and therefore is chosen as the source for this timbre test. Four piano samples are played in sequence.

Listen to each piano sample and ensure that the timbre is consistent. Move around the room and check that the timbre remains the same regardless of the listening position. Check also that the piano location does not appear to “smear” up or down or side to side. The harmonics should line up vertically and horizontally.

Due to the nature of non-perf system center channel reproduction, these can noticeably change from piano sample to piano sample, and especially with position. 

7.   9min 4sec Pink Noise Timbre Test

Pink noise has a very specific character that is familiar to technicians, and is often used not only for calibration, but to check the consistency of loudspeaker voicing and timbre. SMPTE ST 2095-1 Wideband Pink Noise is used for this test.

The pink noise is panned across the screen left to right and then clockwise around the room to check that its perceived character stays the same.

• 9min 25sec  Pink noise panned left to right across the screen

In a perf-screen system, check that the character remains consistent as the sound pans between loudspeakers located behind the screen. Variation may indicate inconsistency of loudspeaker voicing.

In a non-perf screen system, check that the character remains consistent and the panning appears smooth. There should be no big peaks or valleys in the level as it pans.

• 9min 57 sec  Pink noise starting at center and panning clockwise around the room

In a perf-screen system, check that the character remains consistent as the sound pans between loudspeakers located behind the screen and the surround loudspeakers, and through all of the surround loudspeakers. Note that due to the nature of 7.1 system calibration, the volume level will be 3dB lower in the surrounds compared to the screen. This is normal and should not be considered a defect.

In non-perf screen systems, again check that the character remains consistent between screen, surround and surround to surround. The same 3dB lower level in the surrounds applies and is normal. Check that there are no big peaks or valleys in the sound level as it moves.

8.   10min 49sec Soundtrack Reproduction Test

Having checked specific issues in the previous tests, the system should now be evaluated using actual movie soundtrack material that is familiar.

Several clips from “The Patriot” are provided for this section, courtesy of Sony Pictures Entertainment 

This soundtrack is exemplary for its natural sounding voices and very specific, high-quality Foley, sound effects, ambiences and room treatments. The music is purely orchestral and is extremely well recorded, with the imaging of an actual orchestra setup. Throughout the clips, check that these are correctly reproduced from multiple listening positions.

If comparing a perf-screen system to a non-perf screen system, note any differences in voice quality, imaging, panning and room treatments.

• 11min 2 sec “Ride to Town”

This is a classic traveling sequence with big orchestral music, with very specific sounds and locations for the many different places that are shown on the way. Once in town, there is an intimate outdoor conversation that reveals natural voices and panning of offstage sounds. Check that the music and voice character holds up as well as the specific locations of sounds. 

• 12min 44sec “Beginning of Council Meeting”

This takes place in a very live room with the expected reverberation. There are crowd sounds at the beginning, and then individual voices. Check that the voice character is consistent and especially that the room treatment on the voices and Foley sounds correct from screen to surrounds. You should feel like you are in the room with the characters.

• 13min 25sec “Captain Martin Speaks of Being a Patriot” 

This takes place in the same live room, featuring a very well-recorded Captain Martin addressing the Council. Check the voice timbre is consistent and correctly perceived in the center with no smearing. The room treatments on the voices vary subtly with camera angle and should sound natural from screen to surrounds. The variations in treatment should be easily perceived and sound correct.

• 14min 24sec “Wood Scene-Preparing to Shoot Redcoats”

This is an outdoor scene and thus should have no perception of voice treatment. The scene opens with running through the forest, with excellent Foley, panning and sound effects. Check that the panning sounds correct as they run. The outdoor ambience should sound very natural, like you are in the middle of the forest. When they get into position, there is some very intimate conversation between the actors that should sound very present and natural while the forest ambience.

• 15min 50sec  Church Militia Recruiting-Woman Speaks To Congregation

The scene takes place in a reverberant church with various voice actors at the beginning. Check the voices sound natural and the room treatment changes with camera angle. Mid scene, a young woman emphatically addresses the congregation. Check that her voice character is natural, not too strident, and the treatment appropriate as she addresses various members of the congregation.

• 17min 07sec “Ambush Redcoats” Full 7.1 mix

This is the full mix of an outdoor scene with very specific sounds of bullets flying and riccos. Soldiers are heard yelling from various locations, some close and some further away, including in the surrounding forest. The panning shifts with the camera angle in a natural way. Through it all, there is a very specific forest ambience that envelops and makes you feel you are in the middle of it. Check that the panning of the bullets and voices is consistent and correct and that there is no smearing of the image.

• 18min 21 sec “Ambush Redcoats” Surrounds only

This is the same clip as above, but with the screen channels muted. Only the surrounds are played. Check the sounds are perceived to come only from the surrounds and not from the screen and respect the intended panning per camera angle.

END OF TEST MATERIAL CONTENT. 

9.  19min 43sec 7.1 Channel Identification and Calibration Pink Noise” 

THIS SHOULD BE THE FIRST THING PLAYED BY THE THEATER TECHNICIAN, PRIOR TO CONDUCTING THE TEST MATERIAL. IT IS USED TO CHECK THAT EACH CHANNEL IS CORRECTLY REPRODUCED AND THAT THE SPL LEVEL IS CORRECT.

Each channel has a voice slate followed by SMPTE ST 2095-1 Pink Noise.

The calibrated SPL levels should be as follows:

Left (L)-85 dBc
Right (R)-85 dBc
Center (C)-85 dBc
LFE-(LFE) approximately 91 dBc
Left Side Surround (Lss)-82 dBc
Right Side Surround (Rss)-82 dBc
Left Rear Surround (Lrs)-82 dBc
Right Rear Surround (Rrs)-82 dBc