TEST SIGNALS and DEGRADED EXAMPLES loudspeaker reproduction quality
Including the development of a new objective quality assessment method

A first approach for assessing loudspeaker reproduction quality can be found in this 2015 JAES paper. This paper used ideal reference signals recorded in professional sound studios with an expensive Head And Torso Simulator (HATS) in the objective quality assessment. The new assessment method uses the unprocessed digital reference recordings as ideal reference signals and uses a cheap ZOOM H5studio recording device.

TEST SIGNAL Left Right Inphase Outphase    used to check the loudspeaker setup

6 MUSIC FRAGMENTS used as input TEST SIGNALS in the first subjective JAES test as published in the 2015 JAES paper:
b SOLO INSTRUMENT: Accordion
c CLASSIC ORCHESTRA: Georges Bizet / Carmen
d POP MUSIC: Tracy Chapman / Fast Car
e SOLO INSTRUMENT: Bass guitar
f SOLO INSTRUMENT: Trumpet
g MULTIPLE INSTRUMENTS: Percussion

CONCATENATION OF THE 6 MUSIC FRAGMENTS1, including an anechoic recording of my voice for play back level calibration and validation of POLQA (ITU-T P.863)


6 MUSIC FRAGMENTS used as TEST SIGNALS signals in the second and third subjective JAES test:
b SOLO INSTRUMENT: Harpsichord / Bach Sarabande Partita II, BWV 826
c PERCUSSION: Daniel Cross / The Spinroom
d CHOIR: William Walton / Set Me As a Seal Upon Thine Heart
e ROCK: Stanislav Moryto / Per Uno Solo
f CLASSIC ORCHESTRA: Georg Philipp Telemann / Ach, Herr, Straf Mich Nicht In Deinem Zorn
g CLASSIC ORCHESTRA: Antonin Dvorak / Slavonic Dance Op. 72

CONCATENATION OF THE 6 MUSIC FRAGMENTS2, including an anechoic recording of my voice for play back level calibration and validation of POLQA (ITU-T P.863)


In the new objective quality assessment method any music fragment recorded with a ZOOM H5studio recording device can be used in the loudspeaker reproduction quality assessment. If the signals from the JAES paper are used the procedure below can be used for new recordings:
1) Play the 48kHz/16bit stereo TEST SIGNAL Left Right Inphase Outphase to check the loudspeaker setup
2) Play the 48kHz/16bit CONCATENATION OF THE 6 MUSIC FRAGMENTS1 or play the 48kHz/16bit CONCATENATION OF THE 6 MUSIC FRAGMENTS2
3) Use the first item, the anechoic recording of my voice, to set the play back level correctly, the level at the listening place should be the natural level for a voice, the music fragments will then sound significantly louder, close to the normal play back level in music listening
4) Judge all 7 fragments, thus including the voice on the perceived overall quality with a high quality headphone to get an impression of the maximum achievable reproduction quality, thus calibrating your opinion
5) Judge all 7 fragments, thus including the voice on the perceived overall quality, take the impact of the room into account using the 10 point scale below (a bathroom recording will thus get a lower score than a high quality listening room)
6) Record the signal on the listening place using a Zoom H5studio recorder in standard stereo mode (48kHz/16bit)
7) Split the 7 parts and run each MUSIC FRAGMENT with the LoudspeakerAssessmentTool, the voice recording can be used for validation of POLQA


SUBJECTIVE SCALE TO BE USED IN THE LOUDSPEAKER REPRODUCTION ASSESSMENT
10 best quality possible
9   excellent quality (expensive state of the art loudspeakers in a low reverberation high quality listening room)
8
7   good quality (medium/high quality loudspeakers in a low reverberation living room)
6
5   fair quality (low quality loudspeaker, e.g. laptop, in low reverberation living room)
4
3   poor quality (smartphone quality)
2
1   bad quality (unacceptable low quality)
0   worst quality possible


DEGRADED EXAMPLES HATS recording (2015 JAES paper )
Low quality (4)
High quality (8)

DEGRADED EXAMPLES ZOOM H5studio recording (2026)
Low quality (4)
High quality (8)