Speech intelligibility is an important aspect to be considered in a Public Address (PA) system. A PA system in a public transit system announces
train arrival/ departure, general information, and emergency information and instructions to passengers. Poor speaker characteristics, inappropriate
sound levels, reverberations, and delays can degrade sound quality and make even an expensive sound system unintelligible and useless.
Addressing these issues at the design stage through proper planning and acoustic simulation studies would improve a PA systemís overall
performance and customer satisfaction, and minimize the need for expensive post-installation modifications.
A Case Study
One of our public transit customers operating a large metro train system in the U.S. had a PA system with inadequate intelligibility and
insufficient coverage, inviting adverse feedback from the public. GL was tasked to conduct a comprehensive study to (i) assess the existing
system and (ii) recommend steps to improve speech intelligibility and coverage within public areas at select above-ground metro stations.
GLís study would serve as a model for implementation at other metro stations. Our study consisted of testing and analysis, subjective and
objective, to assess intelligibility and current coverage.
GLís engineers used Speech Transmission Index for Public Address (STI-PA) standard to assess speech intelligibility objectively. Enhanced
Acoustic Simulator for Engineers (EASE) software was used to conduct simulation studies and recommend steps to improve intelligibility and
increase coverage area. For subjective analysis, we listened to test announcements at various locations within the PA coverage area and
rated the sound quality as Satisfactory, Needs Improvement, or Unsatisfactory. We made our recommendations based on a need to meet or
exceed NFPA 72/2006 requirements for a Common Intelligibility Score of 0.7.
GLís Solution, Technical Guidelines and Test Plans Development
GL conducted an audio survey of the PA system at two metro stations. The customer considered these two stations to be representative
of other above-ground stations. We conducted initial site surveys of the two stations that:
- Detailed physical measurements of the station architecture.
- Took note of the materials used in the station architecture.
After the initial site surveys, we conducted comprehensive tests and simulation studies to:
- Assess audio quality and speech intelligibility of the existing PA systems.
- Measure ambient noise levels.
- Measure STI-PA.
- Perform extensive acoustic modeling for deriving optimal configurations.
- Recommend steps to expand the coverage and improve speech intelligibility.
- Recommend specifications and vendors for speakers.
GLís engineers created three dimensional computer models using the above inputs. With the 3-D models and ambient noise measurements,
our engineers simulated the acoustic environment of the areas requiring PA coverage. Results of simulations studies were analyzed to determine
optimal speaker type, speaker locations, and speaker power levels for the PA system.
Our studies also simulated sound pressure levels and Rapid Speech Transmission Index (RaSTI) scores. Using the simulation results, GL also
recommended speaker tap settings, orientation, and volume adjustment in order to improve intelligibility in existing coverage areas as well as to
provide coverage in additional areas. Based on our studies, GL recommended a number of steps to improve audio quality within the stations to
meet or exceed NFPA 72/2006 requirements for a CIS of 0.7. We also developed a detailed test plan that our customer can use to assess new
PA systems that will be installed at other metro stations.
GLís study and recommendations would greatly improve the speech intelligibility and increase coverage area within the public transitís metro
stations. The quality improvement would benefit the patrons leading to increased satisfaction.