Guildhall Measurements
Capturing the acoustic of the Guildhall is a lengthy process: planning, organising, recording, measuring, and processing takes some time.
In practical terms, t-he planning process starts with booking the space, planning the measurements and organising equipment. A historically important space, such as the Guildhall, which is now a multipurpose space used for a variety of meetings and events, needs to be available for an extended period of time including times when it is possible to take acoustic recordings without disturbance from extraneous background noise.
In order to understand the planning of the acoustic measurements, two words which are commonly used need to be explained :
Source: A source is typically the generator of sound, this could be a musical instrument or in our context a singer.
Receiver: A receiver takes sound energy and converts it to electrical energy we can record and measure, in the context of our research this manifests itself as a listener/audience member, with measurements taken by an array of specific microphones to simulate an audience.
The interactive app we have produced lets you listen to choral music as if you are standing in the Guildhall, and gives you the ability to move yourself and the singers within the space. Unlike in real-life, the app allows you to move the performers around, and even give them odd instructions such as “Go and stand in that corner and face the wall”.
For more information about the Guildhall and acoustics follow the buttons below.
Planning Measurements
Since the app allows both listeners and singers to move around the (virtual) space we needed to measure many source and receiver positions in the Guildhall. Having a number of source and receiver points means that we can interpolate between these points smoothly to give the illusion of moving around the space.
Standard grid
The image shows a view from the front of the array, four speakers can be seen extending down the length of the Guildhall.
A standard grid of source and receiver positions was chosen to enable generous coverage of the space.
Stage diamond
The image shows a view of the diamond formation, four speakers creating each point of the diamond with an additional three points irrelevant to the measurement.
A stage diamond formation was implemented to reflect the layout of the choir members for the first Architexture project, where a microphone was placed in the middle of a circle of 10 singers as show aside.
With a combination of four loudspeaker sources in a tight diamond formation and a microphone receiver in the centre we are able to recreate the style of the first Architexture piece
Weird Placements
When discussing the musical material we wished to integrate to the app we realised that we wanted a range of musical material. The main motivation of the Architexture project is to investigate heritage acoustics, through choral music.
For Architexture V, we wanted a number of short snippets of music that users of the app could ‘mix’ together on the fly. Some of these snippets fit the space and work with the acoustic in the traditional sense. Other snippets are written for particularly odd singer placements e.g. standing close to a wall facing away from the traditional audience position in order to explore the space in a novel way. A number of “weird placements” were recorded in awkward places, such as the room corners, high in the ceiling, low down near the floor, and even in doorways.
One of our weird placements showing a source and receiver in the doorway leading from the Foyer into the Hall.
Final Plans
Floorplans were marked-up to generate a map of the sources and receivers.
The final plans made for the measurement session. The ‘Omni’ sources refer to turning the speakers to each point of the compass, this is explained later.
Equipment
For this project, we required several types of equipment, as discussed in the 2nd learning exercise (Recording Room Impulse Responses). Impulse response recording requires sources (loudspeakers), receivers (microphones) and an audio interface.
Sources
As we wished to create an acoustic map of the space we required a large coverage of the space. Typically sources in choral performances are unidirectional, this means that they have directionality (they point in a direction). If you have noticed that it is harder to understand speech when the speaker faces away from you, this is the effect of direcitonality.
The app has the ability to control the direciton of the sources (singers). Due to our unidirecitonal sources multiple measurements for each point of the compass were required to form a comprehensive set of results that allowed us this functionality.
Time in this case becomes a limiting factor, with nearly 19 receiver locations and 14 source locations equalling about 250 measurements. Then considering the directional functionality multiplies this by 4 to equal 1000 measurements. If this was to be done with a single source at a time then the time for measurement would outpace the scope of the project.
Thus, we chose to use multiple sources, all 14 source locations were filled with Genelecs, some sourced from the location and some from the AudioLab’s stock. As we were using slightly different models we needed to calibrate the sources to the same loudness.
Receivers
We chose to use three different systems when making the measurements:
mh’ acoustics Eigenmike: This was chosen due to its 32 different capsules allowing for 4th order ambisonics, this translates to better spatial cues and better immersion to the space. The issue with this system is the noise floor of the microphone, All the preamps of the 32 capsules are contained within the microphone meaning there is a large amount of electrical noise.
Rode NT-SF1: This microphone was selected as it offers a great 1st order ambisonic experience with a much lower noise to that of the Eigenmike. This can be used as a good baseline if the Eigenmike measurements come out as too noisy for use.
AKG 300B ‘Blue Line’ with the CK92 Head: The AKG Blue Line was used as an omnidirectional mono measurement. Due to the low directionality of lower frequencies it could also be used to bolster the other measurements in the low end.
Measurement
Dealing with noise of the location was a factor we had expected as the location was in the middle of the city flanked by two pubs. Thus we knew that the later in the evening it got the more we would have to deal with noise. Due to this our measurement process was to capture the high priority locations first.
The typical process was that we would move the receiver to a given location and then rotate all omnidirectional (all direction) speakers to north. The reaper session was laid out in sections, where each section represented a type of source (unusual, stage, main hall), this way if something interrupted our measurement (police siren) then we didn’t have to redo the whole measurement. We would then wait for all the measurements to be taken before rotating the speakers clockwise 90° to East, then take an additional set of measurements, and so on until we came back round to north. Here we would then move the receiver to the next location and the process would start again.