Strictly speaking, a microphone is a single channel of audio input, in that it is a mono signal. USB microphones which include integrated ADC and USB codec can be considered a single channel device.
Dual channel would be a measurement which includes a second channel of signal input in addition to the microphone. This would be an electrical feedback signal taken from the speaker terminals. This second channel provides a timing reference as well as the ability to provide compensation for linear errors between the audio output and speaker, everything within the feedback loop. Both channels need to arrive from the same audio device sharing the same clock, even 1 sample error can be significant. USB audio interfaces are ideal for this task, including at least 2 channels of input, as well as integrated phantom power for the condenser mic, and individual gain adjustment on each input. Standard XLR microphone is a must.
In it’s simplest form, a dual channel configuration can be achieved by simply running a patch cord from audio output to input for the signal reference. This provides a short feedback loop which provides no compensation between audio output and speaker, but does provide the valuable timing reference information. This is often referred to as “semi-dual”.
Single Channel Measurements
Single channel measurements include no timing reference to record the specific moment when the audio leaves the speaker, so generally the timing is aligned using the peak of the measured impulse response as the reference, effectively removing the “time of flight” from the measurement. When combining multiple measurements from multiple audio sources, the lack of timing information is not desirable.
The “work around” for the missing timing information between multiple audio sources is to play both sources together to create acoustic interference, and adjust timing of each audio source in software until a matching response is achieved.
Figure 1 Single Channel Configuration
To design using single channel measurements, there is a simple 3 measurement process that can be completed to determine acoustic offset. Simply place the mic at the tweeter axis, measure the tweeter, measure the woofer, then measure both connected in parallel. In VituixCAD, you can go to Tools -> Auxiliary -> Time Align. The process then is to adjust the delay until the sum of the two individual measurements matches the real measurement you took of the two drivers in parallel. I would un-check the minimum phase "MP" boxes and use the measured phase as-is. VituixCAD will automate this process for you by the click of a button. Minimum phase determination is often suggested for single channel data, but not strictly necessary.
Figure 2 Example: Single Channel Delay Determination
The limitation of this process is that this delay value only applies to these specific measurements only, so it locks the result to a single mic location, and a single distance which is often a fair bit closer than typical listening distance. This may not be a big problem for small speaker, but for large towers the response difference between 1m and 2-3m listening distance can be significant. Additionally, the delay between multiple drivers is a moving target at different angles, and there is an inability to measure each driver on it’s own axis, as moving the mic between measurements for delay determination is not possible. VituixCAD follows modern CTA-2034-A measurement standards, which aligns closely with the "Spinorama" data you may have seen from Klippel analysis at sites such as Erin's Audio Corner or AudioScienceReview. A single channel measurement is simply not feasible to maintain accurate timing information across a full set of spatial data for multiple drivers.
Figure 3 Single Channel Design Example
REW can utilize an acoustic timing reference to overcome these limitations, however I would not recommend this process. The acoustic timing reference requires a second speaker to provide the constant timing information prior to the measurement signal. However, care must be taken to maintain the distance of acoustic reference to mic when moving the mic, and this care must be taken if any measurements need to be repeated at a later date. Dual channel measurement utilizing electrical feedback described below overcomes this limitation, allowing for ease of repeatability and greater reliability overall.
Basic rules for measuring with single channel data:
- Mic location must remain stationary for all far field measurements
- For each pair of drivers, 3 measurements are required to determine acoustic delays between drivers
- Physical driver offsets in crossover should remain at 0, only single delay value is utilized
With dual channel measurements, a second channel includes an electrical reference feedback which can serve a few useful purposes. This feedback signal captured time of flight from driver to mic, so any acoustic delay between multiple measurements or drivers is inherent to the measured information, no extra steps are necessary. The measurement can also be normalized to the reference signal, meaning that only the speaker response is measured, any amplitude non-linearities prior to the reference signal are compensated for, including the inclusion of a protection capacitor.
Due to these benefits, many measurements at many angles can be captured with accurate relative phase by simply maintaining a constant timing reference, since the reference signal will remain at the exact same point in time for all measurements. Off-axis performance can be simulated accurately using real measured spatial data. Because many phase-accurate measurements can be utilized, advanced processing of full 360 degrees of acoustic information can be accomplished, such as determination of in-room response, power response and directivity index which provides much better insight into the speaker performance than a single axis of simulation.
Even though measurement distance is often around 1 meter, simulated results at any distance can be achieved using the spatial information with some interpolation and calculated distance from mic to driver.
The downside of a dual channel measurement system is that it's not quite as "plug and play" as a USB mic. It can be plug and play in a semi-dual configuration, but some extra work to build a simple jig is required to provide full benefit. The jig provides a voltage divider to step the signal voltage down from speaker level signal to line level, as well can be used for driver impedance measurement as well (not covered in this document). Clio, REW, ARTA, and SoundEasy are all capable of dual channel measurements.
Figure 4 Semi-Dual Channel Configuration
Figure 5 Dual Channel Configuration
Figure 6 Dual Channel Design Example With Spatial Information Included
Basic rules for measuring with dual channel measurements:
- Mic distance to baffle surface must remain constant for all far field measurements
- Measure at driver axis for each individual driver
- Centre of rotation for spatial information is front of baffle surface
- Window reference start time must remain constant for all far field measurement processing
Examples
Figure 7 Single Channel loopback measurement of a vintage Realistic SA-1000 amplifier.
Figure 8 Dual Channel loopback measurement of a vintage Realistic SA-1000 amplifier.
Figure 9 Single Channel transfer function of XT25 tweeter with series 50uf capacitor.
Figure 10 Dual Channel transfer function of XT25 tweeter with series 50uf capacitor.
More information on the measurement process for VituixCAD can be found in the VituixCAD help page, under “How to start working with VituixCAD”:
https://kimmosaunisto.net/Software/V...with_VituixCAD
Further Information for Measuring with ARTA:
Further information for measuring with REW:
https://www.roomeqwizard.com/help.html