Preparation of response measurements for crossover simulation with VituixCAD Kimmo Saunisto, 21.2.2017 Measurements Measurement gear Microphone with calibration file Sound card with 2 analog input channels. Left input for mic and right input from output of sound card. Analog output (left) to amplifier to driver under test. Licensed ARTA measurement software version 1.9.0 PC, cables etc. Another option is CLIO 11 with FW-01 and MIC-01. Processing of measurements with CLIO Standard is slower than with ARTA if time window is not in final position when files are saved. Note! Single channel measurement systems with possible latency variations are not recommended for speaker engineering. ARTA should not be used in single channel mode for far field measurements because timing is normalized by the program. Single channel mode is okay for near field measurements of conventional woofers and vents. Impedance measurements 1) Impedance response (with phase) of one woofer if all drivers have own box. If drivers share the same box volume, they should be measured together in series or parallel - like in final connection. 2) Impedance response of one mid-range driver if all drivers have own box. If drivers share the same box volume, they should be measured together in series or parallel - like in final connection. 3) Impedance response of tweeter. Measurement setup with LIMP, justmls etc. See user manual for more information. Warning! Do not use B&O ICEpower or compatible amplifiers having balanced/bridged output and high D.C. voltage in both speaker terminals. Speaker minus terminal should be in the same potential with 0 V terminal of line input.
Gear setup for acoustical measurements Set up measurement gear for semi-dual (or full-dual) channel acoustical measurement. See ARTA manual "Semi-dual channel measurement setup for acoustical measurements": Right and left should be swapped with sound cards having phantom power only in right channel. Prepare manual rotation table with angle scale 0 180 deg, steps 10 deg. Diameter 80 120 cm depending on speaker size. This makes measuring process easy and fast. Measurement parameters: Signal Sequence length Sampling rate Output volume Log-frequency sweep Dual channel mode SPL Sweep 256k 48k 0 db for far field, -20 db for near field (other volume controls constant). Yes Yes while far field measurements. Single channel mode is adequate for near field measurements. about 85 db @1m, or 2.83 V AC with SPL-calibrated measurement gear. Use 2.0 V AC at speaker terminals if actual SPL is unknown and sensitivity of driver is typical i.e. close to 90 db/1m/2.83v.
Terminology Few terms need to be defined first: X-axis is horizontal from left to right when observer is in front of speaker. Y-axis is vertical from floor to ceiling. Z-axis is horizontal from microphone through the speaker to front wall when measuring axial response 0 degrees hor/ver. Far field measurements Measure far field frequency responses of one woofer and one mid-range driver and tweeter at 100 cm. In horizontal plane around speaker to single side: 0, 10, 20,, 180 degrees. Basic rules: All far field measurements must have the same signal level from power amplifier. Amplifier volume, microphone sensitivity and other volume potentiometers should not be touched between measurements of different drivers. 1st order reflections should be avoided or delayed. Measure at elevation of 110 125 cm (close to half of room height). Upper mid-range driver and 3rd woofer from the floor with WWMTMWW to get maximum possible distance for floor and ceiling reflections. Absorb with large and soft pillows on the floor if possible to enable time windows longer than 4 ms. Saved measurement filenames (.pir) must have valid coding for off-axis angle. ARTA style angle coding <name-prefix>_deg[+ -]<num>.pir is okay. CLIO style <name-prefix> <num>.mls where angle <num> is multiplied by 100 is also okay for VituixCAD. For example M15CH002_hor_deg+110.pir equals M15CH002 to horizontal angle of 110 degrees. Elevation of mic is at the center point of driver under test i.e. mic and driver have the same Y-coordinate. Rotation center on X-axis while off-axis measurement sequence is at the center point of driver under test. Rotation center on Z-axis while off-axis measurement sequence: a) Rotation center on Z-axis is common for all drivers if drivers are installed in straight non-stepped baffle or baffle stepping is only few centimeters. Rotation center is typically on surface of front baffle at the tweeter. This way timing and phase differences between drivers are picked automatically with dual channel measurement. Location Z mm of drivers will be 0 mm in the simulator. b) Rotation center on Z-axis varies with stepped baffle. Each driver on different baffle level has own rotation center on Z-axis. Distance from each baffle level to microphone must be constant (100 cm). Differences on Z-axis are entered to the simulator as Z mm of driver, e.g. tweeter 0 mm, mid-range -30 mm, woofer -100 mm. If driver is not located at vertical center line of front baffle, off-axis rotation to direction where distance from driver to edge is shorter. This compensates excessive directivity in time-windowed measurements. Very nonsymmetrical constructions should be measured both negative (to left) and positive (to right) off-axis angles. Angle recipe is -170, -160,, 0,, +170, +180 deg.
Near field measurements Do not touch volume controls of amplifier and microphone input. Decrease volume with measurement program (ARTA) by 20 db. Decrease more with volume potentiometer if mic input clips (red light flashes) when measuring reflex port or midrange driver at near field. Measure near field response of one woofer cone at 5 mm, 15 30 mm from dust cap towards surround. If two woofers have shared box, feed signal to both woofers and damp the other (which is not under test) gently with pillow to prevent midrange frequencies going to mic too much. Measure near field response of reflex port(s) or passive radiator(s). All separately if they have not equal dimensions. Typically mic head is penetrated few millimeters inside the port where tube with constant diameter begins. Measure near field response of one mid-range driver at 5 mm. About 15 mm from dust cap or phase plug towards surrounding rubber. Useful metadata with the measurements Drawing of front baffle; dimensions, driver locations, edge radius (locations specified if radius variates much). Type of drivers, or at least Sd or Dd for magnitude scaling. Dimensions and number of reflex ports or passive radiators. Typical listening distance. Room acoustics measurements in two.pir files. Measurement at listening position, from left and right channel separately, with common small 2-way or 3-way, sine sweep 256k, about 80 db SPL. Sketch of listening room dimensions and layout. Construction materials and location of possible acoustic panels and diffusers. Hired VituixCAD operator can do the rest. If you hire VituixCAD operator, zip all measurement files (.pir,.zma or.mls) files and microphone calibration file (.mic) and metadata and send by e-mail.
Example: WWMTMWW speaker with closed enclosures Exporting far field responses Find earliest sample number It is usually in woofer s IR to 80 degrees horizontal. Sample #434 is selected as starting point (cursor position) for all far field responses measured at 1 meter. One sample too far does not matter too much because this happens only once. Distance from mic to driver s cone is about (434-300)/48000*344500 = 962 mm.
Export far field responses of tweeter Open axial measurement of tweeter. Set cursor on sample 434 as defined earlier. Set marker just before IR peak of the first room reflection. FFT 8k, IR Wnd Hann 12%, PreDelay 0.000ms. Open Export (spatial) Frequency response window. Select Gated. Enter Low 10 Hz and High 22000 Hz. Click Generate FR in text file button. Close the window.
Export far field responses of midrange driver Open axial measurement of midrange driver. Set cursor on sample 434 as defined earlier. Set marker few samples over IR peak of the first room reflection. FFT 16k, IR Wnd Hann 12%, PreDelay 0.000ms. Open Export (spatial) Frequency response window. Select Gated. Enter Low 10 Hz and High 22000 Hz. Click Generate FR in text file button. Close the window.
Export far field responses of woofer Open hor 80 deg measurement of woofer. Set cursor on sample 434 as defined earlier. Set marker few samples over IR peak of the first room reflection. FFT 16k, IR Wnd Hann 12%, PreDelay 0.000ms. Open Export (spatial) Frequency response window. Select Gated. Enter Low 10 Hz and High 22000 Hz. Click Generate FR in text file button. Close the window.
Exporting of near field responses Export near field response of woofer and ports Open near field measurement of woofer. Set cursor on sample #300. Set time window with marker to 600 ms. FFT 32k, IR Wnd Hann 25%, PreDelay 0.000ms. Open FR window. Smoothing 1/9 1/12 oct. Select M+P and export response to ASCII (txt).
Export near field response of midrange driver Open near field measurement of midrange driver. Set cursor on sample #300. Set time window with marker to 50 ms. FFT 16k, IR Wnd Hann 25%, PreDelay 0.000ms. Open FR window. Smoothing 1/9 1/12 oct. Select M+P and export response to ASCII (txt).
VituixCAD Options Verify that Number format of angle coding is compatible with your response files. Reference distance is typical listening distance (but don t set below 2500 mm). Check Mirror missing angles. Verity that all parameters in Power response & DI Calculation group are correct:
Baffle diffraction simulation Simulate cabinet impact response of woofers Include woofers representing average of all woofers. Two woofers represent all four in this case. Set mic on lower woofer. See parameters below. Export cabinet impact response without directivity. Simulate cabinet impact response of midrange driver Include single driver if it represents all. Set mic in the center point. See parameters below. Export cabinet impact response without directivity.
Merging of near and far field responses Merging of woofer responses Amplitude of near field response is adjusted manually with Scale [db] to signal attenuation (with opposite sign) adjusted earlier in measurement program (or until levels are equal within blending range, especially 300 600 Hz). Delay [us] is set automatically while adjusting transition frequency. See other parameters below. Export merged responses with Save button. Merging of midrange responses Amplitude of near field response is adjusted manually with Scale [db] to signal attenuation (with opposite sign) adjusted earlier in measurement program (or until levels are equal within blending range, especially 300 700 Hz). Delay [us] is set automatically while adjusting transition frequency. See other parameters below. Export merged responses with Save button.
Simulation Adjust timing offset for drivers Load tweeter responses. Do not add crossover yet (or disable all filter blocks if already added). Show minimum phase response by checking View->Minimum phase. Adjust Timing [us] to negative until Phase response on top right corner looks like minimum phase (lime) i.e. phase lands to about 0 degrees at top octave. -140 µs (-48 mm) in this case. Enter the same value for other drivers/ways because (semi-)dual channel measurements from same (1m) distance to common z-coordinate on front baffle have equal timing. This improves readability of charts, normalizes top octave of excess group delay plot at 0 ms and impulse/step response peak to 1.0 ms etc. Note! PreDelay of 0.140 ms while exporting frequency responses with ARTA would make the same effect. Leave timing of all drivers to 0 µs in VituixCAD if valid PreDelay was applied in ARTA. If response measurements are not SPL calibrated, adjust Scale [db] until SPL is equal to spec. sheet of driver manufacturer. Enter the same value for other drivers/ways because all measurements from same (1m) distance with equal source voltage & gain have equal net gain. Load merged frequency responses and impedance responses of woofer and midrange drivers. Simulate the crossover.