Sedona, Arizona, USA (Photo by Yuichi)
|=========== AudioTools Measurement Menu and Modules =============
The AudioTools application can be purchased though the appstore.
Left list shows the overall structure of the AudioTools measurement system. The SPL, the Acoustics, the Line Input... are the menu of the AudioTools. And each menu consists of several measurement modules. Modules with (*) are included in basic AudioTools price. The Settings module which is not listed here, are for global setting related to Input/Output/Microphone, and the Support menu includes video instructions how to use each module.
The menu and modules listed are as of Sept.-01, 2012.
|=== The Amplifier Property of iPad internal microphone (Before iOS-6) ===
This is a Gain-Frequency curve of the amplifier for iPad internal microphone. Measurement was done by loop back method where the output of the headset earphone is wrapped back to the headset microphone interface.
The roll off frequency is about 150Hz and the slope is pretty steep. However, this prevents the wind noise and, so called, the pop noise, too. About the internal microphone, please refer to Studio Six Digital web page.
As a bottom line, the internal microphone is not suitable to measure precise audio and acoustic characteristics. However, for some measurements the internal microphone is very convenient including A-weighted measurement. Because we carry only an iPad in the live environment for example. There is no need to carry external microphone and its stand, external box and power outlet etc. Just iPad only !
|==================== Signal Generator Module ====================
This is a main screen of the audio signal generator module. You can choose wave forms including Sine, Square, White Noise and Pink Noise. You can also generate fixed frequency or swept sine by just tapping the screen. This is stand alone generator module which is started from the Utility menu. However, each measurement module has their own wave generator to perform their function. Therefore, there is no need to come back here.
|======================== Fs finding ============================
By using the audio generator, you can find out the Fs of the speaker driver with a simple manner. Connect the components as illustrated. The headset connector is 3 pin mini jack. Do not use the Microphone pin. Only use left or right earphone outputs. To avoid this, you better use the stereo earphone jack NOT headset jack. Connect the speaker under test with series 8-10Ohms resister and measure the voltage by the AC Volt Meter. When 1KHz sine wave is generated, you can hear the sound from the speaker and confirm voltage readings on the Volt Meter. Now you can find out the minimum voltage reading by changing the frequency of the generator under may be 100Hz. This is the Fs. This operation have to be done carefully to avoid damaging iOS devices.
|=========== RTA Module (Real Time Analyzer) Before iOS-6 ===========
This is a example of a frequency response bar graph by using RTA (Real Time Analyzer) module. Please note that the microphone response has steep roll off in this case. So, lower frequency data is not reliable. In case you measure the acoustic property of the room, or you want to know the balance of frequency bands at specified location, this will gives you a good idea to know about properties in such a live environment with very convenient manner. You can easily move location in the space.
| =========== ETC Module (Energy Time Curve) Before iOS-6 ==========
This is a example of the ETC (Energy Time Curve). We can use this just by clapping hands. Then, ETC is obtained as shown. Then, we specify linear portion of the curve by two finders. Then, RT60 is estimated on the upper location of the screen. Still the low frequency limitation exist, however, it is a convenient way to know RT60. In this module, we can not break down the RT60 in each frequency band.
|Left chart displays the ways to connect external microphone to the iPad.
1: A condenser mic is attached through the iAI2 to the doc interface.
2: A iTestMic is attached directly to the doc interface.
3: A condenser mic is attached through the iXZ to the headset interface.
4: A dynamic mic is attached to the headset interface.
However, without using the doc interface adapter such as the iAI2 or equivalent, we can not escape from the limitation of the iPad internal mic amplifier. If we use a doc interface adapter, the mic amplifier depends upon the adapter which is usually suitable to the measurement.
Following several slides show same examples of TASCAM iXZ applications.
|Even there are some limitation of the mic amplifier property, there still
have several advantages when we use good condenser mic.
1: The mic properties are known
2: The mic can have physical re-location freedom from iPad
3: Some user applications are not so sensitive about low frequency accuracy such as under A-Weighted condition.
4: Some applications only compare the gain of "before and after". Therefore, the gain sensitivity itself does not affect the result.
Following is RT60 (reverberation) measurement by Behringer ECM8000 mic and TASCAM iXZ which supplies +48V phantom power. We can attach mic directly to the iXZ or through the extender cable or we can use iXZ extender from the iPad. But later one is not recommended. This way is achieved with only limited amount of investment.
|======== RT60 measurement by the Impulse Response (IR) Module ========
In this case, a swept sine wave is used which is included in the IR module. The "7 seconds sweep" signal which is generated in the IR Module is supplied to the power amplifier. The sound from the speaker is captured by the mic. Then, the IR module will generate the impulse response from the mic data. This impulse response is almost noise free. We can capture the impulse response by just clapping hands. But this one contains noise and does not contain all the frequency bands. The left figure shows clean ETC (Energy Time Curve). There is no special process to reduce the noise. Reduce something will loose some information.
|Originally the IR module calculate the impulse response from the wave captured.
Then, it calculates the ETC and the Schroader integration curve. It finds
out EDT, Early Reflections, Reverberation RT60(from T30), Clarify C50,
C80, CT, Definition, S/N and FFT. All the graphs and the curves are displayed
by the user's choice.
The left bar graph shows the RT60 by 1/3 oct band of the room from previous ETC. This is important data for adjusting the acoustic conditions of a listening room . How much? what type of? absorption material have to be placed in the room can be analyzed.
To measure this, we better have single wide range and Omni directional speaker. But this is nice to have idea. We can use single corn and a full range speaker. Or, just your system and mic location is in the listening position.
|This is a comparison of RT60 data of two rooms. One is for a small room
(orange) and the other is the entrance hall (yellow). There is quite large
difference. The sound? Yes, it defers quite significantly.
If we remember that the internal mic amplifier properties, the gain of the low frequency domain is limited. Therefore, the reliability of the low frequency analysis is a little questionable. However, it still gives us a good idea.
|==The Amplifier Property for iPad Internal Microphone was now improved==
After the following software implemented
= iOS 6
= AudioTools 4.8
the low frequency cut-off of the internal mic amplifier was improved on the AudioTools apps. Therefore, iPad and external mic with TASCAM iXZ connected with the headset interface will have almost flat frequency response as shown in the left photo. Drop at 20Hz is just about 1dB or so. This was measured by the Frequency Sweep function of the AudioTools with the input/output wrap around connection.
|=========================== The iAI2 ==========================
iAudioInterface2 (iAI2) is the best match to AudioTools applications. This will provide 2 channel balanced analog input and output and a microphone interface with +48V phantom power supply. The equipment is driven by external power supply or internal Li-ion battery. The iAI2 interfaces the iPad through the doc interface. Specs are as follows.
Mic amp : +0 to +50dB gain in 9 ranges
Line Input : = Frequency Response (20 to 20KHz +0 to -0.50dB) = Distortion : (THD+n 0.01%)
Line Output : = +10 dBu balanced (Maximum) = Distortion : (THD+n 0.009% at +8.00 dBu
|================== Frequency curve of the iAI2 ====================
Left is the frequency response of the microphone amplifier.
the frequency sweep module
The "Line out" signal is wrapped around the "Microphone" input. Therefore, the frequency curve must be the "Output Line Amplifier" + the "Mic Amp" + the "Input Amp". It looks really flat.
Microphone property is out of above discussion. So, in this case I use Behringer ECM8000 microphone with calibration data. So, I believe I can obtain reasonably nice properties all through the system. You can have a calibration data through the manufacturer or you can create by yourself.
|==================== Frequency Sweep Module ===================
Left is the measured woofer frequency response. Now I have reasonable low frequency curve by using iAI2.
|================== TASCAM iU2 doc adapter box ==================
The TASCAM iU2 is one of the alternatives for external doc interface though there are some exceptions.
Specifications are as follows
- Analog input x 2 (include Mic with +48V phantom power and guitar interface)
- Analog output x 2
- Digital output (optical)
This equipment is not designed for precise measurement purposes. However, this can be used for audio DIY for some of the AudioTools modules with inexpensive way.
Note: Power is supplied by iOS device. If phantom power is needed, external power have to be supplied through Apple USB power using the USB cable which is included in the iU2 package.
(Note: iU2 sales activity is ceased as of Jan. 2013)
|================= TASCAM iU2 frequency Curve ==================
Frequency response of the iU2 internal drivers and amplifiers. Green curve shows the property of the output driver + the mic head amp + input amp. Yellow curve shows the output driver + input amp. The green curve drops about 4dB at 20Hz and he yellow curve drops about 2dB. If we take this into our account, we can use this for our measurement.
For detailed functions and specifications please refer to their web site.
|============= iU2 compliance to the AudioTools Modules ==============
This is a compatibility chart where iU2 could operate for each AudioTools module.
The Black circle means that I have tested.
The White circle means that I have confirmed it worked.
NA means that those are not applicable to the doc adapter.
No mark means that simply I have not tested.
Note-1: The Transfer function module is only exception. Yes, we can measure the 2Ch-FFT but need one external cable.
Note-2: The impedance related measurements can not be done.
Note-3: Software environment I verified is iOS:5.1.1, AudioTools 4.6
|This is a basic idea of the iAI2 (not iU2) block diagram.
There is a line from L-Channel output back to R-Channel input. In the iU2, there is no such feed back line internally. Therefore, we have to have external cable to perform the Transfer Function measurement. The external cable is to connect from "L output" to "R input" for the measurement.
|========== Coherence Measurement by 2 Ch-FFT Module ==============
This measurement is to compare two signals. One is the source electric signal (Reference) and the other is the sound (Mic.) which came from the speaker through the power amplifier and other electronic components. Therefore, this measurement requires two channel input thru external adapter such as iAI2. Two channel FFT function compares two signals at the same time and displays the differences as the Coherence number. The Coherence number "100%" means that the Mic signal is identical to the source signal.
First of all, the time delay between the reference and the Mic. signal have to be adjusted because the sound signal arrives later.
|Major contribution of the delay of the sound is the sound fly time from the speaker to the mic. Delay of electronic equipment is minimal. However, if signal processor is included in the signal path, the delay becomes large.
In anyway, signal delay adjustment has to be done first.
There are two screens in the display in iPad. Upper one includes the coherence and the gain. Lower screen displays the phase. The coherence is shown 100%(up) and 0% (center of the upper screen).
In the left screen, there is up and down in the coherence graph. This is because of the reflection effects from the side wall in the listening environment. So, this disturbs clear position of the vocal or the instrument sound.
|In this environment, special treatment is taken on the wall surfaces. Some
sound absorption materials are applied on the wall surface. Then, the coherence
curve stays almost 100% level.
This improves the positioning of the sound source dramatically.
|================= Frequency Sweep Measurement ===================
In this module, the signal generator sweeps the sine wave from 20 - 20KHz by specified sweep time such as 10, 20 and 30 sec. The line input signal or mic. signal are measured.
Example shows the tone control property of the control amplifier.
Upper curve shows low boost and high boost case and lower curve shows the opposite case.
This measurement can also be used for the Speaker testing. In this case the MIC. instead of he LINE has to be selected at the set-up screen.
|============= Adjustment of Speakers physical locations ===============
In the multi speakers / multi channel amplifier system, it is a bit hard to adjust physical locations of each speaker. There are several ways to adjust speakers however, it is relatively easy to make the adjustment by using the delay finder function of the 2ch FFT (Transfer Function) module.
Left figure shows typical 3 channel diagram.
|2 Ch-FFT (Transfer Function) Module
First, the delay time have to be measured by using the delay finder in each speaker. The pink noise is applied to the system. The sound from the speaker is acquired by the mic. and is converted to the impulse response data. Then, the delay is computed.
Left is the Delay Finder window in the 2Ch-FFT module. Tapping the Find Delay button will give very accurate distance and the delay time.
Compensation of acquired delay time by the delay adjustment knob on the channel dividing equipment.
|In this experiment four channel system was the system under test.
This figure shows the frequency response after the delay adjustment compensation process was made.
In the property, it was fairly well adjusted all thru the channel except 6KHz locations which is in between the Mid speaker and the Tweeter. There is a small dip.
|The delay consists of several factors. The largest one is the sound travel
time in the air. However, there is delay from electric input to sound output
because of the group delay. The speaker group delay is frequency dependent.
Above mentioned delay which is detected by the delay finder is not always
the delay at the cross over frequency location. Therefore, some minor tune
up have to be taken.
The final adjustment is to fine tuning of the delay adjustment under monitoring real time FFT. Then, perfect adjustment was made.
|======= Visualization of the sound field (3D FFT by Smaart I Module) ======
The Smaart I module provides with the 3D FFT function. Which includes frequency, gain and time in one chart. In this application, The "Time" is converted to the "Physical Location". To realize this, the mic moved from right to left or vice versa with constant speed in front of speakers as illustrated in the left fig. Applied sound source is the pink noise which has all frequency band.
Purpose of this is to visualize sound field in the listening position. It is easy to understand the sound interference among multiple sound sources in visual ways.
|The SPL in front field of single speaker environment
Left data shows frequency (X-axis) and gain (Intensity) and physical location (Y-axis). In this case upper portion means the right side of the speaker and so on. The horizontal center line shows the SPL vs frequency at the center axis of single speaker. It is well realize that the directivity properties which is sharpened in higher frequency zone.
In this case, there is no interference observed. However, if the speaker is located near a wall, strong reflection creates up and down in the SPL curve.
|Front field SPL of the stereo speakers
Left data shows frequency (X-axis) and gain (Intensity) and location (Y-axis). The horizontal center line shows the SPL - frequency at the center of two speakers. In this location, the SPL keeps the same level all through the band. When the mic (listener) location is shifted left or right (A in the fig.), then we encounter the dark area which is dip of the SPL caused by comb filter effect in the area of higher frequency.
This indicates that the listener location have to be in the center. Small shift of the listener location creates small number of large dip(s). The large physical shift creates small many dips.
These are one of the example of the measurement.
|============ Impedance Measurement / TS Parameters =============
Impedance curve has a lot of important information in it. To measure the impedance curve the iAI2 is needed. However, it is very simple for the operation.
Connection is very simple like left photo.
Connection to the speaker under test from iAI2 output is done by the Standard jack.
Before starting, the calibration procedure have to be taken. Short cut the output terminal then tap. Then, open the terminals, then tap.
|Starting the impedance curve measurement, just tap the start button after
setting of the sweep time as 20 or 30 sec.
Left shows the measured curve with Re(ohms), Fs(Hz), Zmax(ohms), F1(Hz), F2(Hz) values on the screen.
However, to get TS parameters, it is recommended that the Re have to be measured carefully. Therefore, the Re from the manufacturers data book or measured by other method. In this system Re is identified as impedance value at 20Hz. In case of lower Fs, this is not true. The F1 and F2 are sensitive with Re. So, those also have to be carefully measured.
For TS Calculation you can use the TS Clack from the downlaod page.
Left curve indicates the impedance in the free air environment. Obtained parameters are: Re=7.2ohms, Fs=83.8Hz, Qts=0.9
|To calculate the TS parameters or specifically the Vas, second impedance
curve have to be obtained. There are two ways to get second one.
One is to add known value of weight on the diaphragm.
In this case, 16g (50 yen coin x 4) were added. Then, new Fs (yellow curve) is lower than the free air case because the diaphragm weight gets heavier. Two curves are displayed as left. How to get TS parameters are shown on the TS parameter page in this site.
Obtained Vas=15.3 liters
|One other way is to apply known closed air volume to the diaphragm. In
this case the speaker under test is mounted on 25 liters closed enclosure.
Then the suspension gets higher (green curve) by the air. Therefore, Fs
gets higher. New Fs, Zmax, F1 and F2 are obtained.
Then, calculated Vas=14.4 liters
Impedance curve depends upon the driver conditions including the constant voltage drive or the constant current drive or voltage value or current value or temperature or other factors. It is interesting but not simple questions.
|The iTestMic supplied by Studio Six Digital is convenient way to attach
qualified microphone capability to AudioTools. Left photo shows iTestMic
attachment to iPad, iPod touch and iPhone. The microphone can be used for
most of iOS applications on top of AudioTools. The reason why it is convenient
is as follows.
We can simply attach the iTestMic to the Doc or the Lightning port of the iOS devices. Then, AudioTools and iTestMic communicate each other to exchange mic SPL calibration data which was shored in each microphone. Any additional calibration activity is not needed for SPL measurement.
SPL accuracy : ;/-0.1dB
Frequency response 20-20KHz +/-3dB
Gain ranges : 28dab-105dBA , 48dBA-120dBA
Power consumption : under 50uA
NC Limit : under NC29
Cable length : 40cm
|Coming soon !||Coming soon !|