af 3 C | Measurement Glossary ‘The unique language of audio analysis. by Pat Brown 2D) INMY LIFETIME, the sizeof sophisticated audio analysis systems has evolved from table top, to under-aeplane-seat, ‘0 computer bag, to cell phone. The cost has evalved from the price of « nice automobile to that of « Happy Meal. As such, cher re more audio practitioners than ever equipped 0 perform sophisticated loudspeaker snd room measurements ‘Those who make the decision co get serious about this ae faced with the daunting tsk of leaning che ropes in order ro get meaingfal data from their measurement platform. Te comes down to signal processing, nd the theory and rincples are not unique to audio and acoustics. They arsed by virtually every engineering fed, and even spill nto seem ingly non-technical elds euch s sccounting and photography. ‘Audio practitioners use sofware and hardware rook with deep signal processing rots "The good news is tha textbooks and websites regarding very part ofthe measurement proces ae plentifil The bad news is chat you can Wikipesis-yourslf util kingdom come and never caver al You wil suo find tha the rigid definitions may not even sem t apply ro audi, since they were not devel ‘oped to measure audio systems. In many case they are presented Jn the most concise form possible ~ as mathematical uations [decided to cook up a glosiry of some ofthe terms most Frequently encountered when working with audio analyzers of all types. Since acoustic snalyzersantlyze audio signals this lossy applies to them, 10, have relaxed the rigidity ro ‘ommnnicate the concepts, resulting in definitions cht, hile less gener are moe applicable to how ado prcttionerswse therm. [ts my hope that this wil help you better understand your measurement platform of choice, Time ° Froquency ‘Domain \ > Signal Domain - The X-avs (horizontal of « 2D plot ofan audio waveform, usually a caprured impulse esponse (IR). The strength ofthe signa x plotted on the Y-ais, ether in Enear ‘units (presur or voltage o sa level in dB. The two domains ‘most often ued to analyze signals ae the time domain and the fiequency domain (Figure 1). > Fast Fourier Transform (FFT) A mathematically efficient algorithm for determining the spectral content (Frequency domain) of atime domain waveform. In measurement work, the FFT is usualy performed on the impulse response (Figur 2). > mpuise Response (I) ~ A display ofthe signal ampliude ‘time ofan impulse chat poses through the aytem, Acoustic «samples nlude haneeps and ballon pops in-room. Is the invere-FFT ofthe transfer Finctce, which canbe eaptured using oimpukive stimu (pink noise orlog sweep) Figure 3). Tie a nem Tine me Figure 1: The Time ane Frequency Doma. Figure 3: The Impules Response. 38 Ure sounsiseatora Juyzor4> Log-Squared Impulse Response — A time domain plot ‘hat results from rectifying the IR and displaying it on a log vertical axis. It splays the relative level of the various events, making it easier to judge whether or not an event is significant (audible) Figure 4). » Envelope-Time Curve (ETC) ~ Formerly the Energy-Time Curve, it displays the envelope ofthe impulse response. Iisa sort of “smoothed!” log-squared response, and can aid in interpreting it. Note that the teem “ETC” is used loosely in measurement, and its exactly meaning is specific tothe analysis platform (Figure) > Time Window ~ A technique used to limit the application of the FFT to only part of the impulse response. This yields the transfer fanction for only part of the time record. A time window can be used to reduce the effects of room reflections on the transfer Function, efecrively allowing anechoic loudspeaker measurements to be made indoors. > Symmetrical Time Window — Symmetrical time windows are used when there is significant energy on both sides of the direct sound arrival of the impulse response that must be rejected from the FFT. Iris usually centered around the direet sound arrival (Figure 6). > Shaped Time Window — A time window that is tapered at its edge(s) to avoid abruptly interrupting the time domain data ‘A symmetrical time window is tapered arboth ends. There are ‘aris time window “shapes” available (eg, Hana, Hammin Blackman-Harris, etc.) (Figure 7). > Half-Window — A time window that is tapered on the trailing edge only. Hal-windovs are used when the impulse is near the stat ofthe time record, where theres very low energy ahead of the impulse arrival, Here the use ofa symmetrical window is ether unecessary ort could exclude the direct field acrival (Figure 8) > Asymmetrical Time Window — A time window that is not tapered (rectangular) atthe leading edge, but tapered atthe trail ing edge. The leading edge of an asymmetrical time window mast always be placed before the impulse atival (Figure 9). The leading edge of a half-window ean be placed afer the aval of| the impulse (Figure 8). The half windows implemented by many analyzers are actually asymmetrical windows, and you will not ind universal agreement on which i the correct” implementation > Dual or Muiti-Time Windows ~The use of more than one time window length. This allows the room reflections to be excluded from the high frequency response, but allows the oom into the measurement for computation of the low fre quency response. This appears to emulate the way that humans perceive sound. In effect, at low frequencies (long wavelengths) the contribution of the room reflections become part of the woofers response, and cannot be separated from it, either by the listener or the analyzer (Figure 10) In general, symmetrical time windows are used by analyzers that collec the real-time transfer function, Half-windows are usually used when the operator manually selects the portion Time jc Figure 4: The Log-Squared Impulse Response. iB I Tine ma Figure 6: The Envelope Time Curve (ETC). ne lib | J hh = Figure 6: A Symmetical Time Window (rectangular). | I Time jx Figure 7: A Shaped Symmetrical Time Window. Aalto toa = Ni IN Tie Figure 8: A Half-Window. of the impulse response to be transformed to the frequency domain (using cursors). The objective ofall window types is to produce a frequeney response magnitude response that is fee cof comb filters (frequency response ripples caused by the phase interaction of multiple sound arrivals) leaving the part that can be meaningfully improved using an electronic equalizer ty 2014 Live Sound nteratons! 99= Ls N , =a ate =| / om os \ ‘Log Frequency yer Figure 10: A Qual Time Window. > Finite Impulse Response (FIR) ~ An impulse response of fixed time length. An example ia wave file of the IR of a room or loudspeaker, but it could also be a high or low pas filter used > Infinite Impulse Response (IR) — An impulse response that is generated on-the-fly with feedback of previous sample values In theory, such an IR would never decay to zero. Analog filters are IR, as are some digital fer types. In general, IRs have lower latency than FIRS. > Frequency Response Magnitude ~ A measure of the rela~ tive or absolute level ofthe signal vs. frequency. There is infor- mation about “how much” but not about “when. > Frequency Response Phase ~ A measure ofthe relative or alpsolure phase of the signal ws. Frequency. Theres inforn about “when’ but not about “how much.” In the vast major ity of system tuning applications, the display is of the phase response relative to a time reference, whieh is usually the arival of the direct sound field from the loudspeaker (its impulse response). This time reference is selected by the operator, or automatically determined by the analysis software, > Transfer Function ~ A display ofboth relative magnitude andl relative phase ofthe frequency response on the same plot (Figure 1), Iti the FFT of the impulse response, and the IRs the iF FT ofthe transfer function. This allows the observe to exploit che strengths of ether domain when analyzing the respons. > Absolute Phase Response ~The phase response displayed 40 ve Sound teratonal ly 2014 H Log Frequeney vy Figure 11: Transfer Function of a Bandpass Filter using “time zero” (the time origin of the signal) asthe reference. I always begins at zero degree, and goes negative with inereasing frequency: his is required by causality, which says tht the signal cannot arrive before it is emitted. Te absolute phase response can go negative by many thousands of degrees, depending on the “rime of fight” of the signal between source and receiver. Tis not very usefl for measurement work, buts used extensively in computer room modeling, where the time relationship between virtual loudspeakers in a virtual space must be considered. > Relative Phase Response ~The frequency response phase using a user-selected “time zero” which i selected to make the center frequency of the bandpass filter (c.g, loudspeaker) go through zero degrees at its center fequency. The “correct” time zero is the one that produces the least phase shift through the pass band of the filter, Unlike the absolute phase, the relative phase ean go positive (Figure 12). The relative phase response is an indicator of how well the system an preserve the shape of an audio waveform that passes through it. > Minimum Phase Response — A given magninide response can hhave an infinite numberof phase responses, depending on the fre queney-dependent delay ofthe signal passing through the stem. ‘The minimum phase response represents the minimum posible phase shift fra given magnitude response. Itean be calculated fom the magniade response using the Hlilber Transform, le serves as reference for comparison with the measured phase response ofthe system, Raw transducers are usually minimum pase > Non-Minimum Phase ~ Phase shift in excess of the mini~ mum phase response for a given magnitude response. Loud- wpe ie inciaoewal san = ‘Log Frequency yee Figure 12: Minimum (top) and Non-Minimum Phase Response of a Bandpass Filter, ev ProSounae.corspeakers with analog or IIR crossover networks are often ‘non-minimam phase (Figure 12). > Group Delay ~ A “joint domain’ display of time (y-axis) frequency (axis)‘The phase response can usually be displayed as"group delay” to make it more intuitive for determining the arrival times of various chunks of the spectrum (such as the woofer or tweeter response) (Figure 13) > Linear System — A 1:2 increase in the input signal level produces a 1:2 increase in the output signal level. Compres- sors and limiters are non-linear, and should be bypassed for measurement work > Time-Invariant System ~The response of the system under test is stable vs ime, making the measurement exactly repeatable, Moving the loudspeaker or mic during the mea- surement would produce time variance Since the Time and Frequency domains are mathematically related in an inverse way (for linear, time-invariant systems) the weaknesses in one damain ean become strengths in the other For example, the time domain is more useful for looking at “when” bat the fequency domain is more useful for looking a “what.” Most analyzers let you toggle back and forth between the time and frequency domains, and some display both simultaneously. > Comb Filter ~ A series of peaks and dips in the transfer fanction that are equally spaced in frequency. A comb filter is always result of multiple energy arrivals in the time domain, within the time window being observed. Since itis caused by the time/distance relationship between two sound arrivals, a comb filter wil be unique at every measurement postion. This makes the response impossible to correct by equalization (Figure 14). ANALYZER TYPES [Now some traditional, common-usage definitions of analyzer types and measurement methods: > Real-Time Analyzer (RTA) ~ Measures the frequency response magnitude in real-time. The data is usually octave banded (o factions thereof) and results from pink noise being played through the system. There is no phase data, so this Magnitude — oS = ae og Frequeney rs Figure 18: Magnitude and Group Delay of « Bandpass Fiter 42 ve Soune teratonal July 2014 pus rae Level = “Tine nerny = Frequency ts Figure 14: A Comb Filter caused by multiple energy arrivals. response is “time blind." The RTA displayes the magnitude of each V-octave band, but nothing about their ime relationship, ‘An RTA measures the stimulus along with how it is modified by the loudspeaker and room. Early RTAs were basically a bank of sound level meters driven through fractional-octave filters Modern RTAs measure in the time domain and compute the spectrum using the FFT. The data can optionally be octave- banded for display. This emulates the traditional RTA. RTAs are the simplest “frequeney analyzers” to use. Just place the mie, excite the system, and observe the response. But since the measurement is blind to the arrival time (and dicee- ‘son) of the sound, the observed response may have litle ro do wit what a human perceives when listening at the same posi- tion. This isan aspect of measurement that the more sophis- sicated measurement processes attempt to adcess > Dual-Channel FFT ~ Measures the impulse response (or transfer function) by comparing the signal fed into the system to the signal re-acquired through the measurement micro phone. A dual-channel FFT measures what happened to the reference signal, not the signal itself. As such, many different broadband stimuli can be used (even music). > Discrete Method ~The system is excited with a stimulus, and the resultant IR is manually time windowed {by the place- ment of cursors to mark the window edges) and the transfer fanction observed. Each isa diserete step. If I make changes with an equalizer, I will need to repeat the measurement to observe the effect of the filter > Real-Time Transfer Function ~The process of collecting the time domain signal, comparing it tothe reference signal, ‘ime windowing, and FF'T-ing tothe frequency domain, done at such blinding speeds that the transfer function is given in pseudo real-time and updated continuously. Ie is 2 “running” ‘measurement that can allow system tuning to be performed during a show. The effect of filtering is immediately seen and ev ProSounae.corheard, since the measurement is ongoing > Goherence Plat ~ A frequency domain plot used with the real-time transfer fanetion that indicates the likelihood thatthe observed response is valid, based on the input and output sig- nals ofthe system. High coherence suggests good data. For the discrete method, repeatability isthe “acid test" for good data STIMULUS TYPES > White Spectrum ~ A stimulus with equal energy per He, Examples include an impulse, white noise and linear sine wave sweeps. White spectra have energy that increases with fre- quency (since there are more “Hin the higher octaves than the lower ones), and typically yield poor signs at low fiequencies (Figures 15 and 16). > Pink Spectrum — A stimulus with equal energy per octave Examples include pink noise and log sine wave sweeps. A pink spectra voltage contains more low frequency energy (and less high frequency energy) than an equivalent white spectra vole. “Most general sound system testing methods utilize pink spectra. > Narrow Band Stimulus — A stimulus having limited spee~ tral content, such a a sine wave (Figure 15) > Broad Band Stimulus ~ A stimulus having broad spec~ tral content, ypically covering the entire audible spectrum. Examples include unfiltered noise and impulses. PUTTING IT ALL TOGETHER Here is an example of how all of the terms in the glossary might be used to describe che measurement process used for sound system tuning session: wish to improve the frequeney response of my loudspeaker through equalization. I will collect the impulse response using «pink spectrum signal (pink noise or log sine sweep), fed into the ystem fiom the sound ead oupoc of my dual-channel FFT _measurement platform. Thissignal is also looped back into one of| ‘the inputs allowing the analyzer to compare it to the signal picked Uupby the measurement microphone. The resultant IR is displayed asallog-squared response to better visualize relative levels ofthe sound energy asivals I bypassed the compressor/limiter blocks cof my DSP, to assure that the system response is linea. The mic was placed on the floor plane, and I made sure that no one was moving near the mic or loudspeaker. This, along with turning off the ceiling fans assured thatthe system was time invariant. [A time window is placed around the direct arrival, ro exelude oom reflections. Performing an FFT produces a transfer fane= tion thas fice of comb filters. A shaped time window is used ‘ovoid an abrupt interruption of the time domain signal and the resultant anomalies in the frequency domain, Equalizer fites are placed on bumps in the frequency response magnitude. Since these bumps are minimum phase, the equalization process also improves the loudspeaker’ phase response. The phase wrap at ao Se San! ‘Set ise Speen ro a - See Po Specrum rae ae ‘ener vom Frequency a Figure 15: Spectra of test stimu oss nS Frequency Figure 16: Energy of test stim crossover of the 2-way loudspeakers non-minimum phase due the IIR crossover fiers used in my digital signal processor. Thave tolive with it The phase response could posiby be improved by using FIR cli filters, atthe expense of increased overall group. delay. Since Tam eaning the ster before the event, Tam using discrete measurement method. While any broad band stimalus could be used Iwill use one with a pink spectrum so as to pro- vide equal energy tothe system in each octave band. This assures that I dorit smoke my tweeters while trying to energize the sub- woofer. Late, Iwill monitor the response during the event using aureal-time transfer function to assess the effet of the increased absorption presented by the audience asthe room fills up. Since my earlier measurements used a mie placement that ignored the fioor reflection, minimal adjustment of the equalizers needed for “audience compensation.” The eoherence ploc indicates whether the data can be trusted. It is needed because the spectrum ofthe ‘muse signal is continuously changing, and there may be time ‘variance caused by people moving around. CONCLUSION ‘That's a general overview of the terms and their application. ‘The principles are universal and apply to all measurement platforms. Some of the definitions were slanted toweard thei {general use in audio and acoustic measurements, As with all theoretical concepts, some of the definitions could be chal- lenged on the fine details, especially with cegard to theit ‘general use outside of audio and acoustics. Thats one of the reasons that I love this business. © PAT & BRENDA BROWN dead SyndudCon, conducting ‘audio seminars and workshops online and around the sworld. For more information go to http://wwwasynandeon.com. Jty2014 Live Seund nteratons! 49