Back in the late 1970's audio engineers started using Real Time Analyzers, or RTAs, to provide a live display showing the frequency spectrum of audio signals. These early analyzers worked by using a collection of electronic bandpass filters. Each band was typically one-octave wide and the bands were distributed on center frequencies spaced in one-octave intervals. The most popular octave center frequencies are: 62.5 Hz, 125 Hz, 250 Hz, 500 Hz, 1 kHz, 2 kHz, 4 kHz, 8 kHz and 16 kHz. One full range audio signal was input to the bank of bandpass filters with the output of each filter representing the signal level for that frequency band. Then, the output of each filter would be fed into something equivalent to a VU meter, usually an LED display, to show the loudness of the signal in each frequency band. By placing the level displays side by side they formed a live graph of the audio signal and we were able to see the bass versus the treble energy, or perhaps a peak in the midrange'all in real time as the music played.

In those early days of RTAs the popular analyzers had an LED display with a bar graph that displayed 10 bars as an octave equalizer. Each bar was driven by the output of one of the bandpass filters. In search of better resolution, engineers then divided the frequency bands into even more narrow bands. From one-octave frequency bands we advanced to 1/3-octave bands, going from 10 bands to 30 bands across the audible spectrum. Today even finer bands are used for audio analysis. 

More recently, digital technology has given us a new way to achieve the same results. Today, once we have a signal in the digital domain, such as a PC with a sound card, we can bring to bear the awesome power of Digital Signal Processing or DSP. Modern personal computers now have enough power to do a significant amount of processing of live audio signals.

By using powerful DSP methods, TrueRTA can show you 1/24^th octave frequency bands for a total of about 240 bands spanning the full audio spectrum! The display is very smooth and reveals even the finest detail of the program material under analysis, whether it is music or a test tone.

There are a many uses for an audio spectrum analyzer. Here are just a few:

• Measure the frequency response of various types of audio equipment
• Analyze the distortion characteristics of audio equipment
• Study the noise characteristics of acoustic environments or electronic gear
• Professional audio recording, mixing and mastering
• Audio product design and production testing
• Voice and music learning aid
• Design and testing of musical instruments

Some of the above applications require the use of the analyzer in combination with a test signal from the generator. Others (such as noise analysis or analyzing live or recorded music) require no test signal.

Q: I'm new to this stuff. Do you have a link that can help me learn more about the basics of TrueRTA?

Yes, the AV Science Forum has a link "TrueRTA for Dummies" where they address calibration, settings, and connections in a straightforward manner.

http://www.avsforum.com/avs-vb/showthread.php?t=572477

You have to use the Windows Record Mixer to select the input to TrueRTA. The signal selected for recording will be the signal that TrueRTA receives as its input. See "Getting Started Quickly" under the Help topics for details on opening the Record Mixer.

Sure! We have many TrueRTA users who get excellent results using laptop PC's.

The MINIMUM required to run TrueRTA is a PC with a 500MHz Pentium III processor and 64MB of RAM.

top

There is probably a conflicting file in your Temp folder. Try emptying your Temp folder and installing again.

No problem. Your original registration code will work fine. The product is not keyed to your hardware.

Yes you can easily transfer TrueRTA to your new PC. It will not be a problem.

top

While laboratory grade microphones and mic preamps are relatively easy to find, they can be quite expensive. Inexpensive measurement microphones and mic preamplifiers are more difficult to locate so we have made an attempt to locate a few products that can make it easier for audio hobbyists to set up an audio test lab.

Behringer ECM8000 Omnidirectional Measurement Microphone

This microphone is generally available at music stores for around $50.00. It is an electret condenser type and requires phantom power (+15 to +48 VDC) from the preamp in order to operate. You will need a mic preamp that supplies phantom power in order to use the mic with your computer sound system. Based on the frequency response curve provided with the microphone the response appears to be within plus-or-minus 1 dB from 20 to 20 kHz.

Dayton EMM-6 Omnidirectional Measurement Microphone

Virtually identical to the Behringer ECM8000 but often on sale at a lower price. Individual calibration charts are available for download.

See the table on our home page for more hardware suggestions.

One source for microphone calibration services is Herb Singleton. You can contact him at

Cross'Spectrum Laboratories
info@cross-spectrum.com

 

You can read more about Herb at this link:

http://www.cross-spectrum.com/measurement/about.html

For basic measurements the simple unbalanced mic input on the sound card can work fine. You will probably be limited to multimedia mics or a mic that you build yourself. Noise may become a problem with longer cable lengths.

If you are using a professional balanced microphone (like the Behringer ECM8000 we recommend above) you will need a mic preamp with balanced inputs and phantom power. This usually calls for the use of a professional mic mixer as the preamp and source for the phantom power (see our recommendations above).

Not really. That is just a representative mic cal file. If you cannot have your mic individually calibrated I would recommend using no mic cal file.

The mic's phantom power should have no effect on the measurements...it just powers the mic.

top

If you are using the M-Audio Mobile Pre you will not need to use either the headphone out or mic in on the notebook. All audio will input/output through the USB preamp. Make sure you tell Windows you are using an external sound system; open the "Sounds and Audio Devices" control panel, select the Audio tab and then select the M-Audio Mobile Pre as the default device for both playback and for recording. Click OK to save these settings.

Setup the Windows Playback and Record mixers as follows:

1) At the Playback Mixer: Select the Wave signal for playback and make sure Line-In is NOT selected for playback. Set the Wave fader to a reproducible level (max or 50%)

2) At the Record Mixer: Select the Line-In signal and set the fader to either maximum or 50%. You may want to change this later.

I suggest you use the line in and line out of the M-Audio Mobile Pre for calibration. This means you will need to patch (connect) the line out (RCA) to the line in(1/4") for self test and for the various calibrations.

Now, simply verify that you have a signal passing through the system. Set the generator to output a 200 Hz sine wave at -10 dBu. Turn the generator on. Use some phones plugged into the USB preamp to verify that you have the 200 Hz tone coming through the mixer. Next run the spectrum analyzer by selecting GO in the toolbar. Adjust the frequency limits and dB limits fairly wide in order to see the tone. Is the tone returning to the analyzer? If so, then proceed to the sound system calibration. If not then troubleshoot the setup until you have the signal returning to the analyzer.

The default setting of the Audigy mixer has the Line-In signal playing back through the Line-Out. This places the system in a hard wired feedback loop when you patch the Line-Out to the Line-In for testing purposes. Depending on the level settings, the system may oscillate of just give a bogus frequency response.

Fortunately the fix is quite simple. You just need to set the mixer so it does not route the Line-In signal to the Line-Out. The details are included in the Audigy setup file below:

https://www.trueaudio.com/downloads/audigy-setup.pdf

Also, make sure you review the Quick Start in the TrueRTA Help Topics.

Yes. An external USB sound system can be used where the best precision is required. But it is also possible to get very reasonable results with the built-in sound system. Many customers report good results with the M-Audio products.

Make sure you have the outboard audio system selected as the DEFAULT audio input and output device at the "Audio" tab of the "Sounds and Audio Devices" Control Panel. TrueRTA always uses the default audio input and output devices.

Yes, you can use USB or firewire based sound systems with TrueRTA. The software will use whichever sound system you have selected as your "default" sound system.

top

The Line In and Line Out calibration routines simply calibrate the system input and output levels so that when you specify .500 Volts at the generator you actually get a signal of that level appearing at the sound system output. Similarly for the input. No, nothing will appear to "happen" with these calibrations. The Sound System Calibration will run two sweeps showing the uncorrected and corrected system response. This calibration is not available in the free version.

Yes. Perform the Line Out calibration to calibrate the generator, then do the Line In calibration to calibrate the analyzer and scope.

These are reasonable signal levels. You can calibrate either the balanced signal ( hot to cold) or the unbalanced signal (hot to gnd). Pick the one you will use the most.

The signal level at some specific point in the signal chain can be calibrated so that it corresponds to the level specified at the generator. Most often the signal chosen for calibration is the unbalanced output of the PC's sound card.

I would normally expect a local maximum at the warble tone center frequency. But it is possible to construct the warble either way. In any event you will likely want to retire that CD once you have Quick Sweep working correctly.

No. You only need to calibrate it if you are going to make voltage measurements where the absolute voltage or dBu level must be known. If you are only interested in relative measurements like "there is a peak of +6.25 dB at 500 Hz compared to the rest of the response" then voltage calibration is irrelevant.

If you do not calibrate TrueRTA then the actual Line-In and Line-Out levels will be unknown. When you set the generator output level to -10 dBu the actual output level may be considerably higher or lower. Similarly, an input signal that reads -10 dBu may actually be higher or lower in level.

Using Quick Sweep you can measure frequency responses to within + or - .05 dB. For acoustical measurements the accuracy is typically limited by the accuracy of the microphone calibration. The stability of the AC signal calibration will vary depending on your sound hardware but is generally quite good.

Your TrueRTA is probably miscalibrated. The quick fix is to go to the Audio I/O menu and open the Line Out and Line In Calibration dialogs. At each of these dialogs click on the button labeled "Restore Default Calibration." That should fix the problem.

top

I have to assume that the RTA really is seeing a signal ~200 Hz. One possible cause might be a hardwired feedback loop. This can happen if you are configured for self test (line out patched to line in) and select the Line In signal for playback. The fix would be to deselect the Line In signal at the Play Mixer.

TrueRTA does not support scripting...but you can use two memories to hold upper and lower test "limits" for production testing and it will remember your complete test setup if you save it as a project file.

There are two popular types of broadband test signals that are used for spectrum analysis: pink noise and white noise. 

White noise sounds like the noise you hear when you dial between stations on the radio. It's also commonly referred to as "hiss". White noise has equal amounts of energy for any given linear frequency band. There is as much energy in the range from 0 to 1 kHz as from 1 kHz to 2 kHz. When you look at white noise using an FFT analyzer you see a flat frequency response. That is, all frequencies appear at the same loudness across the spectrum. But if you view white noise with an RTA, the RTA shows the noise much more like our ears hear it. To our ears the white noise is hissy sounding with an emphasis on high frequencies and a lack of low frequencies. That is also how white noise appears on a RTA where it exhibits a 3 dB per octave rise with frequency. 

Pink noise has equal energy per octave or fractional octave band. This means that a 1 octave wide band of pink noise centered around 100 Hz has the same energy (and loudness) as a 1 octave band centered around 1 kHz. The result is that pink noise appears flat when viewed on an RTA but appears to have a response that falls at 3 dB per octave when viewed on an FFT analyzer. To the ear, pink noise sounds 'flat' with the lows and the highs sounding about equally loud. This closer match to what we hear is the reason audio engineers prefer to view live audio signals using an RTA or 'constant fractional bandwidth' analyzer. Pink noise also provides for a more nearly constant signal to noise ratio when making audio measurements.

The low freq limit is 10 Hz. The high freq limit is 48 kHz for sound cards that can sample at 96 kHz (like the Audigy 2), or 24 kHz for the more typical 48 kHz sound cards.

Just save them to separate memories and display those memories. Use Alt+1 to save to memory 1 and so on. See the Memory commands under the View menu. Note the Memory Toolbar is used to display or hide each memory.

You can take a "screen shot" by pressing the "Print Screen" key on your keyboard. This places the image of the screen on the clipboard. Then paste the image into your graphics application and save it in the format you wish. 

The software itself is extremely accurate. By far the largest sources of error will be the microphone and sound card. Fortunately TrueRTA provides calibration routines to remove the response of the sound card and accepts standard microphone calibration files to remove the response of the microphone from your measurements.

There is gating but it is sufficiently wide that it does not exclude room reflections on the acquired sweep. Because of the wide gate, room reflections will influence measurements when testing indoors. For this reason I recommend performing your most critical measurements outside in the open whenever possible. This is really the only way to remove the effect of the room from the measurements. Gated in-room measurements therefore are not particularly useful in the bass range...where the worst problems tend to occur. Measurements made outdoors in the open using Quick Sweep can be truly anechoic while retaining full low frequency resolution. Indoors Quick Sweep shows you the combination of the speaker response and the room influence...as you normally hear the combination. Taking the difference between the ground plane (half-space) response and the in-room response reveals the effect of the room.

You can use TrueRTA to very accurately measure loudspeaker frequency response and also evaluate the distortion performance of loudspeakers. For distortion testing you drive the speaker with a sine wave and directly read the levels of each distortion component in the measured spectrum. For example, a distortion component which is 20 dB below the level of the test tone constitutes 10% distortion. Here are some other levels and the corresponding distortion level:

Yes, it should be fine. We don't have a correction file for the RS meter but I have seen them posted on the internet. Note that with a laptop you will have to be careful not to overdrive the mic input. Make sure to first verify that you have a clean (not overdriven) signal at the scope before attempting any calibration routines. You may have to use a generator level down around -30 dBu with your notebook.

You probably want at least 1/3 octave resolution. By far the most popular version is Level 4 with 1/24th octave resolution.

top

 
Q: I am unable to get a response when I connect a microphone. It does not read any mic input on the PC I am using. Is it the sound card or what?

This sounds like an input signal selection problem. Make sure you select the input to TrueRTA at the Windows RECORD mixer. Here is how you do it: First, open your Playback Mixer by double-clicking on the volume control in the task bar,  then go to the Options menu and select Properties. When the properties dialog comes up switch from "Adjusting the Volume for Playback" to "Adjusting the Volume for  Recording." Click OK. Now, at the Record Mixer select the Mic input. The Mic input signal will now appear as the input to TrueRTA.

You might want to review the Quick Start topic in the on line help system. Once you are comfortable with operating the Play and Record mixers input selection will become routine.

It is normal for any PC sound system to exhibit a broadband noise floor such as you describe. It is usually higher with the mic input because of the extra gain used at that input. Even the very best audio gear will exhibit broadband noise when analyzed this way. So the question becomes: What is the level of the noise?

You may have the Windows Playback mixer set wrong. Open the Windows Playback mixer and make sure that you DO NOT have Line-In selected and that the Line-In fader is at minimum. Only the WAVE signal should be selected at the Play mixer. (a hard wired feedback loop can be created if you are in loopback configuration with the Line-In signal is selected and fader raised...not good) This can cause a wide range of strange problems.

It is possible that you could have a damaged file in your Windows installation. It sounds like your "autoexec.NT" file could be damaged. See the following support topic at Microsoft:
 
http://support.microsoft.com/default.aspx?scid=kb;en-us;324767

I am using Windows 7 Ultimate 64 bit version.  I found that a capture program called 'Tunebite 7' was causing TrueRTA to not start. I uninstalled Tunebite and TrueRTA started working. I will contact the support team for Tunebite.

I am using Windows 64 bit version and had problems with TrueRTA  until I uninstalled a conflicting program called "Soundtaxi".