dsp mobile

For a visual representation of sound energy in different frequency bands it is necessary to transform the digital time domain audio signal into the frequency domain using a mathematical method: the Fast Fourier Transform (FFT). We will not explain the details here – the algorithm is well known and found in many publications and in Wikipedia as well. We will consider only a few aspects of the processes involved.

Digital audio data is represented in samples. The sample rate describes the number of samples which represent 1 second of audio, e.g. a sample rate of 44100Hz means that you need 44100 continuous samples to represent 1 second of your audio signal. The FFT works on blocks of samples which are power of 2, i.e. 256, 1024, 2048 samples coming up to 5.8ms, 23.2ms, 46.4ms at a sample rate of 44100Hz. Each “point” of the FFT represents the energy of a frequency range. Those ranges are linearly distributed over the entire range (0 – 22050 Hz @ 44100Hz sample rate).  In our example with 256 points this would mean: the first FFT point represents all the energy between 0 and 172.3Hz whereas the last FFT point represents the range between 21878Hz and 22050Hz. The human perception of sound energy is logarithmic on the frequency axis. This means that the resolution of the lower frequencies computed with a narrow FFT block size is very unsatisfying because it leaves great gaps in the spectrum! One could argue that the solution to this problem is a larger block size like 65536 samples. Indeed this leeds to a FFT resolution of 0.67Hz per point which is quite satisfying. But: 65536 samples represent almost 1.5 seconds of audio! Despite the computational efforts there is no chance to get details on fast changes of the signal like in transients!

This relation between FFT block size and frequency resolution is the reason for the gaps in the Analyzer’s frequency view at lower block sizes. The shorter the block size the wider the gaps in the lower frequencies – we decided to leave those gaps as they are so you can always see how exact your analysis can be.

For further inside of the relation between block size and frequency resolution we can recommend this: http://en.wikipedia.org/wiki/Short-time_Fourier_transform

The FFT algorithms are based on periodic signals. A block taken out of even a periodical signal is not periodic – the FFT produces frequency leakage. To reduce this problem a window function is applied before calculating the FFT. This window “sharpens” the spectrum but may result in amplitude errors. Depending on the characteristics of the signal and the desired focus in the spectrum there are many different window functions. The Analyzer offers 6 of the most commonly used ones.

Here you can see the effects of different window functions on a 1000Hz sine:

snapshots show from left to right: flat top – Blackman-Harris – Blackman – Hann – Hamming – rectangular

zoomed in from left to right: flat top – Blackman-Harris – Blackman – Hann – Hamming – rectangular

important: if the Analyzer does not show any signal, please make sure that it has access to the built-in microphone here: settings -> privacy -> microphone
The Analyzer is an universal App designed to work smoothly on iPad and iPhone. Due to the different display sizes there’s only one main difference: while the iPad main view provides a frequency and a SPL Meter view on iPhone the SPL Meter view is available only in portrait mode. We integrated an autorotation lock function for the iPhone to prevent unwanted view changes.

Main view

Navigation and zoom are provided using the usual touch gestures:

 

The frequency view shows frequency over time graphs. An overview:

 Snapshots

Workspace

The workspace is the summary of all snapshots and the actual graph. Touch the workspace icon and save the whole workspace (including color settings), export it, change the background color or remove it for clearing the display.

Save & export

Each snapshot and even whole workspaces can be saved in the Analyzers own proprietary data format. All records are listed in the menu under records. They can be loaded, shared via email or iCloud and exported to images (.png) or text files (.csv). When exporting a workspace to a png file all graph and background colors are preserved while the export of a snapshot creates a black on white picture.

Menu

Use thecog wheel in the upper right corner to access the menu. Here you can manage your records, change various settings, visit the store or make use of the sophisticated in-app help.

Records

The left column shows your records. If you select a record a preview appears in the right column. It is zoomable like your main view. If you push the load button the graph shows in your main view. You can delete it, export it to .png or .csv files, send it via mail or move it between local storage (on your iDevice) or iCloud if activated. You can open and share .png and .csv files in appropriate Apps like FileBrowser or Dropbox.

iCloud

If you want your records available on all your devices you can activate iCloud storage. Go to the iDevice settings -> Analyzer and turn iCloud to on. Now you can move your records between local storage and iCloud storage using the export function!

Audio IO settings

The Analyzer uses either the built-in microphone, an audio source attached to the headset input jack or any suitable USB class-compliant audio hardware attached via the dock connector. The iOS automatically detects which audio input and output are used. If the audio IO route changes the Analyzer automatically blends in the current  audio route.

You can select input and output routing in this menu.

When a class-compliant USB audio interface is attached the Analyzer will work always on channel 1&2, 3-8 are not accessible at the moment. We tested successfully and can recommend: Apogee’s One, Duet and Quartet, the RME Fireface UCX, the Alesis IO dock and the Midiface Audioface II.

Analyzer

The frequency range defines the maximum range of the frequency (horizontal) axis.

A single band can either show the sum of energies of each FFT point between left and right border of the band (summed energy) or the mean over the whole band. This decision is made with the band filter mode setting. In sum mode a pink noise input shows a horizontal line and a white noise input shows a line with 3dB / octave  rise. In average mode a pink noise input shows a 3dB / octave drop and a white noise input shows a horizontal line:

It is possible to view a normalized spectrum. If the normalization is activated, the level in the frequency view is always normalized to full scale.

The dB range defines the maximum range (dB) of the vertical axis. It is of course only available when normalization is deactivated.

The block size defines the length of the analysis interval.

With the reference frequency you define the “middle” of the frequency scale. Very common for measurements is 1000Hz. We added 440Hz also because this is the “musical” reference (A4).

Window functions

With the window function you can influence the leakage of the FFT. Available are Rectangular, Hamming, Hann, Blackman, Blackman-Harris and Flat Top. For more information on fft & windowing go here.

Here are some properties and usages for the different window functions:

We are working on the Loudness App. It will be available again in the near future with an all-new functionality!!

Loudness provides loudness, loudness range and true peak level measurement for iPad2 & iPad3 according to the ITU RS 1770-2, ITU-RS 1771 and the EBU R128 and the EBU 3342 recommendations. In combination with a class compliant USB audio hardware like the RME Fireface UCX or the Alesis IO dock it turns your iPad into a qualified studio measurement tool. It comes with presets according to the most prominent recommendations of ITU and EBU and allows for completely user definable settings of many graphical and technical parameters. If you want to test Loudness you can check out Loudness Lite for a time limited fully functional version. Or just contact us!

Loudness screenshots:

Loudness features:

– Loudness, True Peak and Loudness Range measurement according to ITU 1770-2, 1771 and EBU R128 and 3342
– simultaneous display of momentary / short term / integrated loudness and true peak level
– moving adjustable loudness over time graph showing momentary & short term loudness and gate level
– compatible with class compliant audio hardware
– tested & approved audio hardware: RME Fireface UCX, Alesis IO dock, Lexicon Omega, miditech audioface II
– manual and automatic calibration
– user adjustable algorithm parameters: gate levels, target levels, gate range
– user adjustable graphic engine: bar resolution, time resolution, absolute/relative scale, scale range, critical true peak level
– In-App help menu

Technical notes:

The algorithms which are defined by EBU and ITU are computationally quite extensive. That’s why this App does not work on iPad1. The attached audio interface needs to be set to 48 kHz sampling rate.

We recommend to turn on airplane mode and turn off location services in the ipad settings for a smooth performance.