signal-processing | 易学教程

How does a complex number capture phase, amplitude AND frequency in the result of an FFT?

阅读更多关于 How does a complex number capture phase, amplitude AND frequency in the result of an FFT?

问题 I understand that the magnitude and phase are captured in the real and imaginary parts in the result of an fft. But how does each sample capture phase? Is the phase related to the N discrete samples provided in time domain? That is, if the input sample included 44100 samples for a second, then is each resulting value of the FFT represent 1/44100 of the phase? For example, the first FFT value is at frequency 1/44100 and the second value is 2/44100 and so on? 回答1: i think you are saying "phase"

how to convert normalized frequency to actual frequency

阅读更多关于 how to convert normalized frequency to actual frequency

问题 suppose that we have following code function [ x ] = generate1(N,m,A3) f1 = 100; f2 = 200; T = 1./f1; t = (0:(N*T/m):(N*T))'; %' wn = randn(length(t),1); %zero mean variance 1 x = 20.*sin(2.*pi.*f1.*t) + 30.*cos(2.*pi.*f2.*t) + A3.*wn; %[pks,locs] = findpeaks(x); plot(x) end i have generated signal using following command y=generate1(3,500,1); and i have got 501 length sample,now i want to use music method to detect frequencies,namely 100 and 200,assume that number of parameter is 2,so i have

How to begin building a VSTi Plugin?

阅读更多关于 How to begin building a VSTi Plugin?

Im wondering the exact method through which I would go to build a VSTi Plugin is. I don't expect to code the next Massive in a few shorts week, as I have no knowledge of DSP and very basic programming skills. Im sure this is probably above my current level but I figure I'll grow as a programmer if I give myself a high goal that Im deeply interested in. All that being said, Im at a loss as to where to begin. I know that I would need to download the Steinberg VST SDK, but many of the other resources I've searched have given conflicting info as to what framework I need to download, etc etc. So

Compiling FFTW3 for IOS 5.1 on OSX 10.7

阅读更多关于 Compiling FFTW3 for IOS 5.1 on OSX 10.7

问题 A similar version of this question has been asked before (how to compile fftw3 on iOS) about previous versions of IOS and/or OSX, but I am unable to get fftw3 working on a actual IOS device. (Though it works fine in the simulator using the MacPorts "universal" distribution). If anyone has had any success using fftw3 please let me know how you were able to get it working! I am using OSX 10.7 and am trying specifically to run fftw on an iPad3 with IOS 5.1 installed. Thanks. 来源： https:/

Programming a Spectrogram using C [closed]

阅读更多关于 Programming a Spectrogram using C [closed]

问题 It's difficult to tell what is being asked here. This question is ambiguous, vague, incomplete, overly broad, or rhetorical and cannot be reasonably answered in its current form. For help clarifying this question so that it can be reopened, visit the help center. Closed 8 years ago . I am trying to make an audio spectrogram in C and am thinking about using the BASS library: http://www.un4seen.com. I searched over Google and here, but majority of examples are in C++. If any of you have any

R equivalent of MATLAB's filter function

阅读更多关于 R equivalent of MATLAB's filter function

问题 I'm adapting MATLAB code to R and trying to generate a waveform using ARMA formula. Is there a simple R equivalent function for MATLAB's filter to take AR/MA coefficients to build a waveform? npts = 100; a = [1 0.6]; % AR coeffs b = [1 0.25 3]; % MA coeffs e = randn(npts,1); % generate gaussian white noise waveform = filter(b,a,e); % generate waveform 回答1: Hmm can't you achieve that with filter function in the package signal ? require(signal) a = c(1,0.6) b = c(1,0.25,3) e = rnorm(100)

Compute EWMA over sparse/irregular TimeSeries in Pandas

阅读更多关于 Compute EWMA over sparse/irregular TimeSeries in Pandas

Given the following high-frequency but sparse time series: #Sparse Timeseries dti1 = pd.date_range(start=datetime(2015,8,1,9,0,0),periods=10,freq='ms') dti2 = pd.date_range(start=datetime(2015,8,1,9,0,10),periods=10,freq='ms') dti = dti1 + dti2 ts = pd.Series(index=dti, data=range(20)) I can compute an exponentially weighted moving average with a halflife of 5ms using a pandas function as follows: ema = pd.ewma(ts, halflife=5, freq='ms') However, under the hood, the function is resampling my timeseries with an interval of 1 ms (which is the 'freq' that I supplied). This causes thousands of

how to use Band Pass filter(18Khz to 22Khz) in Android

阅读更多关于 how to use Band Pass filter(18Khz to 22Khz) in Android

问题 I need to use Band Pass Filter to filter my real time recording audio. The Range is about 18Khz to 22Khz. I've seen many examples,unfortunatly they are used to detect frequency of audio....Since i'm new to DSP,please ellobrate your answer.... 来源： https://stackoverflow.com/questions/21456630/how-to-use-band-pass-filter18khz-to-22khz-in-android

Audio Manipulation In C++

阅读更多关于 Audio Manipulation In C++

I hope this is the right place to post this and somebody can help. I am a music technology student and I've recently picked up learning C++ as it would greatly help my career knowing a programming language, especially this one since it is used in the video games industry. Anyways onto the main topic. What I want to create is a program (in C++) that lets the user load a 16bit linear PCM WAVE file. Then I want to manipulate the audio sample data within that wave file. I want to either remove every nth sample or randomise them within a certain parameter (±10%). Then write it as a new WAVE file. I

Fourier transform of simple sin wave in matlab

阅读更多关于 Fourier transform of simple sin wave in matlab

I try to show spectrum of simple sin wave as we know a single sin wave with fixed frequency must have on peak in its spectrum I write this code but I can't get this one peak what is wrong in my code: clc nsteps=200;%number of signal elements in time domain i=sqrt(-1); NFREQS=100;%number of elements in frequency domain ddx=1e-9; dt=ddx/(6e8);%separation between each time domain elements lambdai=150e-9; lambdaf=500e-9; freqi=3e8/lambdai; freqf=3e8/lambdaf; freq=zeros(1,NFREQS); for j=1:NFREQS freq(j)=freqi-j*(freqi-freqf)/NFREQS;%desired frequency domain end arg=2*pi*freq*dt; et=zeros(nsteps,1);