Calculate frequency response using FFT and inverse FFT

Hi,
Attached is the program using FFT and inverse FFT to filter a time domain signal. The frequency response of the LPF can be obtained by using the chirp signal from 0 to 5kHz. However, I don't know why the signal obtained from a sine wave input is so strange. The amplitude is wrong and has a envelope outside. Please help to point out what's wrong with that.
Bill
Attachments:
fft filter.vi ‏87 KB

If you check the help text for sine wave.vi you'll see that it generates the sine wave based on the following formula:
yi = a*sin(phase[i])
for i = 0, 1, 2, …, n – 1 and where
a is amplitude,
phase[i] = initial_phase + f*360*i
This means that when you input a=1, f=0,1 and initial_phase=0 you will get a sine wave that is based on samples at every n*36 degrees; i.e. at 0, 36, 72 etc...due to this sample rate you never see the full amplitude (+/- 90 degrees), the wave is clipped at the top. If you input an initial phase of 64 degrees you will get the full amplitude, but the wave is still deformed due to digitalization...
The lower the frequency you put in, the closer the digitalized representation will be to the true sine.
Use the Waveform Generator VIs from the analyze palette if you want to have more control over the wave generation (sample rates etc.). (Not available if you have the base package.)
MTO

Similar Messages

3 Simultaneous Frequency responses using PCI-4552

I need to obtain 3 simultaneous frequency responses (Magnitude and Phase) using a PCI-4552 board. Ch 1 is the reference for the 3 FR's. I succeded in obtaining independent FR's, but I cannot configure the vi with the 3 FR's together. I attach my best try, please take a look at it.
Attachments:
zoom4.vi ‏378 KB

I took a look at your VI. It looks like you have two issues.
First, you are using a single stop button's input to stop several loops. The problem is that the output of the button will not allow the consecutive while loops to execute until the first one has finished (they are waiting for the prior loop to finish to get the final value of the button). If you want to read the value of a single button simultaneously in several while loops, you need to use local/global variables.
You should have an example:
Stopping Parallel While Loops with Reset.vi
... included with LabVIEW which will demonstrate how to properly do this.
Your second issue is that the base analyzer can only do two simultaneous channels. If you really need to do more, you'll need to grab al
l the channel's raw data using the DAQ API instead of the DSA API, and feed the individual captures into the "Frequency Response Function (Mag-Phase)" vi. This should give you the same results, but you'll be using the host computer to do the analysis instead of the board.

Frequency Response using Swept Sin

Does anyone have sample C/C++ code to generate a swept sin wave and get the frequency and phase response data using NI-DAQmx on a PCI-4461 board?
Currently I'm using a PCI-4551 board and can use the helpful NI-DSA C API to do this, but the 4551 card is now obsolete so I need to replace it with the 4461.
Thanks in advance!

If you check the help text for sine wave.vi you'll see that it generates the sine wave based on the following formula:
yi = a*sin(phase[i])
for i = 0, 1, 2, …, n – 1 and where
a is amplitude,
phase[i] = initial_phase + f*360*i
This means that when you input a=1, f=0,1 and initial_phase=0 you will get a sine wave that is based on samples at every n*36 degrees; i.e. at 0, 36, 72 etc...due to this sample rate you never see the full amplitude (+/- 90 degrees), the wave is clipped at the top. If you input an initial phase of 64 degrees you will get the full amplitude, but the wave is still deformed due to digitalization...
The lower the frequency you put in, the closer the digitalized representation will be to the true sine.
Use the Waveform Generator VIs from the analyze palette if you want to have more control over the wave generation (sample rates etc.). (Not available if you have the base package.)
MTO

Difference between LabVIEW FFT and Matlab FFT

Does anyone know why there is an offset in the absolute levels of the results of the FFT function in LabVIEW and Matlab?
See the attached VI for an example of my LV code.
Also attached is the plot generated in Matlab, which seems to be offset by ~37dB.
This was generated using the command 'plot(10*log10(abs(fft(x))))', where x is the same input array as in the VI.
Many thanks,
Dan
Dan
CLD
Solved!
Go to Solution.
Attachments:
FFT.jpg ‏60 KB
FFT.vi ‏22 KB

I see that you are using the FFT Spectrum VI which returns a single sided spectrum. To convert from a two sided transform (ie FFT) to a single sided spectrum with RMS units there is a normalization of sqrt(2)/N where N is the number of points in x(t). You can premultiply your x(t) by N/sqrt(2) and get the same results as Matlab. Or, you could use the FFT.vi so you are comparing transform to transform instead of spectrum to transform.

Frequency Response Function & FFT & Inverse FFT (problem of unit Volts-RMS)

Hello everyone,
I am currently working on a VI in order to compare two analog signals : the first one corresponds to the output signal (my reference) which is sent by my data acquisition card to a shaker and the second one corresponds to the input signal recorded by an accelerometer fixed on the same shaker. The final goal of the VI is to correct the analog output signal by using the analog input recorded signal in order to have the vibrations on the shaker which corresponds to what we really want.
To summary, I have a problem of unit with the Volts-RMS...
So this is my method for the VI :
First, I have to calculate the Frequency Response Function between the two analog signals (output and input). For it, I use the " Frequency Response Function (Real-Im).vi " which returns the complex values of the FRF in Volts-RMS (but I don't want to use this unit).
Then, I want to calculate the FFT of the analog output signal (my reference). There are two different blocs which can be used : " FFT Spectrum (Real-Im).vi " and " FFT.vi ".
The " FFT Spectrum (Real-Im).vi " returns the FFT complex values of the signal in Volts-RMS and the " FFT.vi " returns the FFT complex values in Volts (or say me if I am wrong, thank you). I really would like to use the second one because of the unit.
Then, I divide the FFT just calculated with the Frequency Response Function calculated just before.
For the end, I calculate the inverse FFT of that with the " Inverse FFT.vi " which use the complex values with the same unit than for the " FFT.vi ".
I don't want to use the Volts-RMS unit because I absolutly want to use the blocs " FFT.vi " and " Inverse FFT.vi ".
The problem is that I don't find a bloc which use the same unit for the Frequency Response Function. The " Frequency Response Function (Real-Im).vi " returns only the complex values in Volts-RMS unit. Maybe it is possible to convert it correctly? Or maybe there is an other bloc which can be used in order to calculate the Frequency Response Function with the same init than for the FFT and Inverse FFT ? Because I can't mix everything for the moment...
Thank you for your help,
Best regards,
Sebastien

Hello Preston,
No, I have not use the Sound and Vibration toolkit. I have only used the signal processing toolkit with the two toolboxes " Waveform measurement " and " Transforms ".
But I think that what I have done for the moment in my VI is correct (I have finished the complete VI). But I am not sure of the units (Volts, Volts-RMS...) and I would like to understand.
I have tried with the Sound and Vibration toolkit for the frequency response function (because you say me that it deals with all the unit conversion) and I can obtain the same results than with the " Frequency Response Function.vi " of the toolbox " Waveform measurement ".
But I would like to understand the units (see my previous post please). For example, for the FFT (the result is a complex), why sometimes it is in Volts, sometimes it is in Volts-RMS ? Is it possible to convert it ? How ?
If you want, I can attach on the forum my VI and that will maybe help you to explain me. Maybe it will help other people interested.
And if someone else can give me other precisions or advices about it, do not hesitate.
Thank you for your help,
Sebastien

Microphone's frequency response

Hi,
Is it possible to plot a microphone's frequency response using sound and vibration? If it's possible, can anyone tell me how?
Thanks

Hi popcorn,
If you have the Sound and Vibration Assistant with SignalExpress, there is a step you can add under Analysis - Frequency Domain Measurements called, Frequency Response.
Here's a link to an article that discusses various functions, Sound and Vibration provides. Although the article discusses LabView, many of the same functions are found in SignalExpress as well. The Frequency Response Function can be found at the bottom.
http://zone.ni.com/devzone/cda/tut/p/id/3030
Aki T.

Where can I find the amplitude and phase frequency response when the AC coupling is set on the 4472

My friends,
I need to know the amplitude atenuation but most important, the phase distortion introduced on the low frequency components of the signals when the AC coupling is setted (i.e. when the high pass filter is connected) on the 4472 and 4472B DAQBoard.
I can construct the amplitude frequency response by generating and aquiring a sine waveform of a knew amplitude. But I cannot construct the phase distortion introduced by the circuitry.
However, I assume that this crucial information should be available in the DAQ Manual or in the website of NI, but untill now I haven´t found it.
Thankning in advance,
crimolvic from Chile

crimolvic,
Here are the Specifications and Datasheet for the 4472. They indicate a phase non-linearity of less than 0.5 degrees across all frequencies.
For information on how this varies with frequency, see the attached
spreadsheet. This response was the result of testing on a single
4472. Although this is classified as a "typical" response, it is
not gauranteed.
Have a great, day!
Travis
Attachments:
4472 Phase Linearity.xls ‏21 KB

Intermittent low volume and poor frequency response

I have a 10 month old 30 Gig video IPOD that drops the volume level and loses all its low frequency response at times and then will recover after a while. This happens on a song or songs that I know are properly leveled and that work sometimes and not at other times. The issue seems to be much more prevalent when IPOD is cold or cool. After it has been operating for an hour or so the problem seems to go away. It is very common when first turned on. I am getting ready to return it while it is still under warranty but thought maybe someone has an easy fix.

Returned for warranty repair.

Frequency response function modal analysis

After reviewing the signal analysis functions in DIAdem I have realized them to be a bit limited for modal analysis. I have a couple hammer impact tests that I need to process a frequency response function for, and since this is brand new to me I'm not seeing anything in the embedded function list that is going to help me. I was wondering if anyone out there has a couple of pointers on generating a FRF plot for modal hammer impact tests. I did notice that the ChnFFT2 command allows me to generate a transfer function, coherance, and FFT Cross Spectrum channels for analysis. Though I might be confused and this may be everything I need. My FFT2 settings are below.
[code]
FFTIndexChn      = 0
FFTIntervUser    = "NumberStartOverl"
FFTIntervPara(1) = 1
FFTIntervPara(2) = 2500
FFTIntervPara(3) = 1
FFTIntervOverl   = 0
FFTNoV           = 0
FFTWndFct        = "Rectangle"
FFTWndPara       = 10
FFTWndChn        = "[1]/Time axis"
FFTWndCorrectTyp = "No"
FFTAverageType   = "No"
FFTAmplFirst     = "Amplitude"
FFTAmpl          = 1
FFTAmplType      = "PSD"
FFTCrossSpectr   = 1
FFTCoherence     = 1
FFTTransFctType = "Spectrum H0"
FFTCrossPhase    = 0
FFTTransPhase    = 0
Call ChnFFT2("[1]/Time axis","'[1]/H_1' - '[1]/H_4'","'[1]/A_1' - '[1]/A_4'") '... XW,ChnNoStr1,ChnNoStr2
[/code]

Standard modal analysis has something denoted as FRF. I have a labview application note "The Fundamentals of FFT-Based Signal Analysis..."
Frequency Response Function
The frequency response function (FRF) gives the gain and phase versus frequency of a network and is typically computed as
where A is the stimulus signal and B is the response signal.
The frequency response function is in two-sided complex form. To convert to the frequency response gain (magnitude) and the frequency response phase, use the Rectangular-To-Polar conversion function. To convert to single-sided form, simply discard the second half of the array.
You may want to take several frequency response function readings and then average them. To do so, average the cross power spectrum, SAB(f), by summing it in the complex form then dividing by the number of averages, before converting it to magnitude and phase, and so forth. The power spectrum, SAA(f), is already in real form and is averaged normally.
Refer to the Frequency Response and Network Analysis topic in the LabVIEW Help (linked below) for the most updated information about the frequency response function.
http://zone.ni.com/devzone/cda/tut/p/id/4278
So the options for FFT2 are
No
DIAdem does not calculate a transfer frequency response.
Spectrum H0
DIAdem calculates the transfer frequency response by dividing the FFT of the output signal (A) by the input signal (E): FFT(A)/FFT(E). DIAdem averages the amplitudes of the individual transfer functions.
Spectrum H1
DIAdem specifies the cross spectrum and the auto spectrum for each signal pair. DIAdem calculates the transfer frequency response by dividing the averaged spectra: Middle(cross(A,E))/middle(auto(E)). DIAdem does not average phases, because phases can delete each other.
Spectrum H2
DIAdem specifies the cross spectrum and the auto spectrum for each signal pair. DIAdem calculates the transfer frequency response by dividing the averaged spectra: Middle(auto(A))/middle(cross(E,A))
If you assign the values Spectrum H1 or Spectrum H2 to the variable FFTTransFctType, DIAdem averages and divides the cross spectra and the auto spectra and calculates the amplitudes last.
Which state auto spectrum when FRF is power spectrum.

Flat Frequency Response

I probably shouldn't have to be asking this question since I charge people for my obviously amateurish recording abilities but it's one that I've never had explained to me and one I need to know the answer to......
Let me set the question up this way:
When you get in the car and pop in a professionly produced cd, most people crank the treble control up to 6-10 (on a scale of 10) and the bass up to (4-10) depending on the factory speakers and type of material and whether or not they care what their music sounds like. When you get home and you're listening to the much higher end home stereo still listening to that professionaly produced album, you still reach for those treble and bass knobs and crank them up several notches or if you have a graphic eq you tweak out a smiley face .
There's so much emphasis in the recording world about getting a flat frequency response out of your room with absorbers and bass traps and spreading around the reflections with deflectors, etc...etc..etc..., that we spend thousands of dollars on this stuff and some measuring software to make sure that it's flat. Then we use that flat response to produce music that sounds great and expect that to translate to those cd players and home stereo systems.
(I'll additonally preface my question by saying that I've had no problems getting my music to translate from my home studio to any other playback system, but I'm a little confused about what's going on.)
Now finally my question(s), when we reach for those treble and bass knobs on our car and home playback systems, are we really just trying to make up for the lack of bass and top end in those systems so that we too can achieve a flat frequency response and make the music sound good on whatever system? or
Do we as listeners actually prefer the smiley face frequency response in music and are we taking a cd that itself has a flat frequncy response and making a smiley face out of it so that it sounds good to our ears? (Please don't give me a material/genre answer.)
The reason I ask is because I have to put a graphic eq on my Truth 2031A monitors to make the professional stuff sound good through them, and then I in turn mix my music to sound the same for whatever material/genre of course. (I'm not really interested in any monitor bashers or I would've asked this over at Gearslutz.)
So again rephrased...Do the masses think music sounds good when it has a flat frequency response or the smiley face and if it's the lattter of those, how are we supposed to achieve that when our home studio setup is producing a flat frequency response, do we tune our monitors with eqs like me?
Additonaly I understand that when were talking about flat frequency responses in rooms we're talking about throwing sine waves through a system and ranging their frequencies, measuring them out so that we can detect any over emphasis/deficiencies in the room so maybe this question is a little more towards monitor tuning.

If you look up Fletcher and Munson in Google, you might begin to get a bit of the start of an idea of why this isn't quite as straightforward as it seems.... and I'm not sure that I can give you a complete answer either, although I can give you a few connected but slightly random things to ponder, wearing my acoustician's hat:
The fundamental problem is that when things are quieter (and less distorted, incidentally) our ears get more sensitive to the midrange frequencies, and if we listen to music that way, it invariably sounds as though the bass and treble are out of balance. In cars it's slightly different though; the frequency response of whatever's in there generally tends to be anything but flat - and often over-emphasises the midrange anyway. Treble tends to get absorbed very easily in upholstered cars, and since most car speakers don't have anything like acceptable tweeters in as a rule, it's not surprising that people want to increase the treble. As for the bass - well I'm always turning that down personally, but I know what you mean in principle!
If by a 'commercial' CD you mean one where the vocal is prominent, then yes I can easily imagine why you might as a matter of course want to increase the response at the extremes - it makes sense if you think about it. The mid-range vocal is prominent and probably compressed, so its average level is louder than the backing - this helps it to stand out. But also it distorts the overall time-based response - the backing may well be balanced so it's okay on its own, but that doesn't always translate if you have the wrong vocal settings applied, or at a minimum, applied unsympathetically. And some voices make this significantly worse; for instance Sealion Dying (AKA Celine Dion) makes the most appalling racket in the midrange, and you'd definitely need less of that!
So really I'd say that it's not a Bass/Treble issue, but a midrange one. If you look at commercial CDs in general, you tend to find that the energy distribution is pretty even over the whole audio band, which implies that it falls off at 6dB/octave if you look at it in Audition (this is an energy/Hz thing), but in reality most CDs these days are mixed a little brighter than that - more like a -3dB/octave slope down from about 1kHz, and that's partly to compensate for a lot of things - some of which are cars... You do have to watch out for the distortion issue though - most people don't realise, but you are able to tolerate rather higher levels of non-distorted sound than anything with significant distortion levels in it, and if that distortion is in the midrange, then you'll want it quieter anyway. So decent, over-rated under-run PA systems always sound cleaner and louder but you should beware - they can damage your ears just as much, if not more.
Do car interiors themselves increase the chances of midrange boost occurring? I think it's a pretty safe bet that they do, as a rule, simply because of the size of them, and the treble problem I already mentioned. And if you are trying to compensate for too much midrange, then the rest follows. Most domestic replay systems these days seem to be midrange heavy to me as well - I haven't heard anything cheap recently that had anything like a flattish response - and they really don't suit the rooms they are in either.
If you want to listen to material as it really should be, then you need to experience it live first, I'd say, and then do a direct comparison with what you can hear in your monitors. I'm fortunate - I can do this quite regularly with a variety of material. Do I tend to leave things as flat as they are recorded? Well, it depends on what it is. If it's in any way classical, then sometimes I look carefully at the bass balance, but generally I leave the rest alone. Everything else these days I just get to sound good - and that can mean all sorts of tweaks, depending on all sorts of things. More and more though, I've come to the conclusion that too much midrange isn't necessarily a good thing - but that's mainly because of the general lack of good reproduction equipment around these days.
As for monitors and flatness - well that's not really an issue for most people, compared to getting their listening environment correct. If you have a pair of cheap monitors in a good room, the chances are that the results will sound better than an expensive pair in an uncorrected room - despite what all the monitor freaks on gearslutz might say. These days, even the cheaper ones can sound quite respectable. But flatness isn't an issue with monitors really - a decent impulse response, and low distortion are far more important. Chances are that if a manufacturer has got this right, the monitor is going to be suitably 'revealing' anyway - which is what a monitor is supposed to do.
The one thing you do not do though, is EQ the feed to your monitors - that went out of fashion almost as soon as it came in - fortunately. You fix the room so that it's more truthful. If you EQ the monitor feed it will inevitably only sound good in one place in the room, and that's no use to man nor beast. The only decent things that proper room correction systems can do is equalise the immediate time response to take account of what's actually between the monitors and you - which if done properly can improve stereo imaging no end.
So answers? Well not really. But at least I've explained a few (but not all) of the issues.

50 Hz noise ECG filtering using FFT and its inverse in LabVIEW

I have to write a LabVIEW Program to put the 50 HZ noise in the ECG spectrum using the FFT and then with the Inverse of the FFT I must take it out,but without using any filter.so if anybody can help me with that, I will be thankful.

I have a normal EGG signal,and its Spectrum. Then I add the 50 Hz noise.In the Spectrum I will get this long peak at the x-axis at 50Hz.So,what I'm trying to do with LabVIEW, is to get out this noise without using a filter,but with the FFT and of course with its inverse to reconstruct the signal.
(the EGG signal is created at first with a simulator and not with human body, the signal is going through the "line-in" of the soundcard in the computer)
I hope you understand a little bit what I mean, I will attach what I did in the program till now,maybe you can understand it better.
Thank youMessage Edited by b_rani on 05-18-2005 01:46 PM
Attachments:
Soundin1_030505.vi ‏80 KB

Frequency Response VS FFT for measring frequency response of a audio ouput signal.

We have purchased the Sound and Vibration Toolkit and I have some questions.
From the frequency response example I looked at you measure the input then the ouput and the Vi gives you the difference.
I want to measure the audio frequency response a radio.
So all I have is the output I dont have the audio input to use as a reference.
How would I measure frequency response with the frequency rasponse VI with only the output audio signal?
I was also looking at the FFT example to measure the audio signal frequency response.
But from what I can tell so far this only does 1 channel I want to do both channles at the same time.
Is there a way to do FFT on 2 channels at once and have them output on the same graph?
Thanks for any help you can provide.

Hi,
I was looking through the examples and If you look at the SVXMPL_Multichannel FFT (simulated) examples, this shows how to take an FFT with multiple signals. You just pass in an array of the waveforms. If you need to put your signals together you can use the merge signals VI, which is located in the Express»SignalManipulation Pallette. Let me know if you need any further help with this issue.
Have a great day,
Michael D
Applications Engineering
National Instruments

Filter the data and then do an inverse FFT

Hi everyone,
I know this subject has ever been treated but it's not been solved so here's my question :
First I want to make an FFT on datas. Then I want to filter values (replacing the values I want to filter by 0) and finally make an inverse FFT to get back the time signal without the high frequencies.
Now my question is, how do I modify complex datas or how do I make my modified array acceptable for the inverse FFT vi ?
Thanks a lot
David

I think your problem is that you need some more FFT information. You are uncertain on how the output from the FFT VI is organized. Or more specific how to map the array content to different frequencies
I suggest you read this http://zone.ni.com/devzone/cda/tut/p/id/4278, and this http://zone.ni.com/devzone/cda/tut/p/id/4541. In the last one they used an older labview version. Some of the functions mentioned has minor changes in newer Labview versions. But the theory still apply
If you can answer this question I can help you some more.
If I sample a 5Hz sine wave with 1Khz sample rate for 1 second, and then perform a FFT, at which array indexes will I find information about the 5Hz sine wave
Besides which, my opinion is that Express VIs Carthage must be destroyed deleted
(Sorry no Labview "brag list" so far)

FFT - Frequency Response

Hi everyone!
I'm having a problem on my time domain and fft plot.
In the attached VI below, i made a Chirp signal and perform FFT (left side) and I wanted to have a plot which is moving similar to the Chirp signal I got from the net, which is at the right side of this VI.
I wanted my plot to be moving, or in real-time, I know this is just in the settings, but i cant figure it out. I've tried countless time already to obtain a plot similar to those at the right side of this VI. Also the amplitude of my FFT, is very small from the original amplitude value of 1.
Thanks, for the response.
Attachments:
sample_sweep.vi ‏113 KB

Hi gijude,
it could be so easy with some elementary math:
Reading online documentation like WIKIPEDIA can be a great source of wisdom…
Best regards,
GerdW
CLAD, using 2009SP1 + LV2011SP1 + LV2014SP1 on WinXP+Win7+cRIO
Kudos are welcome

Make frequency response analyser using frequency generator and counter

Hello
Can we make a frequency response analyser using a Frequency generator and frequency counter?
How to add modulation with it? Modulation frequency is to be varied as per the input to given to the carrier!
The outputs are Frequency, magnitude, and the phase as like solartron FRA.
somebody have an Idea for this
awaiting for the solution
thank you
"Thanks with regards "
by
..........Gireesh..........

Hello Gireesh,
You can use a function generator to generate frequencies and use the modulation tollkit and other tools availabe with LabVIEW to do the modulation part . Or you can use the analog output ports on the daq card to generate different frequency signals for the same purpose .This should pretty much serve your purpose.
http://zone.ni.com/devzone/cda/epd/p/id/5646
http://digital.ni.com/manuals.nsf/websearch/AF3615F31CE9656C862576070020B8F7

Calculate frequency response using FFT and inverse FFT

Similar Messages

Maybe you are looking for