FFT cross correlation

Similar Questions

• find the maximum correlation coefficient

  Hello

  I am able 2 analog signals simultaneously from the DAQ 9233 module.

  Whenever there is a trigger, my DAQ will gather X number of samples from the analog module and a cross on the data correlation analysis.

  My question is how can I get the maximum cross correlation coefficient between the two signals (e.g. North is my reference, how many samples should I change my South to get the maximum correlation coefficient)?

  My goal for this?

  This is so that I can get the time between two samples (e.g.sampling to 50 kHz, if I need to move the sample 300 times, so I know my delay is 300 * 1/50000 = 6ms)

  Thank you.

  Attached a picture of my waveforms

  North = my reference signal

  South = my 2nd signal

  Low = Corss correlation between two signals

  I'm sure that the maximum value of the cross-correlation function is the location the highest correlation coefficient.  Using your own example to verify this.

  Search the index of this value.  Subtract the size of the reference (in the X of entry) of the peak value index signal.  The value obtained must be the number of points which is equivalent to one quarter of work.  Divide this value by your sampling frequency (50,000 in your case) and you should have your answer.

• Cross-Covariance

  I need to calculate the covariance of cross of two vectors. How to do this if I do not have MathScript (and therefore impossible to use "xcov")?

  Thank you!

  There are cross-correlation VI in LabVIEW that should work for you

• Timestamp of photon arrival times

  I would like to make a measure of cross-correlation of two photons.  Auto hardware correlators and crossed correlators are available, but expensive.  It seems that many other users resulted in the idea of using the 80 or clocks 100 MHz in the counter of entry for each desired channel, starting the counters at the same time and get every channel of photons in the door of the corresponding counter.  Each photon triggers so the meter to store the total of the clock in its internal register, that is, the time stamp for this photon. Since then carries on a 32-bit counter clocked at 100 MHz only occur every 42 seconds, and average rate of photons will be at least 100 Hz, excesses should be easily recognized and corrected in post processing.  So assume that all we have to do is get meter reads in a buffer.  Bursts of photons with separations of arrival can be as small as few clock cycles.  Average rates may be a few hundred kHz or more on each channel.  We can adjust this rate by turning the power of the laser to the top or to the bottom - the source for the broadcaster that produces the stream of correlated photons.  I suspect many others would like to know which system is better for these types of measures, the 6602 or the X-Series.  If FPGA is a good solution, perhaps we should start a new thread about it.  But for 6602 or X-Series: what is the fastest, it can do so for short bursts (speed for how long) and what is the sustainable rate?  I realize that the sustained rate will depend on the material.  I thought I only saw that the PXI beats the PCI bus in this type of application.  PXI versions are better?  In particular, what are the effects of the FIFO 128 sample on the X-Series compared to the 2 on 6602 buffer solution.  And I noticed something about the speed of the front end for these two choices could be a problem in this regard.  Pulses (photon) TTL on the door will have a width of 10 to 50 nsec.

  For the flow of the series X is going to be much faster than the 6602 mainly due to the FIFO aboard (but PCIe gives more throughput than PCI).  PCI or PXI should have the same flow (and PCIe vs SMU would also have the same flow).

  Forum user (and NOT used) Andrew S displayed some X series streaming of references here.  On a single meter, it was able to maintain the sampling frequency of ~ 20 MHz on PCIe X Series and ~ 8 MHz on USB of the X series.  6602 benchmarks are much lower, somewhere between 200 and 300 kHz (although the system used for benchmarks is old enough).

  The 6602 allows shorter pulses that the X series.  The specifications page indicates a minimum of 6.25 the 6602 ns pulse width.  I saw a specified number directly to the X series in fact.  X series devices include a maximum external base of 25 MHz clock frequency, which assuming duty cycle 50% would be 20 ns high and low times.  50 nsec would certainly long enough of a pulse, but 10 ns * could * be pushing.  You will probably need to do some tests to confirm or not the X series is able to detect all the impulses for your configuration - if not then I would suggest the X series (PCIe-6320 is the lowest cost one) on the 6602.

  Best regards

• LTE application: Signal PSS and EU synchronization problem

  
  

  Just as I suspected, samples received on the receiver side are only cross correlated with the PSS sequence (PCI mod 3) = 0, this is the reason why the synchronization does not work when the eNB transmits with a (PCI mod 3) 0. The PSS sequence (PCI mod 3) = 0 is integrated as a time fixed signal of the domain in the FIR filters although (the two who treats I and rated a Q of the received signal handling complex). To change this and to implement a different sequence of PSS, should open the basic FIR blocks Xilinx generator and change the sequence.

  I got some guys from support LabVIEW Communication an updated version of the "LTE Cross Correlation.gcdl" function that implements a selective control for the value (PCI mod 3) and seems to change the PSS sequence within blocks of FIR. I post here if someone has the same problem. Although I have not tried myself yet.

• How to calculate time filter ADC for USB-4431

  I do a continuous measurement of the aregenerative AOAI.  I need to compensate for the delay of ADC filter both the delay in output filter

  The delay of output filter seems to be well documented in the manual NI DSA (link below)

  http://www.NI.com/PDF/manuals/371235g.PDF

  ADC filter delay is more difficult to find manual DSA says look at 443 x specifications.

  The specification gives a delay of "nominal" of 38 samples ADC filter.  I can't find a more precise definition (link below)

  http://www.NI.com/PDF/manuals/372485c.PDF

  I'm using labview 2010, on XP

  attached is a sample vi.  He uses a loop of producer consumer example that I modified

  I could manually find the total different sampling frequency offset, but I would have preferred a way to calculate it.

  Thank you

  Timm C.

  When you say 'calculate', mean you automatically measure, calculated by the formula, or search in the table?

  It's a good idea to measure the total delay (DAC + CDA + other) of the measuring device to each sampling frequency of measurement used in your application. Best practice is to run this measure whenever the system is powered up or rebooted.

  We recommend that you use VI provided by Sound and Vibration Toolkit to measure the delay and additional delays that may be introduced by the device under test:

  \vi.lib\addons\Sound and Vibration\svt_Calibration\SVL measure Propagation delay .vi (DAQmx)

  If you don't have a SVT/SVMS, you can recreate the measure of delay with LabVIEW complete edition features.

  Start to synchronize your AO and AI using the same implementation used in your application of measure.

  Use a test as a monopulse Gaussian signal and search for relaxation both the signal generated as well as the acquired signal. The difference between the locations of trigger is the delay.

  Or use a test signal as MLS or Chirp, and cross-correlation model to measure the delay.

  But I like to use the VI SVT ;-)

• bugs in the correlate of the IMAQ

  So, I have identified two major bugs in correlate IMAQ that make it unusable for my application (see post to the http://forums.ni.com/ni/board/message?board.id=170&message.id=401078#M401078 for more details about the bugs in IMAQ correlate). Does anyone have a dll or a C++ source code correlation standardized cross that I could use instead?

  -root

  Sorry, I misunderstood.  I did not need CC standard in the past, but I had a quick look at what might be required.  The best reference I found was Lewis (1995) "Fast normalized cross-correlation" (PDF link), which implements using tables of the sum.  If I'm lucky, I'll see if I can code it up (don't look too hard) - but probably using tables instead of the images to be reusable for 3D images.  It seems that the numerator of the NormCC can always be calculated in the frequency domain, and the tables of money speed up the calculation of the denominator.

  Cheers ~ Greg

• Measure highlights in the image

  Hello world.

  I was presented with a challenge.

  Our team of mechanics built a machine that randomnly erodes the surface of a test tube.

  It is an image of a specimen captured at the beginning of a test:

  

  As you can see four white lines are present in the image, the lines are caused by the reflections of the lamps.

  And this is a picture of a similar specimen at the end of a test.

  

  We already found a satisfactory means to measure areas of erosion and it seems to be robust for all colors of possible test specimen

  Now they decided they wanted the machine to measure the length of the white reflection lines (online), because the erosion to some materials only causes the surface to become tangled.

  And not necessarily give obscurity eroded area above.

  an example of such a case of use can be seen below:
  

  I tried several approaches without success.
  Ideally I would like to isolate the highlighted lines, so that the system could identify almost the same lines as those which is hand painted below

  

  Tried methods:

  1)
  IMAQ find Edge, to get the endless lines that best described the lines of reflection highlighted.
  Then using an binning method, the line is segmented, in the place where there is enough light in the picture.
  This method is, however, very sensitive to the horizontal change in lighting and usually either find endpoints, or frosted and middle section correctly.

  2)
  Tuning IMAQ detect lines
  It gives me:
  

  Which is exactly what I want, as I hoped that the lines on the left of the eroded area would also include, and lines lie above and below the highlights would have been excluded, but getting close.

  3)
  High pass filter of the image to improve the edges and remove the effects of the quality of variable light.
  The built-in methods had tend to reinforce the edge of erosion more or equal to the highlights, so I tried to build my own kernel to improve only the horizontal characteristics, but even with a core of 5 x 5 was the best result I got:
  

  Who has not exactly me left to separate the highlights of the eroded area.

  4)
  Finally, I tried a cross-correlation with an image of a horizontal line with a small width.
  

  Yet once I improved more than eroded area the line feature I was interested.

  If you have an idea, that I can use to either improve the highlighted lines, so that the length can be measured I would really appreciate it.
  I tried to adjust gamma, contrast and brightness to make the 'white' lines stand out, but due to the change in light, this also gives the bright Center of the image until the lines are increased in the extremities.

  Thank you

  / ZcuBa

  You can very easily find everything at the top and the bottom line using advanced search of edge. It extracts the values of the pixels form the profile of the line and you should be able to detect all non-uniform pixel values.

  There is a screenshot of the result.

  Good luck

  Dan

• Multiplication of cross-checking (circular convolution)

  Hello

  To make the circular convolution of vectors, I would like to make:

  FFT-> multiplication cross-check (W = u * V in MATLAB)-> IFFT

  for example [1 1 1 1 0 0 0 0] with himself-> [1 2 3 4 3 2 1 0]

  Is there an easy way to do this multiplication on the 50g, or a cross-check order direct cconv?

  Thanks in advance

  Hi, yYou can use the same commands with the 50G

  FFT

  IFFT

  For cross-checking multiplication on the 50g, try HADAMARD / / shift MATRICES OPER NXT HADAM

  RPL mode

  [1 2 3] [4-5-6] HADAMARD

  [4-10-18]

  In rpl:

  [ ... ]

  [ ... ]

  FFT

  HADAMARD

  IFFT

• RMS of the FFT or tip?

  Hello

  I have a sinusoidal signal that oscillates around zero the axis in the negitave and positive. It is to measure the height of the waves on the static water level in a wave tank. I am calculating the FFT but the amplitude is less than the height of the waves in the time domain. Does this mean that I'm the wrong calculation? I don't know wheather I should use the FFT Peak or RMS? I assumed that the amplitude of the FFT should give me the same amplitude as the time domain.

  Any advice would be greatly appreciated.

  Thank you

  Donners,

  It can get complicated, but I suspect that what you see is the result of the frequency of the waves is not not exactly at the frequency of the FFT bins.  For example, suppose that df = 1 Hz. Then each element of the array FFT represents the energy in a "sandbox" a hertz wide centred on the frequencies of the integer.  For a wave with most of his energy to 7.3 Hz TFF likely values not null for trays Hz 7 and 8.  In real systems with noise and light frequency variations, it is not uncomon for data can be spread over 10 or 20 locations. To get the fundamental signal amplitude, you would need to combine the energy in all stores.  Since the FFT produces complex output, the phase component must also be considered. Non-sinusoidal Waveshapes means that part of the energy is in harmonics. A pulse square can have a large enough amplitude, but there's still no broad spectral components.

  The short answer is that generally there is no easy way to get a precise correlation between the amplitude of crete in the time domain and the amplitude of spectral components in the field of the freqeuency.

  Lynn

• Cross fade

  What happened to cross fade between songs in itunes

  Hello

  You can configure iTunes to fade each song in (increase volume) while the previous song is by train, so there is no gap of silence between tracks. This feature is called a transition.

  1. Choose iTunes > Preferences, then click play.
  2. Select crossfade songs.
  3. To change the amount of time it takes to disappear at the end of the current song and the new song to fade, drag the slider to seconds.

  When iTunes play consecutive songs from the same album, a transition is automatically disabled, with no fading between songs (or titles).

  To play songs without fading, deselect Crossfade Song

  12 iTunes for Mac: play songs

• Disabled applications with marked cross circle around him

  I got the application Parallels on my Macbook Pro. Suddenly, an update alrert came when I tried to launch parallels. I hit accidentally YES. It was updated and now I need to enter the activation key once again, that I do not have access to.

  So, I tried to restore the image of Time Machine application (just app restore but not the restoration of the whole image).

  Now I can see the icon of the parallels application with a gray cicrle crossed on this subject. I can't launch the Parallels. Also I can't get the activation key if I reinstall parallels.

  How to fix the restored application?

  Also another question is, my timemachine showed some images of backups, but when I rebooted my macbook, I lost all my previous bodies timemachine.

  Now, I am in a position to fix my parallels application, but may not be bale to do.

  Please, help me to understand:

  (1) why an application gets this circle crossed out around her? and how to fix it without reinstalling.

  (2) why I lost all my images TimeMachine on restart the Macbook?

  Thanks in advance.

  (1) the version of the application is not supported on your current operating system.

  the latest compatible version is V11

  http://www.parallels.com/products/desktop/?utm_campaign=sitelink & gclid = CIaxlbL3q s4CFQNsfgodtAcIfg

  If you purchased previous versions of the software, the key to series may be obtained by communicating directly with them.

  If you do not have a key that you will need to buy a newer version of the software or to restore your computer to an instance where he worked under your operating system through Time Machine

  (2) you follow the instructions of apples? If so, what are your light

  Use Time Machine to back up or restore your Mac - Apple Support

• My Yahoo cursor has changed to a diamond with a cross in the middle and nothing in the pick - able (?) groups. How can I change back?

  My Yahoo slider went from arrowhead to a diamond with a cross in the middle and nothing in the pick - able (?) groups. How can I change back? 8.1 (latest) Windows and FireFox (latest).

  You're not referring to auto scroll?

  • Tools > Options > advanced > general: navigation: "use autoscrolling".

• get popups telling post Annie or cross unauthorized

  "Error: privilege or cross post unauthorized exponent" continues to appear when I use Firefox. I think he did an update this morning.
  MSG appears several times supposedly various pages then msg proposes to stop the other pop ups, but it does not stop them.
  Go to MS Explorer. Firefox unusable.

  Make a check of malware with several malware scanning of programs on the Windows computer.
  Please scan with all programs, because each program detects a different malicious program.
  All of these programs have free versions.

  Make sure that you update each program to get the latest version of their databases before scanning.

  • Malwarebytes' Anti-Malware:
    http://www.Malwarebytes.org/MBAM.php
  • AdwCleaner:
    http://www.bleepingcomputer.com/download/adwcleaner/
    http://www.Softpedia.com/get/antivirus/removal-tools/AdwCleaner.shtml
  • SuperAntispyware:
    http://www.SUPERAntiSpyware.com/
  • Microsoft security scanner:
    http://www.Microsoft.com/security/scanner/en-us/default.aspx
  • Windows Defender:
    http://Windows.Microsoft.com/en-us/Windows/using-Defender
  • Spybot Search & Destroy:
    http://www.safer-networking.org/en/index.html
  • Kasperky free Security Scan:
    http://www.Kaspersky.com/security-scan

  Alternatively, you can write a check for an infection rootkit TDSSKiller.

  • Utility anti-rootkit TDSSKiller:
    http://support.Kaspersky.com/5350?El=88446

  See also:

  • "Spyware on Windows": http://kb.mozillazine.org/Popups_not_blocked
  • https://support.Mozilla.org/KB/troubleshoot-Firefox-issues-caused-malware

• every time I erase content and settings on my iphone 6 after reset at the start of my phone and after crossing the screen as Hello > select language > Connect wifi > my iphone doesnot ask me activation lock even if my unit is also found on

  every time I erase content and settings on my iphone 6 after reset at the start of my phone and after crossing the screen as Hello > select language > wifi connection > my iphone doesnot ask me lock activation that says that your iphone is connected with the old apple ID, please enter the id and password

  I always reset on find my optional equipment please tell me how to activate locking activation so that whenever I have factory reset my phone with finding my camera so my phone always ask an old apple and password

  I do not understand your question, but let me go with what I believe. Looks like you entered in iCloud and erased all the content and settings on the iPhone, and once you go by assigning back up again, you do not see something that you expect to see, for example, a request for an Apple ID. When you go through the installation process to select the language, etc., it must, at some point, ask you to identify yourself with your Apple ID. are you not see this?

  It would be better if you try to describe exactly what you do again. Also, without the help of any sign of punctuation, it is difficult to track everything you ask. Try providing the steps of what you do, and then what you see when you get to the point that you believe that something is going wrong. You mention both an old and new Apple ID, which is rather confusing.