Results 1 to 9 of 9

Thread: Wavelet Analysis vs. CSD

  1. #1
    Administrator Robh3606's Avatar
    Join Date
    Apr 2003
    Location
    Rocinante
    Posts
    8,200

    Wavelet Analysis vs. CSD

    Anyone have any thoughts or preferences?

    I have been looking at CSD plots for quite some time. Wavelets are new to me and appear to have advantages over CSD plots.

    Just running through Clio Pocket 3.0 you can view them in 3 distinct ways.

    Balanced: A combination of Frequency and Time

    Frequency Preferred: Which to me looks like the easiest to understand

    Time Preferred : The oddball to me anyway.

    Here they are posted with the accompanying CSD plot

    All plots labeled.

    Rob
    Attached Images Attached Images     
    "I could be arguing in my spare time"

  2. #2
    Administrator Mr. Widget's Avatar
    Join Date
    Apr 2003
    Location
    San Francisco
    Posts
    9,734
    Quote Originally Posted by Robh3606 View Post
    Anyone have any thoughts or preferences?
    I have lots of random thoughts and preferences, but alas none on this topic.


    Widget

  3. #3
    Administrator Robh3606's Avatar
    Join Date
    Apr 2003
    Location
    Rocinante
    Posts
    8,200
    Quote Originally Posted by Mr. Widget View Post
    I have lots of random thoughts and preferences, but alas none on this topic.


    Widget
    Yeah not exactly what you go down to the corner pub and discuss!

    Rob
    "I could be arguing in my spare time"

  4. #4
    Senior Member
    Join Date
    Jul 2003
    Location
    Orange County, CA
    Posts
    3,603


    Where is says, "Spectra No. 60", are we to infer that the plot is made of 60 points across the spectrum and everything in between is interpolated?

    If this be the case, why are you guys always raggin' on me for using a 61-band RTA? It is certainly easier to implement and interpret.

  5. #5
    Administrator Mr. Widget's Avatar
    Join Date
    Apr 2003
    Location
    San Francisco
    Posts
    9,734
    Quote Originally Posted by toddalin View Post

    Where is says, "Spectra No. 60", are we to infer that the plot is made of 60 points across the spectrum and everything in between is interpolated?

    If this be the case, why are you guys always raggin' on me for using a 61-band RTA? It is certainly easier to implement and interpret.
    Apples and oranges.

  6. #6
    Administrator Robh3606's Avatar
    Join Date
    Apr 2003
    Location
    Rocinante
    Posts
    8,200
    Hello Toddalin

    That is correct. In Clio Pocket you have a choice of 30, 60, 120. I typically use the 60 as it easier to see than the 120 which can get densely packed depending on how long you set the Window. You typically remove the delay caused by distance at the beginning and then go out to your first reflection. That set's your Window/gate time in this case 5.4 msec

    This colorized plot fills the blanks between the individual slices in time. Those slices are 91us each so not a lot of time. You can also look at the CSD without the colorization where you can see each individual slice as a stand alone.

    Each time slice looks at the entire selected frequency band.

    You can also use a cursor where you can pick the frequency and look at each time slice for level individually.

    Even though I have Clio I still use my RTA. You use Clio to get a raw measurement and load that into Leap. With Leap every part change can be seen before you do it so it's really easy and faster compared to doing changes in real time. I can go through a half dozen part changes in a matter of minutes using Leap.

    Here is a "Raw" CSD plot. No filling in the blanks here. Some prefer this over the colorized version.

    Gated Impulse/Window the gate time beginning and end are noted in the red and green text The lines are hard to see in this view.

    Rob
    Attached Images Attached Images   
    "I could be arguing in my spare time"

  7. #7
    Administrator Mr. Widget's Avatar
    Join Date
    Apr 2003
    Location
    San Francisco
    Posts
    9,734
    Quote Originally Posted by Robh3606 View Post
    Here is a "Raw" CSD plot. No filling in the blanks here. Some prefer this over the colorized version.
    Actually, I do have an opinion on that. I prefer the non-colorized versions... but I may have a minority opinion there.


    Widget

  8. #8
    Administrator Robh3606's Avatar
    Join Date
    Apr 2003
    Location
    Rocinante
    Posts
    8,200
    Yes I can see why. I like the color because the coloring makes it easier to see how quickly you are rolling off and what level the "noise" is at as you move away in time.

    It's a preference actually make more sense to post both come to think of it.

    Rob
    "I could be arguing in my spare time"

  9. #9
    Administrator Robh3606's Avatar
    Join Date
    Apr 2003
    Location
    Rocinante
    Posts
    8,200
    Interesting how broadband the noise is in your room. Was looking at in room noise in impulse response measurements. Here is a quick snippet limited by the gate window size.

    Rob
    Attached Images Attached Images    
    "I could be arguing in my spare time"

Thread Information

Users Browsing this Thread

There are currently 1 users browsing this thread. (0 members and 1 guests)

Similar Threads

  1. 2447J Measurements, help with analysis?
    By emilime75 in forum Lansing Product Technical Help
    Replies: 2
    Last Post: 02-19-2017, 08:58 PM
  2. Horn Loudspeaker Response Analysis Program
    By Hoerninger in forum General Audio Discussion
    Replies: 15
    Last Post: 07-26-2015, 07:38 AM
  3. Fourier Analysis
    By 1audiohack in forum General Audio Discussion
    Replies: 1
    Last Post: 12-12-2014, 07:56 PM

Posting Permissions

  • You may not post new threads
  • You may not post replies
  • You may not post attachments
  • You may not edit your posts
  •