Measuring Performance

Discussion in 'Archived Threads 2001-2004' started by SimonInd, Jan 9, 2002.

  1. SimonInd

    SimonInd Auditioning

    Joined:
    Sep 10, 2001
    Messages:
    9
    Likes Received:
    0
    I've just completed (apart from painting/veneering) a DIY sub based on the Adire Audio Low Q Sealed Tempest design. It sounds fantastic although I'm sure it occasionally hit's a room peak or something ( Madonna's Ray Of Light is a pretty good test CD for low Bass ). Anyway, I've just got a Radio Shack SPL meter and downloaded some freeware to turn my laptop into a signal generator so that I can chart system performance. My intention is to manually plot a frequency response of the whole system: 10Hz - 500Hz to check for any peaks and measure the sub's performance along with it's integration with my two main speakers. To do this I was going to pump sine waves out of the laptop and measure the SPL at head level in the main listening position. Now I have a couple of questions....

    Lots of posts mention correction values for the SPL meter, do I need to worry about these?

    Also there do seem to be some spectrum analysis packages out there for PCs. Are these a better way of measuring performance or will my old-fashioned approach suffice?

    I'm not interested in measuring the performance next to the cone as it's an established design and I wouldn't have built it if it hadn't been recommended for it's performance, however I do want to find out how good it is in a real-life listening position and system, and if necessary alter the room position to improve (flatten) the response if possible.

    Any advice would be appreciated.....

    Thanks

    Simon
     
  2. Charles L.

    Charles L. Agent

    Joined:
    Dec 21, 2001
    Messages:
    34
    Likes Received:
    0
    If you dont compensate for different values (0, -10, -15, etc.) then theres no point of charting the response. I'm curious on the down loaded sine waves. Can you put them on disc? I've been meaning to get a more complete catalog of sine waves, maybe thats my answer. Good luck
     
  3. SimonInd

    SimonInd Auditioning

    Joined:
    Sep 10, 2001
    Messages:
    9
    Likes Received:
    0
    If you dont compensate for different values (0, -10, -15, etc.) then theres no point of charting the response.
    Sorry Charles, don't really follow you here? I'm referring to the correction values I've seen on the web along the lines of 10Hz +2dB, 15Hz +1.5dB etc. ie. an adjustment to the reading taken by the meter when a particular frequency is being reproduced.
    I'm curious on the down loaded sine waves. Can you put them on disc? I've been meaning to get a more complete catalog of sine waves, maybe thats my answer. Good luck
    I just downloaded a freeware signal generator program that generates single frequencies and sweeps. I figured I'd use this straight out of my Laptop's soundcard - not ideal but I'd hope it would be good enough?
    Simon
     
  4. Ron Shaw

    Ron Shaw Stunt Coordinator

    Joined:
    Dec 4, 2001
    Messages:
    142
    Likes Received:
    0
    Single frequency test tones are not very good for measuring system response, as they set up 'standing waves'. To get a decent measurement, you need to use pink noise and spectrum analysis.
     
  5. BruceD

    BruceD Screenwriter

    Joined:
    Apr 12, 1999
    Messages:
    1,220
    Likes Received:
    0
    Simon,
    Yes you can accomplish your goal with the RS SPL meter, plus the correction factors you mentioned applied per frequency, plus a test signal sound source.
    Three comments:
    1) Sine waves can really damage speaker drivers
    2) You can buy CDs with appropriate test signals like Stryke
    3) You should visit this website for a great introduction on sub placement and measurement:
    ETF analysis software
    BruceD
     
  6. Charles L.

    Charles L. Agent

    Joined:
    Dec 21, 2001
    Messages:
    34
    Likes Received:
    0
    SIMONIND

    let me clarify. you can take 5 sine waves in the same frequecy lets say 1KHZ. Each one is recorded at a different level than the others. -20db, -25db, -10db, etc. Follow me so far. If you played each one with out changing the volume the -10db sine wave would be the loudest and the -25db would be the opposite. If you dont know what you're putting in you dont know what to expect at the output.

    Example my first sine wave is 40herts -30db. it reads 91 db on my spl.

    my second sine wave is 20 herts-20db. it reads 78db on my spl. The spl readings aren't useful for charting purposes unless you compensate (correction of the values) for the differences. To convert it all you do is add or subtract the values to the reference value in this case -30db add 10 to the -20 db sine wave you get -30db now you subtract the same number from your spl reading 78db is now 68db

    I hope that helps.
     

Share This Page