Test your hearing across the stereo field!

Discussion in 'Location Recording' started by ATAK, Nov 17, 2005.

  1. ATAK

    ATAK Guest

    Hi There,

    I've recently completed a study on various stereo mic techniques (helped by a couple of you on this forum). To quantify my results a bit i've produced a simple listening test which i've made available on-line. Whilst I was creating the test it occured to me that I could also use the experiment as a rudimentary listening test to compare peoples hearing accuracy accross the stereo field. This way I get to gather my research and eveyone that completes the test will get something out of it because I can e-mail their score back to them against the average of all the other people that have submitted the test.

    If anyone has a spare 20 mins (thats all it will take) i'd be really grateful if you would complete it for me. Have a look at the following fior more details:

    http://

    Many thanks,

    Tom

    P.S. you will need broadband and a pair of stereo headphones to complete the test.
     
  2. DavidSpearritt

    DavidSpearritt Well-Known Member

    Tom, what microphone technique was used to record the test, Dummy Head? Blumlein? and if so what model and make of microphone/s were used. Then I have some more questions. :)
     
  3. ATAK

    ATAK Guest

    ***** Just to point out that I have processed the results of the test and submitted my paper so please don't submit any more results becasue I won't be able to get the results back to you *****

    Dave, to answer your questions, I used spaced omnis- 180cm, the NOS array- 90 deg, 30cms apart (i think, cant remember 100%) and a coincident pair- 90 degrees.
    All the mics were AKG C414's and they were all set up at the same time.

    Next Questions (be gentle with me!)

    Tom
     
  4. DavidSpearritt

    DavidSpearritt Well-Known Member

    Tom

    I am not sure I understand what you are trying to test. I think there is some problems with your test conception. None of those mic techniques will produce a solid stereo image with accurate image positioning. Together with different source strength/volume/frequency components, ie (arbitary percussive impulses), there are way too many variables to make this experiment meaningful.

    I also noted that the noise and balance of the two stereo channels to be out, the left channel is slightly higher in volume. This will also skew the data, confuse the listener and make another unwanted variable.

    What is it that you are trying to measure with this experiment. To improve it I would do the following:

    1. Get the electronics calibrated,, so balance is dead centre.
    2. Use Blumlein mic technique with a matched pair, or second choice a dummy head.
    3. Use a consistent and repeatable sound source, in a good acoustic, ie one without excessive reveberation and standing waves.
     
  5. Calgary

    Calgary Active Member

    Dang cool idea though. :cool: :cool: :cool: :cool: :cool:
     
  6. ATAK

    ATAK Guest

    Ok i'll come clean (although I thought I already did in my first post!) My primary objective was to test the accuracy of reproduced localisation accross the stereo field with various arrays. It's not all that interesting for people to take part in so I thought that I would gather my research by providing this "rudimentary" listening test. It was never meant to be a totally accurate and scientific listening test. I am going to get the results back as an average against other participants scores so the inaccuracies of the arrays would apply to all. Also, I repeated some of the same sounds a few times as a safety check so again, it wouldn't matter how naff the recordings were, someone with accurate hearing should place the same sound in the same location.

    I would be surprised if the left channel were louder as there is no mixing involved. Maybe the sound you heard was one from the left resulting in a greater intensity on that side?
    Anyway, the results I got were surprisingly accurate. I filtered out people who varied above a threshold on the repeated sounds and then processed the rest. The localisation properties of the arrays performed pretty much exactly as expected which was almost a bit of a dissapointment because it left me with nothing new to announce! I just backed up everyone elses work, which I guess shouldn't be much of a surprise in a topic that has been studied for 70+ years!

    Next it's on to 5.1 recording which is proving a little more tricky :?

    Thanks,

    Tom
     
  7. RecTeach

    RecTeach Guest

    Hey man I did the test and really enjoyed it but I never got my results emailed to me. My email is donnie@prestopercussion.com

    Thanks!
     

Share This Page