Announcement

Collapse
No announcement yet.

Directivity Errors - impact?

Collapse
X
 
  • Filter
  • Time
  • Show
Clear All
new posts

  • Directivity Errors - impact?

    I've been reading a lot of ASR lately, and quite like Amir's approach (in general), but I really struggle to understand why he places so much importance on directivity. I understand the argument for room independence, but I'm not sure that I've experienced issues with directivity, or if I have, they haven't negatively impacted my enjoyment of the music, and I've listened to speakers that surely have fairly large directivity errors - including the speakers I'm listening to now with an 8" woofer and a 1" tweeter, which while crossed low, probably isn't crossed low enough for Amir.

    In your experience, what is the impact? To be clear, I'm not asking about obvious cases of beaming, like crossing a 6.5" driver at 4Khz, but rather the impacts of directivity error on what would otherwise be considered a reasonable XO point in the DIY world - say low 2k ish.

    Basically, I'm looking for an excuse to play with a waveguide, with all of the pluses and minuses that entails, and this seems like a good excuse to try one out, but with the relative difficulties of getting waveguides right, even well-designed ones, I'm a bit hesitant whether the juice is worth the squeeze in this case.


  • #2
    IME so long as you don't go more than -6dB at 30 degrees off-axis you're good. If you can get it better than that fine, but I wouldn't jump through hoops to do it.
    www.billfitzmaurice.com
    www.billfitzmaurice.info/forum

    Comment


    • #3
      I listened to the interview of Sean Olive by Erin

      https://youtu.be/IEtYH03pfOI

      From 1:03 Sean talks about directivity and it’s level of importance depends on the room, number of channels and the listening tasks

      Comment


      • #4
        Directivity is just one factor, overall speaker in-room response, power response, listening window, and resulting directivity index all play important roles in what you hear. It's important to consider the entire sound field presented by a speaker system, not just a single axis of measurement, and research completed over the years by people like Sean Olive, Floyd Toole, the team at Klippel and others have all been extraordinarily valuable in the interpretation of overall speaker performance.

        In recent years a in-room measurement standard has been developed, CTA-2034-A, which is made available for free here:
        https://shop.cta.tech/products/stand...e-loudspeakers

        This document defines a standard representation for processing off-axis measurement data down to the "spinorama" you may see from Klippel measurements at ASR or Erin's Audio Corner, that include things like early reflections, in-room response and sound power.

        For DIYers, VituixCAD is available with instructions for REW, ARTA, or SoundEasy on how to measure speakers properly and process the measurements in VituixCAD to design a speaker using complete accurate off-axis information. In my own experience, I had one speaker where the design was completed in SoundEasy using simple on-axis measurement, and off-axis simulation was by crossover polar pattern only. This speaker was enjoyable, not not all it could be. I later redesigned the crossover by utilizing complete off-axis data set and VituixCAD, optimizing the design for a balance between the listening window and in-room response. The difference in result was not subtle, and this speaker is now my favourite to listen to of anything I've built.

        When designing a speaker, balance of sound power vs on-axis etc is a bit more involved than simply slapping a waveguide on the tweeter. It is a balance of several factors including baffle shape, driver separation, crossover frequency, and of course each driver's individual response pattern. Even without a waveguide using just a standard dome tweeter with flat faceplate, adjusting crossover slope, driver separation, and crossover frequency can provide good balance of on-axis and off-axis performance. Kimmo (VituixCAD developer) has explained that a basic rule of thumb can be to keep driver centre to center distance at 1.0 - 1.4 times the crossover frequency, 1.2x being ideal. This may appear counter-intuitive to the common design rule provided by many DIYers that drivers closer together is best. Closer may provide a wider vertical polar pattern of the crossover, but may actually provide a very poor power response result depending on the crossover frequency implemented.

        As far as waveguides go, when implemented properly they can provide better balance of sound power vs on-axis response, as baffle diffraction effects are reduced, and can provide some benefit in distortion performance as well. You will see this benefit easily represented in the directivity index chart in VituixCAD. So by all means feel free to play around with them. Not needing to re-invent the wheel - there are several plug and play solutions available, such as Vitaton WG148, and several 3d printable designs from Augerpro that are freely available. Have fun!


        "I just use off the shelf textbook filters designed for a resistor of 8 ohms with
        exactly a Fc 3K for both drivers, anybody can do it." -Xmax

        Comment


        • #5
          I think it is helpful to have a smooth directivity curve and/or power response with no sharp edges or major bumps. It is, of course, only one part of the overall equation.

          Comment

          Working...
          X