Announcement

Collapse
No announcement yet.

fdieck "sounded like music not loudspeakers !"

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

  • fdieck "sounded like music not loudspeakers !"

    fdieck said this about some speakers that were at LSAF in Dallas this past weekend .i dont remember ever hearing those words together before or maybe just not used that way . it got me to thinking if maybe sometimes in the design ,crossover,driver choice that we get away from making music and are just making speakers .im looking for a speaker to do this , im looking for a speaker that does that ,my graphs are to look like this or that ! maybe these are all one in the same . just a thought !?
    donc

  • #2
    Careful driver selection is very important for natural sound. I like to start with drivers that sound decent with a minimal or no crossover. I also like drivers that pull me in and I can just sit and listen for extended periods.

    Comment


    • #3
      Originally posted by Billet View Post
      Careful driver selection is very important for natural sound. I like to start with drivers that sound decent with a minimal or no crossover. I also like drivers that pull me in and I can just sit and listen for extended periods.
      You took the words out my mouth. This is also how I modify drivers by listening with a single capacitor cross over for mids and tweeters.*

      Comment


      • #4
        Originally posted by Billet View Post
        Careful driver selection is very important for natural sound. I like to start with drivers that sound decent with a minimal or no crossover. I also like drivers that pull me in and I can just sit and listen for extended periods.
        So, which ones are those?

        Comment


        • #5
          Originally posted by skatz View Post

          So, which ones are those?
          Kraco.

          Comment


          • #6
            Originally posted by donc View Post
            fdieck said this about some speakers that were at LSAF in Dallas this past weekend .i dont remember ever hearing those words together before or maybe just not used that way . it got me to thinking if maybe sometimes in the design ,crossover,driver choice that we get away from making music and are just making speakers .im looking for a speaker to do this , im looking for a speaker that does that ,my graphs are to look like this or that ! maybe these are all one in the same . just a thought !?
            donc
            First thing is to forget about that whole flat on axis response thing, for most speakers anyway. Put your focus into the directivity index, flat frequency response is only correct if the power response is flat as well. So aim for a constant directivity, or plan to compensate for the change in directivity with a non-linear frequency response. Then aim for low distortion, especially from 300Hz to 7kHz, and an impulse/CSD that stops on a dime. When you combine the power response, lack of distortion and transient response together, you will start to forget about the speaker and just hear the recording.
            "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


            • #7
              Originally posted by dcibel View Post

              First thing is to forget about that whole flat on axis response thing, for most speakers anyway. Put your focus into the directivity index, flat frequency response is only correct if the power response is flat as well. So aim for a constant directivity, or plan to compensate for the change in directivity with a non-linear frequency response. Then aim for low distortion, especially from 300Hz to 7kHz, and an impulse/CSD that stops on a dime. When you combine the power response, lack of distortion and transient response together, you will start to forget about the speaker and just hear the recording.
              +1 The flat, smooth power response is important. Once I understood that, I finally made progress getting the sound I wanted.
              Francis

              Comment


              • #8
                how we do verify flat power response from measurements?

                Comment


                • #9
                  Originally posted by ani_101 View Post
                  how we do verify flat power response from measurements?
                  It is generally represented by analyzing the off-axis response. For Omnimic users, it is easily done by taking the on and off-axis measurements, then normalizing to the on-axis response to remove the speakers natural response from the chart, showing only the change in output at different angles. This can then be plotted as a nice colourful polar chart.

                  The power response is basically an average response of all angles. Once you see the polar chart, you won't have a need to go a step further and combine it down to a single response plot, the "problems" will be apparent in the polar chart. However, Omnimic also features the ability to average many response curves, and even apply different weighting to specific curves, so you generate a plot averaging your on and off-axis response curves, providing a higher priority to the forward facing angles, than say the 75-90 degree response. Here's a plot where I did just that. Black is the on-axis, and red is the weighted average. As you can see I have some excess energy in the 2-4kHz range.

                  Click image for larger version  Name:	weighted average.png Views:	1 Size:	14.1 KB ID:	1375107
                  "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


                  • #10
                    About 25 years ago I "designed" a speaker with all sorts of notch filters and LCR traps it was the flattest*
                    ON AXIS speaker I have ever measured, although it sounded like the worst speaker I ever heard.
                    The problems were easily visible about 15 degrees above and below the tweeter. That's a great*
                    place to start.**
                    Guess xmax's age.

                    My guess: 15. His grammar is passable. His trolling is good.

                    Comment


                    • #11

                      Comment


                      • #12
                        "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


                        • #13
                          A good "lesson learned" I'm sure that the output must be analyzed at multiple angles, at least 30-45 degrees out.
                          "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


                          • #14
                            Originally posted by dcibel View Post

                            A good "lesson learned" I'm sure that the output must be analyzed at multiple angles, at least 30-45 degrees out.
                            Click image for larger version  Name:	Lobbing .jpg Views:	2 Size:	248.0 KB ID:	1375139
                            Guess xmax's age.

                            My guess: 15. His grammar is passable. His trolling is good.

                            Comment


                            • #15
                              The horn itself has very little loss, so what we're dealing with here is the power response of the driver on a particular horn. The horn can (and often does) influence the impedance of the driver, and therefore the power vs frequency, so the plane wave response of the driver is only a start. It really helps to know the DI of the horn vs frequency, to get things sorted out.
                              Francis

                              Comment

                              Working...
                              X