Announcement

Collapse
No announcement yet.

Please help Z axis calculator

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

  • Please help Z axis calculator

    On the Z axis calculator in Omnimic I don't get the measurements it wants with a tape measure.

    Am I measuring distance from mic to baffle, mic to driver A center and mic to driver B center?
    Or and I measuring distance between driver centers?
    Or what?

    It would help if there were some clearly defined arrows indicating which distances to enter.
    Tom

    NEW CNC Site http://cnc.trbailey.net/
    My Projects BLOG http://speakers.trbailey.net/

  • #2
    Re: Please help Z axis calculator

    I have not tried doing this in omnimic. I use this method to find my z offsets.
    https://www.box.com/shared/ouxjjsx0m8bs00cil5iq
    Jeff

    Comment


    • #3
      Re: Please help Z axis calculator

      Originally posted by peepaj View Post
      I have not tried doing this in omnimic. I use this method to find my z offsets.
      https://www.box.com/shared/ouxjjsx0m8bs00cil5iq
      I'd rather not have to load PCD if I can help it. The Omnimic version is faster and probably more accurate.
      Tom

      NEW CNC Site http://cnc.trbailey.net/
      My Projects BLOG http://speakers.trbailey.net/

      Comment


      • #4
        Re: Please help Z axis calculator

        Originally posted by siggma View Post
        I'd rather not have to load PCD if I can help it. The Omnimic version is faster and probably more accurate.
        It is the same method - it is not more accurate.

        Personally, I find it easier to do in PCD, I have trouble making it work in OmniMic. Someone else may prefer doing it in OmniMic, It's a matter of preference. But if you don't like PCD, then don't use it.
        Click here for Jeff Bagby's Loudspeaker Design Software

        Comment


        • #5
          Re: Please help Z axis calculator

          Originally posted by Jeff B. View Post
          It is the same method - it is not more accurate.

          Personally, I find it easier to do in PCD, I have trouble making it work in OmniMic. Someone else may prefer doing it in OmniMic, It's a matter of preference. But if you don't like PCD, then don't use it.
          Basically I can't make it work at all in Omnimic. The instructions are nebulous at best and poorly written on a good day. It's not clear what they want in the little spinner boxes. No arrows or explanation just "get out the tape measure and put the data in the box".

          AND whoa there Mr Data . I did not say I didn't like PCD. You've done an excellent job putting together a superb spreadsheet to solve some rather complex procedures. It's not easy with all those macros. I know how hard it is, I was writing reports and scripts for spreadsheets in the early 80's back when "Symphony" was King. The corporate controller used "Symphony" exclusively for his accounting and report writing so I got stuck figuring out how to write corporate reports with it. I might suggest you move some of the "external" functions to a different spreadsheet and "chain" load the other sheets on demand rather than trying to put it all in a single sheet. It would definitely re-calc faster. Functions like the extra filter calculators, maybe separate out active from passive calculators etc. Just an idea. The reality is I've seen other programs that try to do the same thing as PCD and they cost money and don't work. Bass box pro and Xover 3 for example. Very poorly written software. Thank you!
          Tom

          NEW CNC Site http://cnc.trbailey.net/
          My Projects BLOG http://speakers.trbailey.net/

          Comment


          • #6
            Re: Please help Z axis calculator

            i have only tried the omnimic method once and got so frustrated i gave up. think we need a vidio totorial.
            hello bill.
            " To me, the soundstage presentation is more about phase and distortion and less about size. However, when you talk about bass extension, there's no replacement for displacement". Tyger23. 4.2015

            Quote Originally Posted by hongrn. Oct 2014
            Do you realize that being an American is like winning the biggest jackpot ever??

            http://www.midwestaudioclub.com/spot...owell-simpson/
            http://s413.photobucket.com/albums/pp216/arlis/

            Comment


            • #7
              Re: Please help Z axis calculator

              Originally posted by [email protected] View Post
              i have only tried the omnimic method once and got so frustrated i gave up. think we need a vidio totorial.
              hello bill.
              I couldn't make it work for me either. Since the PCD method was just too easy to pass-up I just stuck with it. When I model based the offset I get, the final measured results match perfectly. That's what you're after.

              Jeff
              Click here for Jeff Bagby's Loudspeaker Design Software

              Comment


              • #8
                Re: Please help Z axis calculator

                I couldn't make it work for me either. Since the PCD method was just too easy to pass-up I just stuck with it. When I model based the offset I get, the final measured results match perfectly. That's what you're after.

                Jeff
                +1
                Jeff

                Comment


                • #9
                  Re: Please help Z axis calculator

                  Originally posted by Jeff B. View Post
                  I couldn't make it work for me either. Since the PCD method was just too easy to pass-up I just stuck with it. When I model based the offset I get, the final measured results match perfectly. That's what you're after.

                  Jeff
                  i will try the jb method again when the oportunity presents.
                  " To me, the soundstage presentation is more about phase and distortion and less about size. However, when you talk about bass extension, there's no replacement for displacement". Tyger23. 4.2015

                  Quote Originally Posted by hongrn. Oct 2014
                  Do you realize that being an American is like winning the biggest jackpot ever??

                  http://www.midwestaudioclub.com/spot...owell-simpson/
                  http://s413.photobucket.com/albums/pp216/arlis/

                  Comment


                  • #10
                    Re: Please help Z axis calculator

                    Any reason I can't use the impulse response method?

                    Click image for larger version

Name:	Two Impulse.jpg
Views:	2
Size:	47.0 KB
ID:	1146110

                    I know the flight time is accurate I've measured it with a tape. 1.615ms is 21.808858 inches and I measured roughly 22.5 from baffle with the tape measure. If I take a measurement of driver A then save it then take a measurement of driver B and load driver A as an overlay I can see both IR at the same time. Sliding the cursor (in ARTA) gives me time and voltage of both current measurement and overlay on the bottom so it's easy to find the peak sample of each individual impulse. In the image below I measure 1.583ms flight time for the tweeter and 1.615ms flight time for the woofer. That gives me 0.032ms arrival difference which coverts to 0.010976 meters or roughly 11 mm. This is a shallow woofer and the tweeter is not flush mounted so it seems reasonable. It also correlates with the tweeter being surface mounted. When I use PCD I get two FR lines that never match up like in the example. It usually will only match over a very narrow range above 4K or so. This feels more accurate to me. whether it is or not I can't say yet. I'll try it with the few drivers I have and see if it works for them as well.

                    Click image for larger version

Name:	PCD impulse.jpg
Views:	1
Size:	116.0 KB
ID:	1146111 This is what it looks like in PCD. Changing the value does little to get them any closer. It just creates peaks and dips.

                    Obviously there is no exact acoustic center in either time or frequency domain of a driver or the impulse would be a single vertical line and only the shape of the impulse after arrival would be different between drivers (roughly representing it's frequency content). I would think it would be more accurate, especially since phase is calculated from the impulse. Can anyone answer if impulse arrival is an acceptable way to calculate Z offset or exactly what part or aspect of the driver this measures?

                    Tom
                    Tom

                    NEW CNC Site http://cnc.trbailey.net/
                    My Projects BLOG http://speakers.trbailey.net/

                    Comment


                    • #11
                      Re: Please help Z axis calculator

                      Originally posted by siggma View Post
                      Any reason I can't use the impulse response method?

                      [ATTACH=CONFIG]32628[/ATTACH]

                      I know the flight time is accurate I've measured it with a tape. 1.615ms is 21.808858 inches and I measured roughly 22.5 from baffle with the tape measure. If I take a measurement of driver A then save it then take a measurement of driver B and load driver A as an overlay I can see both IR at the same time. Sliding the cursor (in ARTA) gives me time and voltage of both current measurement and overlay on the bottom so it's easy to find the peak sample of each individual impulse. In the image below I measure 1.583ms flight time for the tweeter and 1.615ms flight time for the woofer. That gives me 0.032ms arrival difference which coverts to 0.010976 meters or roughly 11 mm. This is a shallow woofer and the tweeter is not flush mounted so it seems reasonable. It also correlates with the tweeter being surface mounted. When I use PCD I get two FR lines that never match up like in the example. It usually will only match over a very narrow range above 4K or so. This feels more accurate to me. whether it is or not I can't say yet. I'll try it with the few drivers I have and see if it works for them as well.

                      [ATTACH=CONFIG]32629[/ATTACH] This is what it looks like in PCD. Changing the value does little to get them any closer. It just creates peaks and dips.

                      Obviously there is no exact acoustic center in either time or frequency domain of a driver or the impulse would be a single vertical line and only the shape of the impulse after arrival would be different between drivers (roughly representing it's frequency content). I would think it would be more accurate, especially since phase is calculated from the impulse. Can anyone answer if impulse arrival is an acceptable way to calculate Z offset or exactly what part or aspect of the driver this measures?

                      Tom
                      As I read this I get the feeling you may be missing the point. This is not about finding the acoustic center of the drivers, it is about finding the relative acoustic offset that gives a mathematical summation that matches the measured summed response you have already taken. In doing so, you assure yourself that your simulation in PCD will ultimately match the final measured response. If you can not get your response curves to match in PCD by entering offsets then you are essentially assuring yourself that your simulation will not match the actual measured response, and so then, what is the use in doing it?

                      I can give multiple examples of everything from my Triton where the midbasses are only 6mm behind the tweeter, to the Sopranos with a 17mm offset, to the Meniscus Caritas with 24mm offset, to speakers like the one I did for Eminence where the horn tweeter actually 77 mm behind the woofer and the Zephyer where the tweeter is 32mm behind the woofer, yet in every case I could get the curves to completely and perfectly overlay my measured summed response with almost no deviation across the entire frequency range, just like I show in my white paper on using PCD to find the acoustic offset.

                      The result of doing this, which I can demonstrate, is that the resultant frequency response in the PCD simulation will exactly the match the final measured system response once the crossover is built and tested. This is what you are after. You want your prediction to be valid.

                      If you are not able to do this with the phase data you are getting from your individual measured responses, then I suggest you use minimum phase data extracted from both frequency responses, and then using this data find the offsets that result in the matching your summed measurement.

                      I know with OmniMic I no longer extract minimum phase data at all. I simply use the measured phase from OmniMic and follow the steps I already detailed in my paper to get a perfect match simply be entering offsets. However, I do not know how other measurement systems may change all of this.

                      Jeff
                      Click here for Jeff Bagby's Loudspeaker Design Software

                      Comment


                      • #12
                        Re: Please help Z axis calculator

                        I was playing around with my new OmniMic after work and I was just about to post a new thread here asking if anyone has used that z-offset calculator in OmniMic. Glad I decided to look here first and find this thread.

                        I have used Jeff's method with PCD and just like him my final measured results matched perfectly with PCD's prediction. The first time I tried it though I wasn't so sure it was right (or I did it right). On that project (my Rushers) the results showed a 32 mm z-offset between the Usher 8945A and the RS28F. Pretty much everywhere I've seen people say to estimate about 25 mm for a 6-7" woofer and flush mounted 1" tweeter. However final measurements confirmed that the simulated FR and phase were dead on using 32mm in PCD.

                        After reading the posts above, I will definitely stick to his method as it works and it is very straight forward.
                        Craig

                        I drive way too fast to worry about cholesterol.

                        Comment

                        Working...
                        X