Harmony Central Forums
Announcement
Collapse
No announcement yet.

Behringer DSP1124P Feedback Destroyer

Collapse



X
  • Time
  • Show
Clear All
new posts

  • #16
    (BTW: technically speaking, you should say "Fixed-Point" and NOT fixed precision. IEEE floating point is fixed precision. Nitpicky, I know)
    You've never heard of fixed point arithmetic before?
    http://www.dsprelated.com/showmessage/23752/1.php

    Comment


    • #17
      You've never heard of fixed point arithmetic before?
      http://www.dsprelated.com/showmessage/23752/1.php

      Of course. Maybe I misread: I could have sworn you said "fixed precision".

      I had an old algorithm book that gave a really clear explanation and code examples: it's actually pretty simple to implement in code, and I learned how to do it back when I was serious about being a game developer.

      Of course a fixed point DSP implements fixed point in hardware, but the concept is going to be the same.

      BTW, that dsp page is quite interesting: I can relate to the quantization step that is mentioned in one of the posts.

      Sort of reminds me of Bresenham's anti-aliasing algorithm for lines and circles, whereby an "error term" is accumulated and then subtracted off at each step: same concept anyway, if not the same implementation.
      Kickin' it in the sticks...

      Comment


      • #18
        Of course. Maybe I misread: I could have sworn you said "fixed precision".
        Nope.
        I had an old algorithm book that gave a really clear explanation and code examples: it's actually pretty simple to implement in code, and I learned how to do it back when I was serious about being a game developer.
        Yah, games used to have to do it all fixed point before PCs had floating point as standard. Game programmers used to cheat horribly to get acceptable performance - can't really do that with audio .

        That reminds me, there is still one of my shareware proggies "floating" about the 'net for those interested:
        http://www.hp200lx.net/anonftp/pub/em87.zip

        Comment


        • #19
          Game programmers used to cheat horribly to get acceptable performance - can't really do that with audio .

          It's not cheating, it's "optimization".


          Kickin' it in the sticks...

          Comment


          • #20
            It's not cheating, it's "optimization".
            Did you say "autotune" ?

            Comment


            • #21
              Did you say "autotune" ?

              God, talk about quantization (fail)....

              (not knocking the engineering, just the results as applied to horribly out of tune input signals. Polished turds....)
              Kickin' it in the sticks...

              Comment


              • #22
                i have read about a program called REW(room eq wizard) being used with these behringer units as auto room analysis and auto correction using the filters in fixed mode.i think mainly for home audio use,but might be fun to play with.u might be able to use that program to manually config the device to ur desires.
                cheers,,,,dave

                ***disclaimer: views expressed by davidss1 are not an attempt to generate sales for CMS.

                www.computer-mix-systems.com

                Comment


                • #23
                  i have read about a program called REW(room eq wizard) being used with these behringer units as auto room analysis and auto correction using the filters in fixed mode.i think mainly for home audio use,but might be fun to play with.u might be able to use that program to manually config the device to ur desires.
                  Yah, kinda scary that them picky audiophiles think the DSP1124P is good enough to put into their systems . Maybe 'ringer got something right here at least .

                  Comment


                  • #24
                    When the filters are Locked they were locked because they were in Single Shot mode when set. Just select the fitler that is locked and put it in Single Shot mode again.
                    I have 12 fixed filters - to go do that to each one would be quite time consuming . For those interested I've set mine up with each channel having 3 parametric filters, 6 "single shot" filters, and 3 "auto" filters.

                    Comment


                    • #25
                      I was pointing out SOME of the various ways that DSP can be used beyond it's capabilities. Many of these terms have taken on different or multiple meanings so folks discussing this should remember that there are many, many ways to exceed design capabilities when you go outside the typical ranges that programmers typically allocate resources towards. You will always want to double-check the programming with an actual signal transfer function test.

                      Since DSP has progressed so far beyond where I stepped off the design bandwagon, techniques, software tools and hardware capabilities have exploded. Since fewer and fewer analog guys are coming out of the universities, I find myself more deeply involved with developing and design wher I am needed most and currently that is analog. Most of the analog guys my age have retired from engineering or moved into sales or management. My semi-retired partner just turned 70 today and he's one of the best analog AND digital engineers I have ever worked with. He worked developing and implimenting data acquisition systems for the Apollo project and he's an engineer's engineer, just not really creative in audio.
                      -------------------------------------------------------------------------------------------------------------------------------------------------------------------
                      Former product development engineer: Genz Benz, a KMC Music/FMIC/JAM Industries Company, continuing factory level product support and service for Genz Benz

                      Currently product development engineer: Mesa Boogie

                      Comment


                      • #26
                        Last DSP programming I worked on was for the F-35 :



                        http://www.jsf.mil/
                        Woo Hoo !

                        Them guys were REALLY anal about testing boundary conditions for sure .

                        Comment


                        • #27
                          My college roommate was an aeronautical/mechanical engineer and went into the AF as an engineer/test pilot. The other engineer here was an instrumentation engineer for NASA on the Apollo projet. The parent company I work for had an aerospace division as well, and there's a darn good reason for why design and testing procedures are so thorough and complex. There's a lot that can go wrong, and there's a lot of possible conditions to consider, so testing involved a wide range og both conditions and variables... a lot of permutations.
                          -------------------------------------------------------------------------------------------------------------------------------------------------------------------
                          Former product development engineer: Genz Benz, a KMC Music/FMIC/JAM Industries Company, continuing factory level product support and service for Genz Benz

                          Currently product development engineer: Mesa Boogie

                          Comment


                          • #28
                            I hooked it up to my laptop today using "Visual Analyzer":
                            http://www.sillanumsoft.org
                            The frequency response of the 3 filters I had was quite a bit off from what Behringer's software predicted so I tweaked it in using pink noise. I ended up with these three:
                            20Hz 37/60 octave -48db
                            20Hz 37/60 octave -48db
                            100Hz 80/60 octave +16db
                            Looks like a very nice 24db/oct 100hz HPF now .

                            However I'm a bit perplexed as to the locked filter behavior I saw the other day. Now any locked filters return to single shot mode upon a short power cycle or if you switch to another preset and then switch back. Neither of those actions would clear them before .

                            Comment


                            • #29
                              I hooked it up to my laptop today using "Visual Analyzer":
                              http://www.sillanumsoft.org
                              The frequency response of the 3 filters I had was quite a bit off from what Behringer's software predicted so I tweaked it in using pink noise. I ended up with these three:
                              20Hz 37/60 octave -48db
                              20Hz 37/60 octave -48db
                              100Hz 80/60 octave +16db
                              Looks like a very nice 24db/oct 100hz HPF now .

                              However I'm a bit perplexed as to the locked filter behavior I saw the other day. Now any locked filters return to single shot mode upon a short power cycle or if you switch to another preset and then switch back. Neither of those actions would clear them before .


                              Does your software show phase as well? 2 48db per octave filters in series gives you a 96db per octave slope. The amount group delay may be quite great. I wonder if it rings as well?

                              Dookietwo

                              Comment


                              • #30
                                Ah, I was wondering when the "phase police" would raid this thread .
                                Does your software show phase as well?
                                No.
                                2 48db per octave filters in series gives you a 96db per octave slope. The amount group delay may be quite great. I wonder if it rings as well?
                                They are not 48 db/oct filters - they are bandpass filters with a 20Hz center frequency, a 37/60 octave bandwidth, and set for -48db of cut.

                                I suspect the phase gets pretty squirrelly at and below 100hz but as it's not a crossover there isn't another driver "below" its frequency cutoff that I have to match up to so I'm not all that worried about it .

                                Comment













                                Working...
                                X