Announcement

Collapse
No announcement yet.

DAYTON AUDIO IMM-6 vs YPAO microphone

Collapse
This topic is closed.
X
X
 
  • Filter
  • Time
  • Show
Clear All
new posts

  • #31
    Stash (looks like you've only been around here about 6 mos?), you DO realize that by using some free software (I use ARTA, but there are several others) and a "calibrated" mic, that most guys on here DO take "accurate semi quasi anchoic measurements" in their living rooms? Using "gated" measurements, you can can get anechoic results down to about 300Hz (w/a gate of about 3.3ms) as long as you measure right in the center of a room no smaller than 8 x 8 x 8 ft (maybe 8 x 8 x 10', w/the mic-tweeter axis running diagonally across the room). That's really what a large portion of this board is all about - and the way almost all of the bettter systems (many of which are offered by PE as well as other vendors) have been designed.

    I don't have your Yamaha mic. The curves I posted were made from the 2 curves (RTA?) that YOU posted. Those are your mics.
    If you care to post your imm-6 cal file here, I can make you a cal file for your YPAO so they'd (theoretically) match (at least they should match under YOUR measuring conditions - that you used).

    Comment


    • #32
      Chris,

      Haha. I guess I'm behind the times. I've read of quasi anechoic measurements. I assumed they were more fantasy and wishful thinking under extremely rigid conditions. I did understand such measurements could be had down to a few hundred hertz and near-field techniques would be used to approximate the balance of the low end. Although discussions of baffle diffraction effects and how to adjust the curve to compensate is still a bit beyond me.

      i will post the calibration file tomorrow. Some of what you and others do here goes well beyond what limited knowledge I possess. I am up to learning. Thank you for the offer.

      I'm old, and maybe stubborn at times. But I always enjoyed math and electronics. And I've been fascinated by audio since the late 70's. The 80's were a trip with regards to hifi. I did take trig and a little calculus many many years ago. I would like to better understand some of the concepts discussed nowadays.

      Thank you for the offer, Chris.

      Comment


      • #33
        So wait... Chris Roemer, you interpolated and graphed the original images posted?

        Also, Stash here's how you do quasi-anechoic measurements: http://audio.claub.net/software/FRD_...0to%2010Hz.pdf
        Electronics engineer, woofer enthusiast, and musician.
        Wogg Music
        Published projects: PPA100 Bass Guitar Amp, ISO El-Cheapo Sub, Indy 8 2.1 powered sub, MicroSat, SuperNova Minimus

        Comment


        • #34
          I just "PrtScr"'d Stash's plots into MSPaint, drew a (pink) line from band to band (top, centers), pasted that into FPTrace and traced.
          Then I boosted the lower mike by +7dB (to align them somewhat) and brought both into an XO sim. Easy schmeezy.

          Comment


          • #35
            Originally posted by Chris Roemer View Post
            I just "PrtScr"'d Stash's plots into MSPaint, drew a (pink) line from band to band (top, centers), pasted that into FPTrace and traced.
            Then I boosted the lower mike by +7dB (to align them somewhat) and brought both into an XO sim. Easy schmeezy.
            Very interesting. It's really difficult to glean those differences in an RTA bar chart, which I had also interpreted as having minimal differences.

            Stash Definitely well into flogging a dead horse territory here. The differences in the mics are larger than it seems, however for your described purposes I think you're fine IMO. In the event you decide to dive into crossover design with measurement data, you may want something better. For your consideration...
            Electronics engineer, woofer enthusiast, and musician.
            Wogg Music
            Published projects: PPA100 Bass Guitar Amp, ISO El-Cheapo Sub, Indy 8 2.1 powered sub, MicroSat, SuperNova Minimus

            Comment


            • #36
              Wogg,
              That is still a bit over my head. I won't be building anything else for a while anyway.

              When I put together plans for my redesigned MACH ONES, I referred to the original's crossover. Using the ersa crossover calculator online I designed a Riley-Linkwitz 3-way 2nd order. I like the R-L designs as they supposedly offer flatter response at crossover points, given the drivers are the same SPL at those points. Was not expensive.

              For the woofer, 1200Hz LP, on PE I found terminal cups with 2.22mH coils in closeouts for about $2.50 a pop, which are very close to the specified 2.25mH. I rewired the coils to run in series with positve terminal. They were originally wired to negative terminal. I used 17uf electrolytic NP caps, less than $1 ea on PE, pretty close to the specified 17.6uf, wired in parallel to the woofer.

              Original MACH ONE mids crossed over at 800Hz to 8000Hz. I found a Timpano driver with a usesble response of 500Hz to 8000Hz, very close match. 2.25mH with 17.6uf again. for low crossover. I did same as I had done for the woofer, removing the 2.22mH coil from its terminal cup. Crossover calculator specified high crossover cutoff as 0.23mH and 1.76uf. Coils not in stock at PE. I ordered 0.27mH air coils. Close enough for me. Caps are 1.5uf, which I made by combing a couple of caps I had on-hand. I include a L-pad to turn down the output of mid horn without affecting its response.

              For super tweeters I found Pyramid horns, 4 ohms. Nice quality. I like Pyramid. I know they are affiliated with Pyle. But Pyramid once had a good reputation. Not a fan of Pyle products.
              made a HP filter at 8000Hz by ordering an assembled HP crossover from PE. Originally designed to be 3 5Khz at -12dB into 4 ohms. Consists of 0.21MH coil and a 3.3uf cap.
              So I modified it. I added a 4 ohm resistor in series with the tweeter, to make it an 8 ohm driver. Then I added the 3.3uf caps (which came with the Pyramids), in series with the 3.3uf caps on the crossover, essentially changing the value to 1.65uf, pretty close to where I needed to be. The coil is also close, within about 0.02mH. The HP crossovers are now -12dB at roughly 8000Khz and only cost me about $4 ea.
              I added a L-pad to the tweeter circuit as well, to lower its output without altering response.

              That"s about as much as I know about crossovers lol.

              I did not want to go into 3rd or 4th order due to phasing issues and sheer expense of parts. For my tastes, 2nd order crossovers still sound natural, not overly processed.

              Also, thx for the paper you emailed. I will begin reading it tonight. Too busy getting the house ready for guests lol.

              Comment


              • #37
                yeah. Flogging a dead horse. Someday, I'll treat myself to a decent calibrated mic. I won't use it enough to warrant a big expense. More for fun for me. I'll keep an eye open for specials, or maybe a good used mic. No worries.

                Comment


                • #38
                  Originally posted by Stash View Post
                  I like Pyramid. I know they are affiliated with Pyle. But Pyramid once had a good reputation. Not a fan of Pyle products.
                  You might have that backwards. Pyle was a reputable company that was purchased by Pyramid.

                  Comment


                  • #39
                    Thx DavidRoberts,

                    I wasn't sure who owned whom. I only know Pyramid and Pyle are related. I made the distinction because some consider Pyle and Pyramid to be one and the same.

                    My point was I bought a pair of Pyramid super tweeter compression horns for use in my speaker building project, and I am quite happy with them. They are not overly sharp, shrill or fatiguiing. With a 4 ohm damping resistor in series, the Pyramid tweeters offer a softer than expected treble. Yet, detail and nuance is not lost. Very pleased.

                    Not a fan of Pyle. Sorry. On more than one occasion, I've found Pyle products to be lacking in quality and/or their performance claims not met. And that goes for all Pyle products. Electronics, audio as well as their drivers. Overrated and underwhelming.

                    Comment


                    • #40
                      Chris & Wogg,
                      As a final note on this thread. I took a screen shot of your red-green comparitave graphs, Chris. On paper I will log the compensation required to bring the YPAO mic in line with the imm-6. Create a simple 30-40 band calibation file. I do not have any sophisticated software.

                      Then I will fine tune my calibration file using the calibrations which came with the imm-6.
                      Lol, I only have a printout of that file. I did not think about downloading it at the time.

                      I currently have a pair of DAYTON AUDIO tower speakers, T652. They have the cheap polycarbonate tweeters. I know the overall response is claimed to be 45Hz-20Khz.

                      Sometime into the next week, I plan to measure the T652 response. Use it as a rough reference speaker, to see if my measurements are close or way off.

                      The speakers are not heavy. I was going to place one speaker on a makeshift stand, so their cemters are about 4.5ft high (I have 9ft ceilings in liv-rm).
                      Angle them to project sound from the back of house towards the front, to reduce standing waves (my first floor is a more open concept).
                      set up my YPAO mic at 1m, on-axis with the tweeter.

                      I will try measuring the same 20Hz-20Khz correlated pink noise. And then I will take some sound burst measurements at maybe 10-15 bands. I found some on YOUTUBE. I've read sound bursts are more accurate as they are short enough as to not allow reflective waves to alter the measurement.

                      Is one way more accurate than the other?

                      Will I still need to do near-field (within 1cm of woofer) for frequencies below 100-200Hz?

                      And finally, is there a posted graph of the T652 with which I can compare my results, to see if I am even in the ballpark?

                      That's it! Thx again.

                      Comment


                      • #41
                        Can you do gating?
                        If you can't close the mic "window", I don't know how you'd keep reflections from your measurements, even with a "burst".
                        Here are the response files that I plotted on my graph:
                        IMM-6.FRD
                        YPAO.fRD

                        These are .FRD files (Hz, dB, phase).
                        I think a mic cal file would not use the 3rd column.
                        Attached Files

                        Comment


                        • #42
                          Chris,
                          thank you for the files. Downloaded 👍.

                          I've only read a little about gating. If center of speaker system is 1m from floor and mic is 1m from floor (forming right triangle), I will have a gate of 4ms ( 1.41m) or a low-end limitation of 242Hz?

                          I am assuming I will need REW or some such room acoustics program. The REW will compute (eliminate) reflective waves, smoothing the graph That's about all I understand so far.

                          For bass measurements I understand I will have to use near-field. I only recently began reading about the effects of baffle steps. I've always known of baffle interaction. But until recently, I've never looked at the actual math.

                          There is a slim chance I could also try outdoor measurements. My yard is fenced-in (6ft wood). Quiet neighborhood. I imagine I could place a speaker on a pedestal, 3ft high, in center of my backyard. I'd need about 50 ft of speaker wire. If the neighbor's dogs aren't out, ambient noise floor might be reasonable. Would such a measurement be more accurate?

                          I'd have to do it when no one was home. Or my family will think I've finally lost it lol. Only my nephews would understand. That is, the curiousity. Both are into audio. Kind of how I influenced them. One nephew is a self-taught guitarist. Played in a heavy metal band. He is knowledgeable about studio recording.

                          Comment


                          • #43
                            Your math doesn't seem wrong. I know that a 3.33ms gate roughly equates to 300Hz (they multiply to about 1000). And that equates to 4ft from all "surroundings". A 4ms gate is close to 250Hz. 33ms would get you down to 30. But if your mic "axis" (usually to the tweeter, and usually horizontal) is only 1m (about 3.3'), then you'd set your gate shorter, like 2.8ms, and your FR cutoff would be maybe 340Hz? With 9' ceilings I'd measure 4.5ft off the floor. Your gate could then be about 3.6ms - good to maybe 270Hz?

                            When I use ARTA (that one's free), you set it to "Impulse Response" and it does a fairly quick sweep (20-20k I assume) - takes maybe 5 sec.
                            Then it displays the Impulse Response. That does NOT look like an FR curve. It looks more like an EKG (w/only a single "blip"). There's a "flat line", then you see a big spike (when the initial sweep tones reach the mic from your speaker) and then the "jaggies" dampen out, less and less, until a secondary blip shows the first reflections arriving. That one's probably a magnitude less than the initial, but you use it to set your gate time so the reflections get gated out. Then it tells you your cutoff freq.

                            Pretty sure what the software does is it sends out a "tone" (a sweep, really - but it knows exactly when each freq. goes out). THAT "tone" travels at the SOL (speed of light basically) through the amp chain and drives the VoiceCoil. THEN the sound travels at the SOS (sound) across to your mic (then at the SOL back through to your computer software). With a gate of 3.00ms, the software only cares about the 1000Hz part of the sweep from a little before it's "arrival" to a little after. If (during a 5 sec sweep) the 1kHz freq. goes out at exactly 2.0 seconds, then the software only listens to (considers) the 1kHz tone up to 2.0 sec. + 3.0ms, or 2.003 seconds. After that time, it doesn't bother measuring any SPL from 1kHz any more, and moves on. Any SPL at 1kHz that arrives after 2.003" is considered "reflection".

                            (I just made all that up - someone set me straight if I'm wacky - PLEASE.)

                            Comment


                            • #44
                              I will look up ARTA. Yes, that sounds about right from what I've read. The reflected waves reach the mic a few ms following the original waves (the wave you want to register). The room acoustics app will eliminate the late arriving waves, thereby displaying only the original wave, as would be the case in a true anechoic chamber.

                              Comment


                              • #45
                                I will now close this thread. I tried looking up ARTA. Not available.
                                I tried searching ARTA on the internet. All I saw were apps recommending where to go for Asian food.
                                Tried searching REW in Microsoft. Found some assorted nonsense. Nothing to do with sound measurements.
                                I went on the internet to search REW. I can download it for free, but Microsoft warns it may not be safe. That's enough for me. I hate all this internet spyware crap.
                                Found some studio app on Microsoft for $7 that looked worthless.
                                Giving up on the idea for now.

                                I did write up a calibration file for the YPAO, based on the imm-6 graph and then fine tuned with the imm-6 calibration file. So it's ready to go, whenever. about 30 bands from 85Hz to 20Khz. I will at least be in the ballpark.

                                I may take my speakers outside into the backyard one of these days. Wouldn't be too hard. I have enough mic cord to sit on my patio and control the entire scenario. Such a readout should be fairly close, I imagine, without the use of room acoustic analyzers.

                                BTW, I was turned off by the REW website. "Analyser" is not spelled with a "S". Disgusting. Butchers the English language.

                                Now I want Asian food.

                                Comment

                                Working...
                                X