1. Register NOW and become part of this fantastic knowledge base forum! This message will go away once you have registered.

Looking to devise a repeatable microphone comparison test.

Discussion in 'Microphones' started by Littlefish, May 2, 2012.

  1. Littlefish

    Littlefish Active Member

    I am looking for ways to set up a repeatable microphone test for all of the microphones that my company offers, which will remove any performance variables.
    I was thinking about tracking as transparently as possible (Earthworks or Brauner KHE ) into Earthworks 1022 into Avid HDIO16 and then perhaps miking the output of this performance through a genelec 8050.
    This way I would collect a database of the relative differences between microphones accurately.
    Any suggestions of repeatable methodologies, a transparent capturing source microphone/pre combo, would be appreciated.
  2. audiokid

    audiokid Staff

    I'm about to do the same thing with a few high end converters. I'm going to mic a source coming from a sound system.
  3. RemyRAD

    RemyRAD Guest

    Honestly, I think you guys are really off base with that? The only real way to compare microphones is through the same preamp and on the same source. But then there is also the difference in spacing, height, distance from the microphones diaphragms. The best way to compare microphones is by listening. That's why we have ears. Actually it sounds like you have too much time on your hands?

    Have you considered blunt force trauma?
    Mx. Remy Ann David
  4. Boswell

    Boswell Moderator Distinguished Member

    I agree. You need to remove the recorded/reproduced element in the sound source from these tests. What you should do is get one of those Japanese violin-playing robots. Only something like that will produce repeatedly a sound source that has no artificial elements in it. These robots will be all over our concert halls in a year or two. Until the Unions get wind of what's happening, that is.
  5. audiokid

    audiokid Staff

    Come-on! maybe the miic test but how would you do a converter test then?
  6. RemyRAD

    RemyRAD Guest

    Just as Boswell and I described it. You need to stay with the same microphone preamp. You can also utilize a Yamaha Disclavier player piano to gain a truly repeatable and consistent performance as your sound source. I still have my Yamaha C 3 from which their Disclavier was originally based upon and later the C 7 with the Disclavier player feature. I certainly wouldn't use no speakers no how no way as I feel that would not be valid. You could also perhaps offer some music school students that opportunity however? You merely have to tell them to play the directions to Carnegie Hall. Or set up the microphones in a very tight cluster while utilizing a series of identical preamps into your nonidentical converters. I've always hated recording speakers. They sound just like, well, speakers. So if you're going to be a perfectionist about your equipment you should utilize a natural sound source. Perhaps some birds in a cage? Simply because you really want to know what these things are going to sound like with a real sound source and not an electronic representation of a sound source. Utilizing speakers is fine when you are only running a frequency response test. But that's not what this is you're talking about. So either do it right or you're fired! LOL

    How's your singing voice?
    Mx. Remy Ann David
  7. audiokid

    audiokid Staff

    Of course, I am thinking about comparing only some converters I have here > using the same mic and preamp for all of them. I've thought dangling keys might be a good source but a full range speaker mic'd in a room seemed pretty accurate too, no?
  8. RemyRAD

    RemyRAD Guest

    Actually I don't think so? It's still an electronic gizmo. So you'll be hearing more that electronic gizmo than you will the actual sound source and what it has to offer that you can glean from your converters. So I really think that's a flawed approach? Sure, I would imagine you would still hear a difference but will that difference be as valid as it should be? I'm really not trying to be a rabble-rouser here so please, don't think that's what I'm trying to do. I just know what I know.

    Know way!
    Mx. Remy Ann David
  9. mberry593

    mberry593 Active Member

    Microphone consistency (high school physics lab)- I have done this in the past to match pairs of microphones.

    Let's start with dynamic omnidirectional mics. Things get more complicated beyond that. This is a consistency test, NOT a quality test.

    Basically the idea is to compare a test microphone to a reference microphone, null out the differences and look at what remains on a spectrum analyzer.

    You need to construct a resistive mixer. This isn't expensive or difficult. The values are not critical and precision, low noise parts are not necessary. Connect pin 3 of each mic to 10k resistors. Connect pin 2 of each microphone to 20k /10 turn pots. Don't be tempted to use really small values here...we don't want one microphone to be loaded by the other. Tie the outputs together out of polarity, send them into a decent preamp and on into a spectrum analyzer. You will need a considerable amount of gain in that preamp as you will be losing some level in the mixer. If you don't have a spectrum analyzer, get the Blue Cat plug-in - it is excellent and has the charm of being free.

    Set up a noise generator into a reasonably high quality speaker. Positioning is a little tricky. You need to have both microphones fairly close together so they hear the same soundfield. They also need to be at least 10 feet (more is better) from the loudspeaker. Crank it up as loud as you reasonably can....remember we have lost some level in the resistive mixer. The idea is to adjust the pots to match any gain differences. What results after the outputs cancel out is the difference between the microphones. It is important to keep well away from AC power as this test setup will have poor hum rejection due to the imbalance caused by the variable resistor.

    You will need to do this test twice to completely describe the microphones.

    First, outdoors. The idea of being outdoors is to look only at on-axis response. There will be some reflection from the ground so keep the microphones as high as possible.

    Next, indoors. What we are looking at is how the microphones compare in their handling of off-axis sounds. So don't bother to put up any foam. We want the wall reflections. You don't even need a good room, remember we are NOT measuring quality, we are measuring differences from the reference mic. The results will be strange at the low frequencies so try not to get too worried about that.

    More -

    Obviously if you want to compare directional mics, you will need to do an additional test with closer spacing to evaluate proximity effect differences. Phantom power complicates the setup a little but can be done with care.

    Converters are harder. Noise may not exercise some important differences. That's beyond my pay grade.

    Good luck!
  10. audiokid

    audiokid Staff

    No worries Remy, I always appreciate direct honesty. No BS or candy coating is ever needed for me. I'm not doing this for fame and fortune. The music comes first and its only why I am here.
    I indeed understand exactly what you are all saying about the Mic tests. I gave up on mic comparisons like this long ago. Poping two mics up and comparing the one that works for that situation is how I do it. I don't even read mic comparison threads but I do read threads on micing and mics that are used. Just never trust any shootouts.

    Looks like you guys are going to get a good microphone discussion going for the OP so I won't go on too much more on the converter quiz but I would like to ask this rather than starting a new thread.
    I'm looking for very noticeable differences. If the results are subtle, I'd keep the more user friendly and call it a day.

    As an example, I used two difference converters in my mobile rig to record choirs the last few years. My tests were by no means scientific but the audio differences between the two were clear enough that I had no problem selling one over the other. One that everyone raves about sucked to me.
    If the sound quality is so close and you need to split hairs ( on anything really) I'm with the majority in saying, converters are over rated. So far though, I don't believe that to be true comparing low end to mid level to high end products in general. So for converters, I'm looking for a smoothness that is identifiable in the high end converters. Once you hear it, you know it kind of thing. If it takes a null test to tell the differences. I'm not that concerned.

    It sounds like it would be interesting to do a Pepsi challenge with a full range speaker and a live source to see if both would still identify certain sonic resemblances.

    Back to the OP.
  11. Littlefish

    Littlefish Active Member

    So I like some of the ideas presented! The first and most important property of this test is that it is EASILY REPEATABLE. It will be done whenever new microphones come in.
    Reducing the variables to me would mean using the same EXACT sound source every time on every test, with the same preamp, converter, and gainstaging into the daw, miked from the same distance of the source.
    Objectively we would fire white noise in an anechoic chamber and get different on and off access responses. This data is already available for the most part though, and doesn't give my customers the subjective experience of being able to hear the relative differences between microphones, which is the aim.
    I was thinking of capturing several different sources with the most transparent microphone, preamp, converters money can buy, and then remiking them using the same genelec 8060 monitor, the same distance from the speaker.
    Of course we will be capturing the properties of the speaker and preamp/converter combo twice. However a/b switching between these level matched recordings would highlight the relative differences between how the microphones capture their "source."

    The help I am looking for is in seeking a very repeatable way of capturing recordings which highlight the relative difference in microphones, holding as many variables as equal.

    Thanks for your help!
  12. mberry593

    mberry593 Active Member

    I'm thinking of those 'player' pianos that use punched paper. Are there any other robotic instruments?
  13. RemyRAD

    RemyRAD Guest

    Mike, yeah, Yamaha came out with their version of a MIDI controlled player grand piano in the early 1990s. They called it the Disclaviere, which was a take off from the New England Digital Synclaviere. The New England Digital thing was purely electronic. Where the Yamaha utilized MIDI, firing variable pressure solenoids of some type, in their C-3 & 7 actual grand pianos. It's an awesome instrument. Just like computers of today, it goes way beyond punched paper tape. It utilized a computer disk drive and had MIDI ports as well. In fact some of the nice hotels in and around Washington DC had put some of them in place as opposed to having a real guy playing them. So those things were putting hotel piano guys out of a job before we were at NBC. You can purchase pre-programmed performances I think even ones performed by Van Clyburn?

    You know when we were still in radio, the late Mark Biggs and I were very seriously thinking about investing in the New England Digital Synclaviere. We went as far as sitting down with the New England digital folks about it. At the time, they were still $250,000. We passed on it. It was an FM synthesizer coupled with one of the first disk drive-based sample record and playback devices. And that was an awesome beast. But like many innovative companies, they were put out of business by the proliferation of modern day computers and software by the mid-1990s. Sting was one of the first on board and New England digital sponsored a special demonstration with him at one of the legendary New York nightclubs where only folks who had attended the AES could obtain a free " aluminum " ticket to this demonstration. He came out with his current band at the time who all began to play. And as they played, one musician at a time would walk off stage while you still continued to hear their performance. It culminated to being just him onstage with the entire band being played back and not sounding any different from when the band members were all playing live. That post dated what Todd Rundgren did back in the 1970s when he came on stage with an MCI console and multitrack recorder and was booed off stage! LOL. Years later, I bumped into Todd Rundgren in the keyboard room at Chuck Levin's Music Center. He was sitting there checking out a keyboard as I was walking through the place. Nobody knew it was him. I had a nice talk with him and actually discussed that ill-fated concert of his because I thought that was a cool concept. He's an awesome dude and was way ahead of his time. I loved playing his stuff back in the mid-1970s when I was at WKTK, years before I came on board at WKYS. Around that same time, I made my way down to speak to Marty Enghauser, who told me to come back after I got my First Class FCC Ticket. Once I did that, it was Scott Standiford that hired me. When I had first stopped down, I was given a tour of the station and sat and talked to Joe Cipriano for a bit. I was really aghast at those old RCA boards. I thought for an NBC owned station that they would have had more modern equipment? Then I couldn't believe that they had pulled apart a custom API board but I found sitting in pieces in maintenance. It was shortly thereafter that I built up my own custom version utilizing whatever pieces of bread board I could find. It was Joe Hall that gave me all this blah blah about not putting transformers in it which I said we needed. But no! And I designed those soft on/off switches for the disc jockey's mics utilizing those Opto devices for smooth fade up & fade down without those obnoxious clicks. That was an awesome little broadcast board I built. And I used those API 325 cards for the outputs. Joe wanted something better and I found a couple of those Dean Jensen 990 Op-Amp modules lying around the shop so I used those. And we had terrible cross talk on the NEMO inputs because of the " no transformer " directive from him. Oh well, you can't fix stupid. So you could never leave one of those interlocking NEMO input buttons selected until you were ready to go on air with them as a result of that. Boy... those were the days. I think Darius got my custom API board? Maybe he still has it? I haven't talked to him in ages since I spoke to Skip, over at Fox. I thought I was good at cutting commercials until I heard Skip's work. I think he's still at Fox? And when you and I were still working together, you might remember that I had an awesome home production studio utilizing and all germanium transistor Phillips 12 x 4 console and an Studio Technologies, stainless steel EMT plate along with my couple of 1176's & 8 API original 550's supplementing the Phillips along with my 3 custom built Scully's & Neumann U-67's when you could still pick them up used for around $400 each, so I had 2. Still do. I parted with that Phillips 12 x 4 when I got my MCI JH-10-16 and also sold my API 550's to Paul Wolff, who I got to know back in 1978. He had just taken over API from Datatronix and he told me he really needed those. So I helped him out by selling in those for only $250 each. He later gave me the second prototype he had built up of the 3124 which I called serial number " 1/2 " along with his first production run which also included serial number 11 which I still have to this day. The second prototype actually had no serial number. And the Phantom supply was on little perph boards. I actually like these first run units better than their current versions. And it was because of me that he came out with the 2510 Op-Amp because in the first 3124 mixer version is output for the mixer section was with NE 5534's. So I scolded him and that he modified mine to have his new 2510's for the mixer outputs. The 2510's was just a 2520 without the pair of output drive transistors. And then I was a happy camper. In the late 1970s I had an eight pack of Neil Muncey, Saul Walker, MELCOR 1731 microphone preamp's. Those 1731 Op-Amps were the precursor to the API 2520's which were virtually identical. And Paul knew that because he was in touch with Saul Walker and was not only making 2520's, he owned API for 13 years. Remember when he was part owner in " No Evil " studios in DC? Before he ever owned API? They had a crappy "APSI" modified PA console. I remember he was so excited when he told me they had purchased a used custom console from Sunset Sound in LA that had API 312/325 & 550's in it. I had been the studio designer, chief engineer of Hallmark Films & Recordings in Baltimore after my days at FLITE THREE in Baltimore. Which was right after George Massenburg had left when they were still called Recordings Inc.. They had offered me the chief engineer job when I was just 16. But I told them that was much too responsibility for a 16 year old. So I got my mentor Tom Bray the position since after 23 years working for Hopkins he got laid off. Then the following year I got in at 17 as a production engineer and trial by fire music engineer LOL. It seems like only yesterday? Where has the time gone?

    Oh to be young again...
    Mx. Remy Ann David
  14. Littlefish

    Littlefish Active Member

    Wow. Off topic. Way off topic!
  15. Littlefish

    Littlefish Active Member

    So same mic pre is possible. so is converter. Disclavier isn't.
    All microphones capturing exact same live performance is not in any way feasible for several reasons.
    With these being what you have to work with, how would you best put together a test which captures the RELATIVE differences between microphones.
  16. RemyRAD

    RemyRAD Guest

    Why do you think that the Disclavier wouldn't be viable to evaluate how a microphone picks up an actual acoustic instrument is beyond me? It's a repeatable performance playback device of a true acoustic nature. Microphones can be placed in exactly the same positions and the playback re-performed precisely the same way each time. Unlike a single performance with a gaggle of microphones all in slightly different positions. And so it is possible. And it's practical. It's personal and it's acoustic. And if you are evaluating the true capability of a microphone consistently placed in the same position using the same preamp and converter, you can't get any closer. We're not talking about a laboratory test procedure but a real-life test evaluation.

    And how can anyone feel this is not on topic? You guys really slay me? Where precisely are you coming from? And how would you go about this yourselves? It doesn't even have to be performed in a studio. It can be performed in an actual performance Hall. Such as at a university, a church Tell me please because I am missing something here.

    Consistency is good
    Mx. Remy Ann David
  17. Littlefish

    Littlefish Active Member

    Why will a $12k Disklavier not work, seriously? For the purposes of my test, give me two very compelling reasons why you might suppose it would not work? Two that come to mind are size and price.
    Both not at all practical.
    I can come up with a litany of other reasons why one might consider this MORE mechanical, less natural, and less repeatable and accurate of a method then simply re miking.
    We are not talking about a hypothetical method here.
    I was looking for some help in developing something that might work for me personally, as a service to benefit others.
    As interesting as it might be none of the historically rich information you wrote about spoke to the original topic. I appreciate the fact that you would try to be helpful.
  18. RemyRAD

    RemyRAD Guest

    Well a $12,000 grand piano is both practical and professional in any studio. And with a computer-controlled acoustic piano, it's a lovely way to evaluate microphones. Now they don't have that... that's your problem. So you got something to save up for. You'll blow that much money on equipment and more for any professional studio. And you'll blow that much of a piano. I wish my Yamaha C-3 had the computer interface. But I bought it back in 1978 and I'm not ready to replace it just yet because I don't need to evaluate microphones. Besides, I think evaluating microphones is beginner BS. You just buy them and you put them in your collection and you keep collecting them. If you don't like one, you sell it. And that's what professionals do. Beginners need to evaluate microphones. So go to your local music school and book some time with the Yamaha with the computer audio interface to evaluate your microphones, sonny.

    I think you got too much sun in Miami?
    Mx. Remy Ann David
  19. Littlefish

    Littlefish Active Member

    Wow, you really don't get it.
  20. Boswell

    Boswell Moderator Distinguished Member

    I know both Remy and I have been a bit frivolous in our suggestions, but what were were trying to emphasise is repeatability. You can't begin to make meaningful measurements of differences between microphones unless everything else is exactly the same. This means not only using the same cables, pre-amps and converters, but also placing each mic under test in exactly the same place in the room and giving it a repeat of exactly the same sound source to listen to. You can't do it by putting two mics side-by-side, as they not in the same place. Believe me, a couple of inches matters, even in a mic test.

    The other point is that numerical measurements of microphones describe only one aspect of their differences. What matters more is how they sound, and no-one has devised a laboratory or studio test that reliably correlates with how listeners rate individual mics. That day may come, but it's not here yet.

    You would also run into problems if your sound source is a reproduced recording. Although precisely repeatable, a reproduced sound is already coloured in a way that could easily mask differences between microphones. Hearing something like a fairground organ may sound like a reproduction, but its punched card input is only an early example of programmed control of a sound source in the way that MIDI is the programming instruction language for a solenoid-driven piano. The sound itself is absolutely live and does not have the hidden artifacts of a recording and reproduction chain incorporated into it.

    Many years ago, when I first got 96KHz/24-bit recording capability, I did some tests under studio conditions using a mechanical hammer to operate the mallet on a crash cymbal. I was amazed how much of the high-frequency (>20KHz) and low-amplitude (<90dBFS) effects were removed by processing the recording down to the CD standard of 44.1KHz/16-bit. The human ear can perceive differences in transients at both high-frequencies and low amplitudes that are not shown up by steady-state audiometric frequency tests. It's performance in this sort of area that sorts out really great mics from the simply good ones.

Share This Page