1. Please take 30 seconds to register your free account to remove most ads, post topics, make friends, earn reward points at our store, and more!  
     
    TalkBass.com has been uniting the low end since 1998.  Join us! :)

Mix Driver Size Cabinet Testing

Discussion in 'Amps and Cabs [BG]' started by dhsierra1, Mar 25, 2013.


  1. As promised from this thread: http://www.talkbass.com/forum/f15/so-410-15-bad-but-210-15-good-969182/

    I'd like to kick off a thread on mixed cabs, what has to be one of the most controversial and heated debates in TB as far as I can tell.

    But in this case what I'm proposing is to help set up a consortium of interested parties who have the time, experience, and equipment to help get down to brass tacks: in what ways do cabinets with different sized drivers working full range and with out crossovers work and not work?

    What I'm seeing now is a growing consensus that mixing cabs doesn't suck 100% of the time. But what is missing is a deeper understanding of why in some instances it does and doesn't work.

    The idea is to present a collection of data collected in the real world by real bassists using real gear. A simple and reproducible test protocol needs to be hashed out and agreed upon. This is directly analogous in the engineering world to setting up test committees in such bodies as the ASTM or ISO. Thankfully we don't have to reinvent the wheel as there are already standardized tests in the speaker world that can be used as is or modified/simplified.

    With dirt cheap digital electronics and the world wide interwebz it can be done by a large number of people regardless of geographical location.

    Combining these two concepts: round robin testing using a standardized protocol and using a large number of participants with different rigs is analogous to pharmaceutical or medical device clinical trials. No two patients are quite the same even thought they may qualify to participate in a test treatment for some condition provided they meet some sort of inclusion criteria. There are statistical methods to collect, collate, and analyze data and present trends and results from experimental designs of this nature.

    What will inevitably happen is the initial data will be confusing and the test protocol modified as real world experience is gained. Like Clauswitz said, once the first shot is fired, the battle plan becomes obsolete. But at least it's a template and guide to get you off the ground.

    FWIW a little background about me so you know where I'm coming from in all this: my day job is in the medical research field and I'm an engineer with science degrees as well. I am not an acoustic engineer but I've been a musician pretty much all my life (I'm in my early 50s), a bassist for nearly 40 years. I've also designed and built speaker cabinets for PAs, bass, and guitars since the early 70s and have stayed abreast of the science and technology of musical equipment as best an amateur with a day job elsewhere can. I've set up and chaired an ASTM committee to establish test standards and test round robins in my field so I'm familiar with how to go about setting up test consortiums and manage them. I've also been involved in more clinical trials than I care to remember, including designing them.

    That said: I am not an acoustic engineer and I would really like some help from those who are: there are limitations in the medical test world analogies to the world of acoutics and MI. I am willing to take some time to help get this going but again, help is appreciated. My goal is simple: I'd like to get to the bottom of this and see the bass community help itself instead of bickering and arguing. Makes TB an unpleasant place to visit and participate in (I admit adding fuel to a few fire fights here so my doo doo stinks too). We can answer this question to some degree using the right tools and with reason, supporting the art of making music, which is what it's all about, yes? Yes.

    So........where to begin? Looking at the posts in the above-referenced thread starting around post #180 there's suggestions as to what kind of signal(s) we should be using, gear, conditions. So let the dialog continue here, please.

    One more thing: it's unimportant what your bias is in this debate, I have mine but I will -I have to if I'm following sound scientific practice- suspend disbelief and let the data do the talking. Hopefully we can all do this.
     
  2. aborgman

    aborgman Supporting Member

    Jun 12, 2007
    Ypsilanti, MI 48197
    Input: white noise

    Measurement: frequency response at 5 degree intervals from -90 degree to +90 degrees away from head on.
     
  3. will33

    will33

    May 22, 2006
    austin,tx
    A. How in the hell did I miss a thread that long???:confused::p


    B. I'm all for it.

    Phase response is usually the culprit in these deals. It's also difficult for most people (including myself sometimes) to interpret the data and "hear in my head" what it's going to sound like. The problem though shows itself in differing response around the listening area in relation to the rig that aren't explained by dispersion and/or room reflection alone. Recordings and/or response charts should show this. By comparing those to a matched rig, one should be able to pick out which anomalies are caused by the room and which are caused by the speakers themselves. Taking measurements/recordings where you hear hotspots/deadspots will show this.


    Anyway, getting ahead of myself here...allow me to catch up. Equipment/conditions will vary from person to person. As much detail as can practocally be provided is a good thing. A sort of consortium/baseline procedure is definitely needed, with some way to quantify all the different equipment, situations, etc. that will come in.

    I'll be setting myself up with a measurement rig here in the next couple months and can add data.
     
  4. seamonkey

    seamonkey

    Aug 6, 2004
    AES has papers on measuring loudspeakers outdoors. They are not free. Bill Fitzmaurice has described this several times in his past posts.
    i.e.:
    http://www.aes.org/e-lib/browse.cfm?elib=2318

    Once measured, you can store them in an EASE DB containing many cabinets:

    EASE Focus to allow people to model the mixing and matching of cabinets in acoustic spaces
    http://www.afmg.eu/index.php/products.html

    News from October 2012:
    "Major Enhancement: Conventional Loudspeakers

    As with every AFMG software release, the October 2012 update to EASE Focus 2 brings several interesting features that make work even easier and quicker. EASE Focus 2.2 now supports conventional “box-type” loudspeakers ..."
     
  5. good suggestions and directions so far. I think the idea of using the established test standards as a template are the way to go.

    I say this because a lot of people will not have the $, expertise, or gear to perform the gold standard tests.

    In clinical trials you have Phase I testing which means a limited number of patients and very limited parameters to measure and subsequently analyze. The goal: does this new drug/device even work and can we see a difference from the control.

    In our case: is this thing working and can we measure something/anything?

    Then based on that experience you roll into Phase II you have larger numbers of subjects with more test "arms" of different doses/conditions and more parameters you're measuring and analyzing.

    Phase III is the full monty where you're seeing if what would be the marketed dose/device/therapy works and how well. Uses large numbers of subjects with very sophisticated test protocols incorporating all sorts of combinations and procedures for data capture and analysis.

    So you want to book-end first with what you know and what you got, then build from there.

    But I think there will be those who have that kind gear and experience should go that route and share the results here.
     
  6. Mr. Foxen

    Mr. Foxen Commercial User

    Jul 24, 2009
    Bristol, UK
    Amp tinkerer at Ampstack
    Clipped from a BFM post elsewhere:

    This all needs to be done at various angles around the cab for dispersion stuff. Also its way simplified, but gives some idea what what you are dealing with.
     
  7. Agreed multiple angles horizontally is necessary. Not sure how many people are going to be able to string up a mic 4 meters (about 13 feet) above a cab laying on it's back though, esp those living in more urban environments ;)

    White noise makes sense as one of the inputs to be testing. Every 5 degrees to +/- 90 degrees off-axis? Again, for backyard/garage amateurs that might be a bit hard to do, at least for Phase I. But yes, I realize that is an industry test standard.

    How would one (the average interested bassist with minimal test gear) go about measuring the amount of wattage going into the speaker cab? Standardizing power in is going to be critical.

    But yes, this is the kind of simplicity with rigor we need.

    Downunderwonder in the thread that started this suggested some gear that's simple and (nearly) universal as test gear: a Shure SM57 microphone and Audacity. Not perfect but for Phase I seems to make sense since the SM57 is something probably everyone here has (or has access to) and Audacity is freeware.

    Audacity has a spectrogram display (frequency vs. "energy") that might do the trick.

    Any thoughts on this? I'm not terribly hip to digital recording software so please critique.
     
  8. will33

    will33

    May 22, 2006
    austin,tx
    Good a place to start as any to get everybody on the same page.

    The 57, like most mics, has it's own response curve, in it's case, a rolloff of the very low frequencies. Some of that type of stuff could be allowed for/corrected for.

    Sort of what I was getting at saying detailed descriptions of test equipment, procedures and environment in order to correctly interpret the data that is gathered.
     
  9. Mr. Foxen

    Mr. Foxen Commercial User

    Jul 24, 2009
    Bristol, UK
    Amp tinkerer at Ampstack
    Oh yeah, you need to bury the cab flush with the ground, and fill in round it, so its coming from a flat plane.
     

  10. So with the higher excursion drivers you can then plant your deeper rooted roses, keeps the gophers away from their delicate and delicious roots :D
     
  11. +1, precisely what I'm driving at ;)
     
  12. Measuring mics would be nice but relatively few of us have those.

    Open field would be nice but kind of exclusive.

    Huge numbers of measures is problematic too. White noise for hours on end won't go down well here.

    Can the audio engineers come up with a protocol that mere bass players can get onboard?

    I was thinking on the lines of ears as measuring devices and recording musical sounds, since the chief objection to charts is "I can't hear a chart" (even though experienced people can form a useful imaginary sound in their head from them )

    If others can go to town with white noise and 5 degree plots etc that would be neato.
     
  13. agreed, that's why I like your suggestion of using a universally available model, the tried and true Shire SM 57. A test mic it's not, but like I said, gotta start somewhere so us DIYers can start with something that'd be easily available.

    The standardized 5 degree testing and such should be done by those who have the chops and gear because it'd be nice to see how the Phase I test method compares to the gold standards. Unlike medical trials we can also do "Phase III" in parallel, I don't think any cabs will be in danger or sue us for malpractice :D
     
  14. acmebass

    acmebass Commercial User

    Mar 22, 2013
    Englewood, CO
    Owner/Designer, Acme Sound/Acme Low B
    The question was: "But in this case what I'm proposing is to help set up a consortium of interested parties who have the time, experience, and equipment to help get down to brass tacks: in what ways do cabinets with different sized drivers working full range and with out crossovers work and not work?"

    If you're looking for a generalized answer to that question, which will apply in all circumstances, forget about it. You won't find one.

    Here's your protocol: Get onstage with the two speakers, and play bass. If you like the sound, you like the sound.

    Obviously, I'm being a bit facetious, but if you're looking for a generalized methodology for predicting the performance when two unspecified cabs are used together, don't waste your time. It ain't gonna happen.

    Even with two specified cabs, and even if you have vast amounts of test data pertaining to both, the blizzard of variables involved would make it a severely daunting and time-consuming task. It would be much easier and more effective to just put the two together, run some curves, and hope the curves translate well to a live environment. In other words, even with two specific cabs, it would be less trouble and more rewarding to simply get them onstage and decide if you like the way it sounds.

    But to come up with some generalized way of predicting how any two cabs will work together? Even with test data, as I said, it would verge on impossible. But without extensive empirical data, forget it. You're wasting your time. Take the cabs to the gig, turn them up, and play bass. Does it make you smile? There's your answer.
     
  15. Scratch one potential helper and move on.
     
  16. What's the minimum size of room that could be used without making a total mess? Assume carpet on the floor, drywall lining. Are we pushing proverbial uphill with a stick trying to do anything indoors?
     
  17. I understand there can be a blizzard of data and variables, many of which are out of your control. But I do that for a living with something more complicated than speakers, with all due respect to you and your experience in this field (and btw I really like ACME cabs) ;)

    I'm not expecting we're going to get to the bottom of understanding and predicting how two (or more) cabinets, the same or different, interact. What I'm looking for is to begin the process of trying to understand rationally what some of the variables are and their response. You have to start simply and under controlled conditions, and then based on that experience and results expand from there.

    Perhaps it's Quixotic but you have to start somewhere and see where it goes. Maybe we crash the car and muddy the waters but I'd rather die trying than watch yet another thread here degenerate into the Hatfields vs the McCoys. There's some sharp people with a lot of experience here (including you) so I don't see why we can't at least attempt to move the ball forward in trying to understand what's going on. And make it practical for the working bassist.
     
  18. 1958Bassman

    1958Bassman

    Oct 20, 2007
    PINK noise, not white noise. There's a difference and pink noise is a standard for testing.
     
  19. If you understand the basic room response to white noise and what the resonance modes are then you have a chance. You'd probably have to "read the room" and then factor that into whatever you're reading with the cabs.

    Again, an audio engineer chiming in would be helpful here, though right now I'm furiously reviewing a few textbooks on the subject. For music purposes it seems the minimum sized room that's practical would be about 1500 cu ft ("Master Handbook of Acoustics, 4th Ed", by F. Alton Evert). In the US with 8 ft ceilings (pretty common) that's a 14 by 14 foot room.
     
  20. Mr. Foxen

    Mr. Foxen Commercial User

    Jul 24, 2009
    Bristol, UK
    Amp tinkerer at Ampstack
    Not in a room is the best way. Quite a lot of the world is outside. No mess from walls that way.
     

Share This Page