Page 1 of 4 1234 LastLast
Results 1 to 10 of 31

Thread: What measurements matter to you?

  1. #1
    Devil's Advocate Adam_MSS's Avatar
    Join Date
    Aug 2010
    Posts
    1,578
    Real Name
    Adam

    What measurements matter to you?

    What measurements matter to you?

    I've been thinking a lot lately about how to help move this site forward and produce some benefit to the community. Goal is to eventually be a valuable resource for not only discussion but also unique information. To me, part of that is clearly gathering and sharing some objective data. I've got prices for Klippel gear, Audio Precision analyzers and other test gear but I want to make sure that the investment is made wisely and delivers some return for the membership.

    A Klippel would be nice, but I've at times wondered how much utility people were getting out of the extra data it provides versus lesser measurement systems. A Praxis or Soundeasy setup clearly doesn't have the same level of capability but it seems like people are better able to make use of the info.

    On the electronics side does an AP analyzer provide tons of extra actionable information that other testing methodologies don't? Or can we see most of what we want with an oscilloscope and some more basic computer based measurement gear?

    So the original question is aimed at helping me evaluate where the value lies. I've got all these ideas in my head and I'd like to prioritize based on what provides the most utility to members. (I've also got a CNC router idea brewing)

    Anyway, I'd appreciate your thoughts on what measurements matter to you. Toss out anything. This is brainstorming.
    Last edited by Adam_MSS; 09-21-2011 at 01:53 PM.
    You don't use science to show that you're right, you use science to become right. - R.Munroe

    The important thing in science is not so much to obtain new facts as to discover new ways of thinking about them. - W.L.Bragg



  2. #2
    Senior Member cvjoint's Avatar
    Join Date
    Mar 2011
    Location
    Northern California
    Posts
    801
    Real Name
    George
    I like the performance based method, Praxis or Sound Easy, stuff like that. HD at various power levels can give you insight into the motor goodies without actually having a Klippel and imo can be even more useful. A battery of tests with Sound Easy can make for very good information. The Klippel gives parameters of interest but are not essential for the end customer. In sum I would go with SE because it's the cheapest and most useful if used properly. Money would be spent on speakers instead or shipping.

    Tests I'd like to see done:

    HD at different power levels, and ballsy ones like 90db, 95db, 100db, 105db at the very least (1m). Maybe the tester should stop testing when THD is above x level.

    Power compression tests with a common car audio HP filter to keep the driver from reaching mechanical limits first.

    IMD tests if available.

    CSD tests keeping all axis units fixed over speakers and the FR flattened out if SE permits. Other time domain tests as recommended by Linkwitz.

    TS parameters using the box method.

    Off axis plots.

  3. #3
    Founding Member Subwoofery's Avatar
    Join Date
    Sep 2010
    Posts
    353
    Real Name
    Kelvin
    After you get your testing gear, I'd like to have a thread that compares 2 drivers and explain why 1 driver is better compared to the other. That would help newbies (like me) understand better and make a better use of the data

    Thanks,
    Kelvin

    PS: Love this forum

  4. #4
    Devil's Advocate Adam_MSS's Avatar
    Join Date
    Aug 2010
    Posts
    1,578
    Real Name
    Adam
    Good input. I've already got soundeasy and some other basic gear so getting moving in that direction is easy.

    I'm home shopping at the moment and a good room for testing is high on my list of requirements (my wife is on-board as she wants to contain my hobby to one area of the house).

    The test box method makes a lot of sense to me and is what I had planned to use for parameter measurement.

    Good points on some higher power levels to get better representation of actual use.


    Kelvin, good idea. I think we'll probably all debate what constitutes "better" but that's probably the best part of the discussion.
    You don't use science to show that you're right, you use science to become right. - R.Munroe

    The important thing in science is not so much to obtain new facts as to discover new ways of thinking about them. - W.L.Bragg



  5. #5
    Founding Member
    Join Date
    Aug 2010
    Posts
    180
    I would say polars but only for finished speaker designs and not so much individual drivers unless it is a unique diaphragm, has a special device to alter dispersion (more advanced then a typical phase plug).

    Higher level for sure, I want to see driver limitations and that is not going to show much @ 1W.

    I don't need to see what the MFG provides (or should provide) as a verification to see if they are falsifying data.
    Last edited by durwood; 09-21-2011 at 05:47 PM.

  6. #6
    Tester Extraordinaire ErinH's Avatar
    Join Date
    Aug 2010
    Posts
    1,247
    My $.02 as a tester and coming from my POV:
    • Klippel data -
      • Few people really understand (hell, I still don't fully and am just now getting to the point where it makes sense off the top) and fewer care. Adam, IMO, don't invest your money in to it. I know you don't like DIY but let me handle it from that perspective and link people. Of course, I know your background and I know your motives and issues with DIYma. That said, I personally would NEVER buy a Klippel for any site. I just don't think there's much worth and certainly I don't believe there's any payback. Not unless you can get one for $5k or less (thumbing an estimated worth to me).




    • SoundEasy/Praxis -
      • This is where, IMHO, the real meat is. If you twisted my arm and said I could have only one form of data and made me choose... I'd choose HD levels with SoundEasy. Why? Because MFG's are actually pretty responsible with their FR. Sure, some may try to fudge it, but truthfully, I trust most after comparing what I get to what they get. Plus, in a car, FR becomes such an animal that I wouldn't put all my eggs in to that basket. HD is where I want to see results. I want to see just how low I can cross a driver and at what levels it starts to turn sour. I chose 93-95dB for my personal tests because I was using others' drivers. In the future, when I start testing again (much sooner than I think most expect) I'll be doing some full bore testing but ONLY if the owners are OK with me potentially blowing their shit up. SE does HD testing differently than others and it will be pretty demanding. Don't let the numbers fool you. for drivers averaging in the 85dB @ 2.83v, 95dB starts tipping scales of power compression. This is typically prevalent in the 'fullrange' drivers. Once you climb above the mid-90's for dB scale, you're getting really, really loud. Also, you have to factor in the mic's ability. Most mics will clip and some will flat out die. I've read others' trials and tribulations as well as shared many e-mails with highly regarded testers on this matter alone.
      • These programs also give us FR which I like to see. It helps me determine where cone modes are and how usable a driver really is. Again, for car environment it's not crucial, but I think it's important to have. More than anything, it helps you determine why you're hearing what you hear. This is one of the two forms of distortion (linearity); the other being distortion (non-linear) which you'll obtain from the HD results.
      • IMD in addition to HD results would be nice. I never got around to posting this myself but SE has the feature and I did use it a couple times. You can set the tones you want, up to 3 IIRC, and play them all at the same time, singularly, or two at a time. However you'd like.




    • Impedance & T/S -
      • At the least, you should have Impedance results. And, most of us get our start here. I always head straight for the impedance spike to see the shape and magnitude of this because it's going to effect the amplifier in addition to what we hear. Not only do I want to see impedance for low end roll off, but high end attributes; the higher you go, the more inductance limits you.
      • T/S certainly has it's place because it gives us the most basic data set to use and within it being Fs and Qtc. Re as well, but I put more weight in sensitivity unless I'm worried about amp stability for high loads. We can use t/s to plug in to modeling programs and do some comparisons, too. I want to see the "other" stuff as well. IE: Sensitivity, Bl, Mms, and Cms (though, honestly, I don't really look at this much). I want to see what the motor is capable of, I want to see what kind of mass the cone has so I can predict if there'll be any weight to the impact and I really want to see what the measured sensitivity is so I can compare apples to apples in terms of power ratings.




    • Electronics -
      • RightMark is where I started but you really should use that with an attenuator and rock amps with it. I honestly haven't really given it much thought lately but you'll want to make sure you've got enough attenuation to drive the amp to a high enough load and not clip the soundcard input. You can make your own voltage divider or buy an XLR attenuator, depending on what your gear is. I'd go for the latter.
      • Testing headunits and processor with RightMark would work, too. It's what I used and Frank uses it as well. It has some limits, though, but for comparison purposes I think it's cool.
      • SoundEasy can be used to test HD. I started to do this one day but got busy. I need to look in to it again. This would give you the various orders of distortion.




    I second Kelvin; we need a discussion format as well. It would be great to talk with Vance Dickason and see if he would mind giving his $.02 now and again for data. I've done something similar with Patrick Turnmire and the Klippel. It doesn't have to be Dickason... maybe just someone with some solid knowledge in the realm.


    I hope everyone here realizes the magnitude of making a room really test worthy for speakers. It's quite an undertaking and really needs to be respected. I spent 3 months building and testing and prepping the room before I could get things acceptable. Adam, if you'd like any input from someone that has been down this road recently, you've got my number. I'm more than happy to discuss things with you. I would suggest looking in to this to help you control reflections for your gated measurements. I went all far field and got down to 200hz reflection free. Next time, though, I'll do NF/FF stitching. I have a list of "next time" things I'm more than happy to share.
    Not only the task of prepping the room but the pure time it consumes. Adam has a baby and travels. Same as I. It's not an easy task to juggle and getting results isn't very quick. With SE you have to switch cables around and sometimes it's easy to get caught up and have to re-run a test. Again, I just ask that as this progresses you all give credit where it's due, because Adam's about to get his ass kicked. I hope I can share the load, soon, though.

    That's all I have for now. I really am looking forward to moving. I've got plans drafted for the rec room which includes a few steel beams and a test stand. ;)
    Last edited by ErinH; 09-21-2011 at 07:32 PM.
    Your ears: The best tools you have... and they're free, too!

  7. #7
    Senior Member
    Join Date
    Sep 2010
    Posts
    430
    The best measurements would be how far from 0 I can keep my checking account.

  8. #8
    Senior Member
    Join Date
    Sep 2010
    Location
    Miami
    Posts
    194
    I agree with Erin on Rightmark's abilities. It is limited by extremely professional standards, but it is flexible enough to point out any audible flaws in components. And the version we use (ie 5.5) even lets you change THD and IMD frequency and amplitude parameters. All that is needed a fairly affordable and flexible soundcard that measure well at 24 bits and >44.1kHz (for amps testing). I see anything thing more precise than this for this purpose like a tailor taking your measurements down to the .1 mm.



    You just need the right gear and the correct setup basically.

  9. #9
    Devil's Advocate Adam_MSS's Avatar
    Join Date
    Aug 2010
    Posts
    1,578
    Real Name
    Adam
    Quote Originally Posted by ErinH View Post
    My $.02 as a tester and coming from my POV:
    • Klippel data -
      • Few people really understand (hell, I still don't fully and am just now getting to the point where it makes sense off the top) and fewer care. Adam, IMO, don't invest your money in to it. I know you don't like DIY but let me handle it from that perspective and link people. Of course, I know your background and I know your motives and issues with DIYma. That said, I personally would NEVER buy a Klippel for any site. I just don't think there's much worth and certainly I don't believe there's any payback. Not unless you can get one for $5k or less (thumbing an estimated worth to me).

    Totally agree. It's good information but it comes at a high price and is thus a poor value proposition unless I can acquire a Klippel at a substantial discount. It's never going to pay for itself in this sort of application so it's sunk cost.

    Quote Originally Posted by ErinH View Post
    • SoundEasy/Praxis -
      • This is where, IMHO, the real meat is. If you twisted my arm and said I could have only one form of data and made me choose... I'd choose HD levels with SoundEasy. Why? Because MFG's are actually pretty responsible with their FR. Sure, some may try to fudge it, but truthfully, I trust most after comparing what I get to what they get. Plus, in a car, FR becomes such an animal that I wouldn't put all my eggs in to that basket. HD is where I want to see results. I want to see just how low I can cross a driver and at what levels it starts to turn sour. I chose 93-95dB for my personal tests because I was using others' drivers. In the future, when I start testing again (much sooner than I think most expect) I'll be doing some full bore testing but ONLY if the owners are OK with me potentially blowing their shit up. SE does HD testing differently than others and it will be pretty demanding. Don't let the numbers fool you. for drivers averaging in the 85dB @ 2.83v, 95dB starts tipping scales of power compression. This is typically prevalent in the 'fullrange' drivers. Once you climb above the mid-90's for dB scale, you're getting really, really loud. Also, you have to factor in the mic's ability. Most mics will clip and some will flat out die. I've read others' trials and tribulations as well as shared many e-mails with highly regarded testers on this matter alone.
      • These programs also give us FR which I like to see. It helps me determine where cone modes are and how usable a driver really is. Again, for car environment it's not crucial, but I think it's important to have. More than anything, it helps you determine why you're hearing what you hear. This is one of the two forms of distortion (linearity); the other being distortion (non-linear) which you'll obtain from the HD results.
      • IMD in addition to HD results would be nice. I never got around to posting this myself but SE has the feature and I did use it a couple times. You can set the tones you want, up to 3 IIRC, and play them all at the same time, singularly, or two at a time. However you'd like.

    Soundeasy is a pretty powerful tool for the task. I agree it makes sense to use it to get some high value data that allows people to make reasonable comparisons between drivers. If you are going to be undertaking some similar testing there is probably some opportunity to coordinate our setups in order to allow some comparison from one data set to the other. We may not be able to get perfectly aligned but I bet we could get close if we used the same baffle setup, similar room treatments, similar baseline testing methodologies, etc.. It would be easy to verify by swapping a couple sets of drivers and comparing our respective results. It also would provide each of us with a credible resource for troubleshooting any issues we encounter.

    Quote Originally Posted by ErinH View Post
    • Impedance & T/S -
      • At the least, you should have Impedance results. And, most of us get our start here. I always head straight for the impedance spike to see the shape and magnitude of this because it's going to effect the amplifier in addition to what we hear. Not only do I want to see impedance for low end roll off, but high end attributes; the higher you go, the more inductance limits you.
      • T/S certainly has it's place because it gives us the most basic data set to use and within it being Fs and Qtc. Re as well, but I put more weight in sensitivity unless I'm worried about amp stability for high loads. We can use t/s to plug in to modeling programs and do some comparisons, too. I want to see the "other" stuff as well. IE: Sensitivity, Bl, Mms, and Cms (though, honestly, I don't really look at this much). I want to see what the motor is capable of, I want to see what kind of mass the cone has so I can predict if there'll be any weight to the impact and I really want to see what the measured sensitivity is so I can compare apples to apples in terms of power ratings.

    I think I'll compare the woofer tester route and the soundeasy route and see which seems to provide more accurate and repeatable results. The closed box method seems like a pretty feasible way to get a good handle on T/S parameters.


    Quote Originally Posted by ErinH View Post
    • Electronics -
      • RightMark is where I started but you really should use that with an attenuator and rock amps with it. I honestly haven't really given it much thought lately but you'll want to make sure you've got enough attenuation to drive the amp to a high enough load and not clip the soundcard input. You can make your own voltage divider or buy an XLR attenuator, depending on what your gear is. I'd go for the latter.
      • Testing headunits and processor with RightMark would work, too. It's what I used and Frank uses it as well. It has some limits, though, but for comparison purposes I think it's cool.
      • SoundEasy can be used to test HD. I started to do this one day but got busy. I need to look in to it again. This would give you the various orders of distortion.

    I need to get a better handle on this. I've heard all sorts of accolades and complaints about RMAA and I have to assume the differences in opinion are largely due to setup so it makes some sense to invest time in getting it right. We've definitely got enough of a brain-trust here to figure out a viable solution.


    Quote Originally Posted by ErinH View Post
    I second Kelvin; we need a discussion format as well. It would be great to talk with Vance Dickason and see if he would mind giving his $.02 now and again for data. I've done something similar with Patrick Turnmire and the Klippel. It doesn't have to be Dickason... maybe just someone with some solid knowledge in the realm.
    Agree that discussion is the key. Data without some discussion about what it means is useless to all but the most advanced users. Early on folks need to see the discussion back and forth about the good and bad qualities revealed by the testing and how they do/might manifest themselves in use. Getting some addition input and perspective from someone like Vance would be a neat addition.

    Quote Originally Posted by ErinH View Post
    I hope everyone here realizes the magnitude of making a room really test worthy for speakers. It's quite an undertaking and really needs to be respected. I spent 3 months building and testing and prepping the room before I could get things acceptable. Adam, if you'd like any input from someone that has been down this road recently, you've got my number. I'm more than happy to discuss things with you. I would suggest looking in to this to help you control reflections for your gated measurements. I went all far field and got down to 200hz reflection free. Next time, though, I'll do NF/FF stitching. I have a list of "next time" things I'm more than happy to share.
    I'm definitely going to pick your brain and hopefully learn from some of your tribulations. Can't have you going through all that effort for nothing :p.

    Quote Originally Posted by ErinH View Post
    Not only the task of prepping the room but the pure time it consumes. Adam has a baby and travels. Same as I. It's not an easy task to juggle and getting results isn't very quick. With SE you have to switch cables around and sometimes it's easy to get caught up and have to re-run a test. Again, I just ask that as this progresses you all give credit where it's due, because Adam's about to get his ass kicked. I hope I can share the load, soon, though.

    That's all I have for now. I really am looking forward to moving. I've got plans drafted for the rec room which includes a few steel beams and a test stand. ;)
    I definitely need to get a decent jig-box put together to turn cable swapping into "switch flicking". I can see how it could get frustrating quickly to have to go backwards.

    I really appreciate the insight and I'm sure I'll be relying on you pretty heavily to get this going once we get moved into the new house.





    Quote Originally Posted by durwood View Post
    I would say polars but only for finished speaker designs and not so much individual drivers unless it is a unique diaphragm, has a special device to alter dispersion (more advanced then a typical phase plug).

    Higher level for sure, I want to see driver limitations and that is not going to show much @ 1W.

    I don't need to see what the MFG provides (or should provide) as a verification to see if they are falsifying data.
    The higher level testing is going to need some structure. There's got to be some standardized methodology to figure out what level is appropriate for each driver in order to provide some valid comparisons from one model to another. I just want to make sure I don't misrepresent a driver by pushing it past it's limits when I don't do the same to others. We'll all need to discuss further to figure out how to get accurate and comparable results without condemning each driver to failure on the test stand.



    Quote Originally Posted by winslow View Post
    The best measurements would be how far from 0 I can keep my checking account.
    Can't make any promises there. :gr_grin:
    You don't use science to show that you're right, you use science to become right. - R.Munroe

    The important thing in science is not so much to obtain new facts as to discover new ways of thinking about them. - W.L.Bragg



  10. #10
    Devil's Advocate Adam_MSS's Avatar
    Join Date
    Aug 2010
    Posts
    1,578
    Real Name
    Adam
    Quote Originally Posted by t3sn4f2 View Post
    I agree with Erin on Rightmark's abilities. It is limited by extremely professional standards, but it is flexible enough to point out any audible flaws in components. And the version we use (ie 5.5) even lets you change THD and IMD frequency and amplitude parameters. All that is needed a fairly affordable and flexible soundcard that measure well at 24 bits and >44.1kHz (for amps testing). I see anything thing more precise than this for this purpose like a tailor taking your measurements down to the .1 mm.

    You just need the right gear and the correct setup basically.
    Good info. We'll probably need to start a thread dedicated to getting an RMAA based system set up properly as it sounds like it could be a discussion all it's own. That would be helpful for me in my setup as well as provide some insight to anyone else looking to do their own testing for comparison against mine.
    You don't use science to show that you're right, you use science to become right. - R.Munroe

    The important thing in science is not so much to obtain new facts as to discover new ways of thinking about them. - W.L.Bragg



Bookmarks

Posting Permissions

  • You may not post new threads
  • You may not post replies
  • You may not post attachments
  • You may not edit your posts
  •