pinkmouse said:
😀
I've been on such a wave.. just caught the crest and was dumped into the sandbar/coral below (..not fun).
I like the "prophetic woofer".. catchy (and in-line with most marketing claims being made today). 😉
So far no scientific or mathmatical explanation for the assumption of distortion. Nor any real world examples of the imagined distortion. This is turning into a 500 page legal document, size 3 font, on the proper design and implementation of a peanut butter & jelly sandwich.
- James
- James
AudialDelta said:
This is a general discussion forum. What did you expect.. an AES paper you could call your own and then turn into a commercial product?
No, it seems more your style to blanket a product type as inherently bad and not provide a viable explanation of your claim.
/Mod hat off.
Right, I've got the papers filed for the patent on the "Prophetic Woofer". Thanks for the name redunzelizer, I'll put a cheque in the post once the royalties start coming in. 😉
Ok, I think what Scott is worried about is a distortion mechanism, and one that should be relatively easy to check. We need to produce a noise burst, filter it with the same crossover type as used in the complete speaker, then compare the raw signal and the result from the driver to get a distortion magnitude. I'm not convinced personally that with a good driver it will be that great, but I think there may well be an AES paper in it for a person with the right test equipment.
Right, I've got the papers filed for the patent on the "Prophetic Woofer". Thanks for the name redunzelizer, I'll put a cheque in the post once the royalties start coming in. 😉
Ok, I think what Scott is worried about is a distortion mechanism, and one that should be relatively easy to check. We need to produce a noise burst, filter it with the same crossover type as used in the complete speaker, then compare the raw signal and the result from the driver to get a distortion magnitude. I'm not convinced personally that with a good driver it will be that great, but I think there may well be an AES paper in it for a person with the right test equipment.
AudialDelta said:
No at all. I do think that significant excrusion is a problem, particularly when pairing with other drivers that are mechanically dissimilar at higher spl's where excursion would be increased. (..and this is all in the context of "faster" which is only a part of this thread.)
Additionaly, if you don't think my explanation on this particular topic is viable - step up and disprove it rather than saying "everything thus far is cr@p".
pinkmouse said:
I think that it would also be relativly easy to measure. I think the more difficult thing would be determining relevance. In otherwords under what "constraints" would it be audible (and the converse). (..and I think this is the biggest problem we have in audio currently - i.e. what makes an audible difference and what doesn't.) While something may not be that "great" measurably - it very well could be audibly. (..and of course the opposite as well.)
Ok, so if we could get a range of magnitudes of distortion, at least that gives us more information to help decide on drivers. I'm sure you would agree, wether you can hear it or not, lower distortion from the primary distortion producer in the system has to be a good thing.
And if the differences are minimal, we have also learnt something, and we can start looking at other reasons why we might be able to hear differences.
And if the differences are minimal, we have also learnt something, and we can start looking at other reasons why we might be able to hear differences.
pinkmouse said:
Always. I'm not of the "Higara school of thought". The problem though is that removing some distortion may well increase other distortion (measurable or not). Still - you have to start from some point.
Mod, I am not attacking anyone, sorry if It appeared that I was. I am merely pointing out flaws in an assumption as your formula's and linked formula's have been unable to convince otherwise. Please delete this post if it violates the forum rules or if you feel your posts already show enough reason to discount the assumption of high excursion being inherently "slow".
Steve, exactly what type of proof would you like? How about a listening test? Oh yeah, I already indicated that I performed that test and a high Xmax driver passed with flying colors. Both posters who feel like doing math on the board have indicated there is no reason to rule out a high X-max driver for SQ duty. But why not do your own listening test before you eliminate a great product from your personal stereo or HT.
As Pink pointed out, and I knew it would be pointed out and therefore didn't bother, it is easy to do an impulse response test. Any test that is done needs to factor in the crossover, amp, box, etc. It's a test of the whole system. At that point, we can just use our ears for the test. And we are back at the top of the above paragraph.
I'm not sure why you would do this -
"particularly when pairing with other drivers that are
mechanically dissimilar at higher spl's where excursion
would be increased"
Are you speaking of pairing it with a midbass, midrange and tweeter? Combined with your desire to compare this sub in an unspecified enclosure to a horn loaded sub I really don't know what it is you are after in this thread. Different enclosure style comparisons should be in a seperate thread.
Now if this thread is looking for a true ultimate subwoofer that is like the search for the garden of eden. Or the magical box.
Would you like an impulse response measurement without a crossover? How about a waterfall graph to show rise and decay? How much lag time is needed to reach the audible limit - maybe do a google search for that answer? Test with a passive or active crossover? Will any of these convince you of anything without hearing it with music for yourself? If not, why not just listen to one or more of these subs?
- James
Steve, exactly what type of proof would you like? How about a listening test? Oh yeah, I already indicated that I performed that test and a high Xmax driver passed with flying colors. Both posters who feel like doing math on the board have indicated there is no reason to rule out a high X-max driver for SQ duty. But why not do your own listening test before you eliminate a great product from your personal stereo or HT.
As Pink pointed out, and I knew it would be pointed out and therefore didn't bother, it is easy to do an impulse response test. Any test that is done needs to factor in the crossover, amp, box, etc. It's a test of the whole system. At that point, we can just use our ears for the test. And we are back at the top of the above paragraph.
I'm not sure why you would do this -
"particularly when pairing with other drivers that are
mechanically dissimilar at higher spl's where excursion
would be increased"
Are you speaking of pairing it with a midbass, midrange and tweeter? Combined with your desire to compare this sub in an unspecified enclosure to a horn loaded sub I really don't know what it is you are after in this thread. Different enclosure style comparisons should be in a seperate thread.
Now if this thread is looking for a true ultimate subwoofer that is like the search for the garden of eden. Or the magical box.
Would you like an impulse response measurement without a crossover? How about a waterfall graph to show rise and decay? How much lag time is needed to reach the audible limit - maybe do a google search for that answer? Test with a passive or active crossover? Will any of these convince you of anything without hearing it with music for yourself? If not, why not just listen to one or more of these subs?
- James
Back to the "Ultimate..."
A few notes on my personal experiences with high x-max drivers.
While I personally will under most (if not any) circumstances prefer a driver with large x-max, I never got to the (dangerous) belief that more displacement capabilties per area will give any improvements per se. Even with such a driver I'd still stick to a design, where that driver is implemented "as if" it had low to average max excursion. The effects of "displaced overlayed transients" are there (and audible), but I would not be able to present any useful measurement setup to quantify this scientifically at the moment. As stated, at least with my types of designs I do consider pass band response above 2kHz of any (sub-)woofer as still very important. Often there is a "shallow" x-over slope with these too. So, Doppler-effects would be the most prominent occuring then, but should IMHO also be taken into account with high order filters.
Doing so (treating a high x-max as an average x-max) will reward you with a vastly extended dynamic headroom (in relation to "initial design specs") and would be the way I'd recommend to go in most cases. (Yet, often budget / space considerations will limit / prohibit that goal)
There is also something else people often forget about (woofer) distortion spectra. Let's say a woofer is crossed over at 200Hz (lowpass) with a slope of XXdB/Oxt. Now with a plain stoopid 1st Harmonic (aka "base frequency") playing at 150Hz at a given level, the 2nd order distortion will show at 300Hz *not* bandwith-limited at all. Even worse, 3rd order / 5th order distortion will show up at 450Hz / 750Hz respectively. Any of these 150Hz*N Harmonics stay "untamed", as if there was no x-over at all. No wonder, the valid x-over frequencies & XXdB/Oct slopes for harmonics *multiply* with according harmonic order. Did not yet see a lot of people thinking about that.
My next two millimeters of underhung thinking...
regards, redunzelizer
A few notes on my personal experiences with high x-max drivers.
While I personally will under most (if not any) circumstances prefer a driver with large x-max, I never got to the (dangerous) belief that more displacement capabilties per area will give any improvements per se. Even with such a driver I'd still stick to a design, where that driver is implemented "as if" it had low to average max excursion. The effects of "displaced overlayed transients" are there (and audible), but I would not be able to present any useful measurement setup to quantify this scientifically at the moment. As stated, at least with my types of designs I do consider pass band response above 2kHz of any (sub-)woofer as still very important. Often there is a "shallow" x-over slope with these too. So, Doppler-effects would be the most prominent occuring then, but should IMHO also be taken into account with high order filters.
Doing so (treating a high x-max as an average x-max) will reward you with a vastly extended dynamic headroom (in relation to "initial design specs") and would be the way I'd recommend to go in most cases. (Yet, often budget / space considerations will limit / prohibit that goal)
There is also something else people often forget about (woofer) distortion spectra. Let's say a woofer is crossed over at 200Hz (lowpass) with a slope of XXdB/Oxt. Now with a plain stoopid 1st Harmonic (aka "base frequency") playing at 150Hz at a given level, the 2nd order distortion will show at 300Hz *not* bandwith-limited at all. Even worse, 3rd order / 5th order distortion will show up at 450Hz / 750Hz respectively. Any of these 150Hz*N Harmonics stay "untamed", as if there was no x-over at all. No wonder, the valid x-over frequencies & XXdB/Oct slopes for harmonics *multiply* with according harmonic order. Did not yet see a lot of people thinking about that.
My next two millimeters of underhung thinking...
regards, redunzelizer
What if the reason we weren't thinking of that is that we don't plan to ever use a crossover point higher than 120 Hz, preferrably 100 Hz. And don't plan to use less than a 4th order low pass? Oops, did we just introduce 360 degree's of phase delay? And maybe we down fire the sub toward thick carpeting and padding to further diminish the 3rd and 5th order distortion. The stuff from about 250 Hz down has little attenuation from the said carpet.
I think the only time the posted scenario would surface is in a set-up useing small front or computer speakers. In which case we would likely have a lot more distortion if we let them attempt to go lower.
I think the only time the posted scenario would surface is in a set-up useing small front or computer speakers. In which case we would likely have a lot more distortion if we let them attempt to go lower.
AudialDelta said:
You are free to do so any time, and IMO there definitely is no general rule, there are only *your* very personal preferences. Just like everyone you'll have to suffer from the limitations of your own designs, and my ears would want to take quite (a) different pain(s) instead of the ones you proposed...
So, my experiences might only be usefull to people heading towards a similar direction, if ever... 😉
No, there are no rules in DIY. That's what allows it to be individualized.
But in your system, what is the purpose of the high crossover frequency? What is the rest of the speaker system? Having knowledge of the complete system may make your explations more reasonable. Maybe you are using them as the low frequency driver in full range speakers or similar.
But in your system, what is the purpose of the high crossover frequency? What is the rest of the speaker system? Having knowledge of the complete system may make your explations more reasonable. Maybe you are using them as the low frequency driver in full range speakers or similar.
AudialDelta said:
Well "high crossover frequency" is just a point of view. You could use the same 10" drivers as woofers (~200Hz) for keeping excursion lowest on the midbass drivers. You of course could have crossed at 85-120Hz, then having just the very same excursion issues on double the drivers. Why create a problem you can easily avoid?
BTW: We also successfully created a two-way sytem with the same 10" and a wave-guided dome tweeter. X-over then at ~2kHz. Of course that 10" was a "nice" and very "controllable" device.
So what's "high" then? 😀
AudialDelta said:
Must I repeat? I won't. 🙄
regards
redunzelizer said:
Well I admit when I was thinking of subwoofers I was thinking of something seperate from the woofer or midwoofer. But we live in different parts of the world and what we "normally" see in use is different. When I saw the thread topic I was thinking of stand alone subs. Systems such as what you described were popular as studio monitors in the 1980's. Professional studio monitors still use those designs. But the majority of the consumer market is working hard to make everything tiny. And that is why people such as I would prefer to build our own. My listening room is small enough to be filled with sound from 5 - 7 inch mids / midwoofers.
"High" then in my application is determined by where it is easy to pinpoint a subs location. I can easilly low pass the sub up to 100 Hz without it sounding as if it is somewhere other than where the front speakers set the stage. And my front speakers are large enough that I don't need to go any higher than that.
- James
AudialDelta said:
Its scott not steve. (but hey, steve is good name.)
Its not what type of proof I would like, but rather what type of proof do you have.
I don't have a problem with subjective comments. I recognize that you think a high x-max driver is good (or "passed with flying colors"). But there are 2 basic problems to your comments:
1. What do you think is "passed with flying colors"? Would someone else think that? Would most others think that? In other words a little more information would be not only helpful, but neccesary for anyone to respond meaningfully.
2. Have you made a comparison with the alternative? If not your subjective impression is completely one-sided.
Yes, I have heard numerous sub designs - those with higher x-max drivers and those that utilize greater surface area at the cost of volume and multiple drivers (and or larger drivers). I've also heard corner loaded bass horns (..though I'm not sure you can call them sub-woofers). My subjective impressions have already been provided - in a fair amount of detail.
Yes I was refering to using a sub with midbass, midrange, tweeter, etc.. Go back and read what I've written (#5 on page 10). Hopefully you'll understand the context then. The use of an enclosure (and what type) is entirely on point here because that will effect the mechanical operation of the driver (..as crossovers do in there own way).
As to graphical measurements: I given quite a bit of information that might provide a starting point for exploration through measurements. As pinkmouse and redunzelizer have mentioned concerning the lowpass character - there would need to be test of varying freq. and spl to account for a non-infinite lowpass character. Beyond that there are two significant points: 1. The test (or series of tests) should effectivly emulate a real signal. 2. The test will need to include a testing of subjects and their responses to differences (..with a placebo included).
- Status
- Not open for further replies.