@Hydrogen Alex - - Your grasping at straws, dude... Normally I just let trolls do their thing, but I am in the mood to bicker a bit... So I will play your game, using your rules...
You have a history of presenting your opinion as if it were fact. In this thread early on you advised
There have been other odd/strange/irrelevant suggestions you have made, we could go through them point by point if you wish.
It wasn't so long ago you dug your heels in and insisted you were right, only to realize you were wrong. I mean, thinking that the Sd of a 10" woofer is equal to the area of a circle 10" in diameter, that is a pretty fundamental error most newbies would not make. Anyone can make a mistake, but to insist you were 100% right, page after page... You used strong language to insist that (experienced people) were totally wrong, and then, pop, about face, sorry I was wrong, lets move on... no apology to anyone, no acknowledgement that you wasted everyone's time.
https://www.diyaudio.com/community/...-for-3-way-studio-monitor.409809/post-7616257
https://www.diyaudio.com/community/...-for-3-way-studio-monitor.409809/post-7616667
And you were concerned that I might be discouraging new folks...
Rather than clutter up this thread, you should start a new thread. Hydrogen Alex trashes hifijim - that would be a good title. Then you can troll me all you want, I will respond as best I can.
j.
You have a history of presenting your opinion as if it were fact. In this thread early on you advised
This is opinion, your opinion... it is not a fact. Furthermore, it goes against the advice of some very accomplished designers who have quite a few outstanding designs. Focusing on CSD plots to the exclusion of other data is not wise, and it is irresponsible to present this to someone new to DIY as if it were a fact, rather than your pet theory.
There have been other odd/strange/irrelevant suggestions you have made, we could go through them point by point if you wish.
It wasn't so long ago you dug your heels in and insisted you were right, only to realize you were wrong. I mean, thinking that the Sd of a 10" woofer is equal to the area of a circle 10" in diameter, that is a pretty fundamental error most newbies would not make. Anyone can make a mistake, but to insist you were 100% right, page after page... You used strong language to insist that (experienced people) were totally wrong, and then, pop, about face, sorry I was wrong, lets move on... no apology to anyone, no acknowledgement that you wasted everyone's time.
https://www.diyaudio.com/community/...-for-3-way-studio-monitor.409809/post-7616257
https://www.diyaudio.com/community/...-for-3-way-studio-monitor.409809/post-7616667
And you were concerned that I might be discouraging new folks...
Rather than clutter up this thread, you should start a new thread. Hydrogen Alex trashes hifijim - that would be a good title. Then you can troll me all you want, I will respond as best I can.
j.
Really, do please quote in full.
So which part of "Here are my top tips" states that my tips are anything other than my own top tips...?
As I clearly state, I believe time domain is far more important than frequency domain.... Happy to disagree with you and your type.
Mmmn, lets!
This must be from another post I assume... Oh my James, who is the troll now
Are you referring to a few months ago when I posted AJ designer sims and wasnt sure if the driver size should be based on driver diameter or cone area plus 50% of surround width.? Please confirm... Thanks James.
Oh how sweet James, please accept my apologies for offending you Oh my, I picked the wrong option, I thanked GM for pointing it out, corrected it and apologized for my mistake... How dare I, what outrageous behavior indeed..
Oh, you just told me I apologised when I realised I had made a mistake... Now you accuse me of not apologising.... Do pleae troll more accurately James, after all we are just having some banter old chap... Tally Ho, looking forward to your next volley
ttps://www.diyaudio.com/community/threads/which-10-woofer-for-3-way-studio-monitor.409809/post-7616257
Your links dont work when replying via quote...
Yes, and with well founded and detailed explanation to support them.
No, that would be a waste of my time and you are doing a grand job right here.
You can get the T34A for 200$ less, can be used as low more or less. Has a much wider dispersion then T34B, wider then domes 1" domes even.
Also for a reasonable price the wavecor TWA30xx are worth checking out.
Read Robs post, also be aware that most of the BE domes are coated also in the manufacturing process.
And the foil diaphragms tends to deform more then shatter into tiny pieces, so if that is your concern foil diaphragms should suit you fine.
You will also see that most of tve Materion diaphragms have polymer surrounds, not be, which furhter increases durability etc.
@jheoaustin,
Like you, I am planning to experiment a 4-way enclosure, now that I met success with my 3-Ways one :
This time, it will be a large sealed one :
View attachment 1317510
Speakers are :
- Beyma T-2030 1.25" Tweeter.
- SB Acoustics Satori MD60N 2.5" Dome Midrange.
- Beyma 8M60N 8" Cone Midrange.
- Beyma 12BR70 12" Woofer.
Still with a passive serial/parallel Xover, the planned transition frequencies are :
W ---> CM = 300Hz
CM ---> DM = 3000Hz
DM ---> T = 7kHz
As you choose it equally, I opt for a large 8" Cone Midrange, in order to have a good low-mid tone, notably including the main part of the voices in its range. By the way, that's what I did on my 3-Way enclosure, and it's a very positive experience, offering a very lively tone, somewhere like if it was a "Wide-Band speaker completed with the rest", if I can say so...
T
Wow, this is very interesting, and congratulations on your success of the 3-way design.
On your 4-way design, your crossover is more like for woofer - mid - tweeter - super tweeter 4-way. Satori dome can go pretty low, so 3kHz seems a bit too high. Could you explain me your design choice here?
Anyway, I am very interested in the outcome of both 3-way and 4-way ones. Please update us on your progress. Also, what is the tweeter in your 3-way design?
It's a filler driver not a midrange like the 3" ATC and Volt. That's why the response is the shape it is rather than flat and why it has a linear deflection of only half a millimetre (i.e. that of a tweeter not a midrange). If you want to use it as midrange rather than a filler driver it is only suitable for the upper midrange. Here is a link giving the gist of filler drivers which are possibly less relevant with FIR filters than analogue ones. Here is a link illustrating why blindly pressing buttons to get flat lines with FIR filters doesn't result in high quality crossovers which some in the thread seem to be advocating. The radiation pattern of the drivers and how they combine for both the direct and reflected sound influences perceived sound quality in a room. If the flexibility of FIR filters is wisely used it can result in a modest improvement on what can be achieved with analogue filters but today it still tends to require a bit of thought and checking to reliably achieve the best results.
Hi Andy,
Who is recommanding to blindly press buttons on FIR to achieve good results?
Of course it's a non sense, but there might be a misunderstanding: IF THE ACOUSTIC DESIGN OF LOUDSPEAKER have been thought to take care of potential issue ( and as you pointed directivity is one of those potential issue) THEN implementing a FIR complementary filter ( xover) is indeed as 'simple'* as targeting a flat line for drivers ( for something like +/- 1 octave more on both side of passband involved).
It's far more simple than implementing a passive filter for non electronic engineers. And it usually makes FIR users even more respectful for people able to implement passive xovers. At least this is the case for me and other members advocating for the simplicity of implementing FIR over passive xover. And it kind of 'fit' nice for multiways design, might not for others...
Grimm audio paper? Hmm... do you want to trigger another discussion about the ability of FIR filters to correct implementation of passive design, we know some of our fellow members are not convinced by this kind of use... Lol!
Worth a read of the Grim paper anyway, it's a great one. 👍
* simple? Well maybe if you are able to extract the 'right' informations from measurements. It's true for passive design too... it's simpler in that all further steps involved in a passive design are moot with FIR xovers once you have a flat behavior within the bandpass range...
Who is recommanding to blindly press buttons on FIR to achieve good results?
Of course it's a non sense, but there might be a misunderstanding: IF THE ACOUSTIC DESIGN OF LOUDSPEAKER have been thought to take care of potential issue ( and as you pointed directivity is one of those potential issue) THEN implementing a FIR complementary filter ( xover) is indeed as 'simple'* as targeting a flat line for drivers ( for something like +/- 1 octave more on both side of passband involved).
It's far more simple than implementing a passive filter for non electronic engineers. And it usually makes FIR users even more respectful for people able to implement passive xovers. At least this is the case for me and other members advocating for the simplicity of implementing FIR over passive xover. And it kind of 'fit' nice for multiways design, might not for others...
Grimm audio paper? Hmm... do you want to trigger another discussion about the ability of FIR filters to correct implementation of passive design, we know some of our fellow members are not convinced by this kind of use...
Lol!Worth a read of the Grim paper anyway, it's a great one. 👍
* simple? Well maybe if you are able to extract the 'right' informations from measurements. It's true for passive design too... it's simpler in that all further steps involved in a passive design are moot with FIR xovers once you have a flat behavior within the bandpass range...
Last edited:
Thanks for posting this Andy, even I find this paper interesting Also its written by two of the smartest audio engineers in modern day loudspeaker design. I will comment some more once I get home and time to re read it slowly... It takes me longer these days!
Also your own comments made me think and I have a question re FIR Vs IIR Vs passive...
In an active three or four way system personally I would go for the best DSP /Eq crossover I can afford, but for all the purist analog guys, do you know of any papers (my god I am worrying myself now!) or even better actual case studies / working systems which have all drivers driven actively but only use passive components in front of each driver?
Thanks
A.
Also its written by two of the smartest audio engineers in modern day loudspeaker design. I will comment some more once I get home and time to re read it slowly... It takes me longer these days!Also your own comments made me think and I have a question re FIR Vs IIR Vs passive...
In an active three or four way system personally I would go for the best DSP /Eq crossover I can afford, but for all the purist analog guys, do you know of any papers (my god I am worrying myself now!) or even better actual case studies / working systems which have all drivers driven actively but only use passive components in front of each driver?
Thanks
A.
^Isn't it the case for the many system implementing active filtering in analog?
I mean in pro world ( studio/P.A./live system) we didn't waited for dsp to implement active filtering.... such systems are in use since 70's...
An analog loudspeaker management system ( iow a xover) is just what you describe: passive components before amp ( line level filtering).
Could be done in many ways, fully passive, active,... whatever. In the end it's the levels involved which change : and doing it at line level solve a lot of potential issues ( reliability, accuracy, price/performance ratio) wrt typical passive xover found in loudspeakers. That said as always there is compromise at play...
I mean in pro world ( studio/P.A./live system) we didn't waited for dsp to implement active filtering.... such systems are in use since 70's...
An analog loudspeaker management system ( iow a xover) is just what you describe: passive components before amp ( line level filtering).
Could be done in many ways, fully passive, active,... whatever. In the end it's the levels involved which change : and doing it at line level solve a lot of potential issues ( reliability, accuracy, price/performance ratio) wrt typical passive xover found in loudspeakers. That said as always there is compromise at play...
Last edited:
Thanks Krivium, I was hoping to find some case studies using high end domestic drivers in WAF acceptable cabinets... PA gear is just too big and uses a very different set of design compromises.
My goal is to focus down on the precise advantages of DSP/Eq Vs passive on a driver to driver basis, not just complete system.
IE We all know (well those of us who have tried both) that a old school 3 or 4 way fully passive loudspeaker will be totally outperformed by a a fully active DSP/Eq version. BUT, that is a system advantage to the DSP... Ie upgrading from one stereo amp driving everything, to 3 or 4 stereo (or 6 or 8 mono blocks) driving everything.
I am looking for data / listening tests where a fully active multi amp system swaps between a single DSP/Eq and 6 or 8 passive dedicated driver crossovers. Ideally with matching budgets... ie a $1000 butget DSP Vs a total of $1,000 passive components.
I think this would a fascinating (as Spock would say) experiment.
My goal is to focus down on the precise advantages of DSP/Eq Vs passive on a driver to driver basis, not just complete system.
IE We all know (well those of us who have tried both) that a old school 3 or 4 way fully passive loudspeaker will be totally outperformed by a a fully active DSP/Eq version. BUT, that is a system advantage to the DSP... Ie upgrading from one stereo amp driving everything, to 3 or 4 stereo (or 6 or 8 mono blocks) driving everything.
I am looking for data / listening tests where a fully active multi amp system swaps between a single DSP/Eq and 6 or 8 passive dedicated driver crossovers. Ideally with matching budgets... ie a $1000 butget DSP Vs a total of $1,000 passive components.
I think this would a fascinating (as Spock would say
) experiment.No matter how hard your tried, you’d never be able match the response or phase of the two so not a very reliable study foundation…..you’d have to rely on subjective response….
Passive components do impart their own sound…..I think we all agree on this. Active filters do the same….so each would be adding its own salt and pepper to the goulash!
As a recording engineer Alex, I can only add my own personal observations…….and in that, an engineer ‘could’ record a source with any given microphone and then shape to preference with EQ and gain to achieve the desired tone from the source. BUT professional engineers don’t often do this…..instead, we try multiple microphones each with its own response profile until we find the one that works best with the source natively. Minimal or no EQ added when tracking…..no phase induced swings from EQ. Why?……because it’s audibly better and can easily be observed during the process. Granted, we have an advantage here in that it’s an isolated track in a fixed acoustic space as opposed to a multi track master in your environment…….but thats not the point of interest…..it’s that there IS an audible difference and improvement pre EQ.
These examples have been repeated for decades with thousands of mics in as many studios and on millions of tracks….many of which we use as reference for our subjective purposes. And amongst other things, PHASE matters and EQ shifts phase. When a mastering engineer produces a track, IF we intend to impart a point of reference to that track…..try to do as little harm as possible……careful with that EQ Eugene.
Last edited:
https://linea-research.co.uk/white-papers/
This is a goldmine for DSP geeks but the proof of the pudding...
I bought a second hand one on Ebay for £500 and it sounded fantastic, but it had a very hard life in clubs... Beer stains included! It sadly died after 6 months but I still wonder just how good would a new one be...?
https://linea-research.co.uk/wp-con...s/Datasheets and Brochures/ASC48 Brochure.pdf
100% agree, I am a fan of active DSP and really think most passive designs are a poor mans compromise... But there are a lot of music lovers who still believe all DSP is evil and only passive/analog can reproduce the "emotion/soul/heart" etc ( enter your own audiophile review BS!) experience... I am trying to find hard evidence to test this.^Isn't it the case for the many system implementing active filtering in analog?
I mean in pro world ( studio/P.A./live system) we didn't waited for dsp to implement active filtering.... such systems are in use since 70's...
An analog loudspeaker management system ( iow a xover) is just what you describe: passive components before amp ( line level filtering).
Could be done in many ways, fully passive, active,... whatever. In the end it's the levels involved which change : and doing it at line level solve a lot of potential issues ( reliability, accuracy, price/performance ratio) wrt typical passive xover found in loudspeakers. That said as always there is compromise at play...
Thanks.
A
main function of passive or active xo is to set the transferfunction, to get some desired acoustic frequency response. In ideal world it doesn't matter which technology you use as long as the transfer functions match well enough.
If you compare active system, where driver is connecter directly into the amp outputs, vs. passive where there is impedance metwork between amp and driver, it's the impedance that makes difference between the two, when transferfunctions make close to identical frequency responses. This leads to conclusion that if someone says analog is better this might be one reason.
Differences include: passive network might attenuate and lowpass amplifier noise. Also it could reduce some driver motor distortions entering acoustic domain. Should and could, because implementation matters For example at woofers cone resonance peak tamed with DSP is just compensation before amp, so amp noise and driver distortion is still "amplified" by the resonance, uncompensated. Cut it passively with series impedance notch instead, and also amp noise and motor distortion is "preconditioned" not to peak wotg the resonance.
On the otherhand active could be much more flexible, one could tailor the response to whatever, really make it ruler straight, and match Left and right easily, adjust in situ, and so on. Which is some reasons why active systems are better.
Answer to your question through is following. Logic from above is that one should use both, manipulate impedance and transfer function relatively independently, to get best performance.
Do not forget that in order to make the EQ nail the system response, you must, unfortunately, consult the folk you just downplayed with multiple posts or become one yourself, but, you could try without as well, because it's live and learn stuff.
If you compare active system, where driver is connecter directly into the amp outputs, vs. passive where there is impedance metwork between amp and driver, it's the impedance that makes difference between the two, when transferfunctions make close to identical frequency responses. This leads to conclusion that if someone says analog is better this might be one reason.
Differences include: passive network might attenuate and lowpass amplifier noise. Also it could reduce some driver motor distortions entering acoustic domain. Should and could, because implementation matters
For example at woofers cone resonance peak tamed with DSP is just compensation before amp, so amp noise and driver distortion is still "amplified" by the resonance, uncompensated. Cut it passively with series impedance notch instead, and also amp noise and motor distortion is "preconditioned" not to peak wotg the resonance.On the otherhand active could be much more flexible, one could tailor the response to whatever, really make it ruler straight, and match Left and right easily, adjust in situ, and so on. Which is some reasons why active systems are better.
Answer to your question through is following. Logic from above is that one should use both, manipulate impedance and transfer function relatively independently, to get best performance.
Do not forget that in order to make the EQ nail the system response, you must, unfortunately, consult the folk you just downplayed with multiple posts
or become one yourself, but, you could try without as well, because it's live and learn stuff.
Last edited:
Hi Alex,
Well you focused on P.A but i stated Studio before... and for the reason it's there i've got the most experience and it's indeed closer to typical domestic use.
On a driver basis? Well i don't think a system that way nowadays, the whole total of a loudspeaker is more than the sum of parts involved imho ( oh my god, Durkheim idea's brought to loudspeaker design! Sociology to loudspeaker... lol!).
In that i think this is what Jim was trying to express in previous posts, well that's what i got from reading your exchange. Who cares anyway...
I disagree multi amping is a 'privilege' for dsp, my previous post was about that: multiamping and active filtering are two different things both bring improvements and issues by themself and both are often linked because once you make a design choice as multi amp then it seems logical to go active filtering too. But it's not mandatory.
I think if you want to makes apple to apple comparison you should let cost aside: an analog processor will cost more than a digital LMU ( Loudspeaker Management Unit). Dsp is too vague imho, it cover anything done digitally and i fear a synth have nothing to do with what interest us.
You have a start hypothesis which is digital LMU is always better than anything else. I would not concur about this. I think it's case related.
I agree the digital LMU have gained on some areas with my threeways but it didn't solved acoustical design issues my loudspeaker have. Sound better to my ears too but i would not bet it would please everyone: it depends what you listen to, your own bias.
There is a bunch of objective technical improvements from going from passive to multiamp with active filtering and most of them are clearly described by Rod Elliott's article( biamping not magical but close to it part1/2) at ESP's site.
You could find more or less the same in 'Recording Studio Acoustic' ( Focal Press) where P.Newell list the pros/cons of such systems too. That said He and Hidley used Kinoshita's passive 2 way monitors...
I once was sure digital LMU was THE answer. Nowadays i think every technology have it's place, it depend of implementation ang goals of a design. There is no absolute.
I had the chance to play with a bunch of loudspeakers/systems and implemented the kind of comparison your after: it's in no way nothing more than anecdotal evidence but i tend to prefer rendering of active filtering/multi amp system.
There is a kind of 'clarity' and 'stability' of acoustic 'signature' with system like that. From reduction in IMD and the filters to not have to deal with high peak power transient help to maintain the same 'sound' independent from ouput level ( some monitors sounds goods only at some level not others, multiamp doesn't show this behaviour as much ime).
That said the 'best' loudspeakers i've heard were passive filtering... they had some 'coherency' to them i find less easy to implement in a dsped system. Maybe it's a bit of 'blured' line making them sound good to me, idk?
Well you focused on P.A but i stated Studio before... and for the reason it's there i've got the most experience and it's indeed closer to typical domestic use.
On a driver basis? Well i don't think a system that way nowadays, the whole total of a loudspeaker is more than the sum of parts involved imho ( oh my god, Durkheim idea's brought to loudspeaker design! Sociology to loudspeaker... lol!).
In that i think this is what Jim was trying to express in previous posts, well that's what i got from reading your exchange. Who cares anyway...
I disagree multi amping is a 'privilege' for dsp, my previous post was about that: multiamping and active filtering are two different things both bring improvements and issues by themself and both are often linked because once you make a design choice as multi amp then it seems logical to go active filtering too. But it's not mandatory.
I think if you want to makes apple to apple comparison you should let cost aside: an analog processor will cost more than a digital LMU ( Loudspeaker Management Unit). Dsp is too vague imho, it cover anything done digitally and i fear a synth have nothing to do with what interest us.
You have a start hypothesis which is digital LMU is always better than anything else. I would not concur about this. I think it's case related.
I agree the digital LMU have gained on some areas with my threeways but it didn't solved acoustical design issues my loudspeaker have. Sound better to my ears too but i would not bet it would please everyone: it depends what you listen to, your own bias.
There is a bunch of objective technical improvements from going from passive to multiamp with active filtering and most of them are clearly described by Rod Elliott's article( biamping not magical but close to it part1/2) at ESP's site.
You could find more or less the same in 'Recording Studio Acoustic' ( Focal Press) where P.Newell list the pros/cons of such systems too. That said He and Hidley used Kinoshita's passive 2 way monitors...
I once was sure digital LMU was THE answer. Nowadays i think every technology have it's place, it depend of implementation ang goals of a design. There is no absolute.
I had the chance to play with a bunch of loudspeakers/systems and implemented the kind of comparison your after: it's in no way nothing more than anecdotal evidence but i tend to prefer rendering of active filtering/multi amp system.
There is a kind of 'clarity' and 'stability' of acoustic 'signature' with system like that. From reduction in IMD and the filters to not have to deal with high peak power transient help to maintain the same 'sound' independent from ouput level ( some monitors sounds goods only at some level not others, multiamp doesn't show this behaviour as much ime).
That said the 'best' loudspeakers i've heard were passive filtering... they had some 'coherency' to them i find less easy to implement in a dsped system. Maybe it's a bit of 'blured' line making them sound good to me, idk?
Last edited:
This is the choice i've made too but mainly because i'm lazy and know how to set up a system fast with a digital LMU. But like you i use a serious unit were implementation was top notch ( Dolby Lake). So all audiofool BS i don't get... and as i've been guinea pig in double blind test i know i'm biased, that sighted test have close to zero value, that group influence your opinion, etc,etc,...
So i encourage you to do this but be aware of OUR own LIMITATIONS in doing so. And be honest to yourself too, not always easy to admit our own bias...
Totally agree, our own bias and belief will dominate in a sighted test, that is a major point and I am guilty of it big time!
But I have a secret weapon... My wife!
She is a singer and pianist and simply does not care about or have tech knowledge of gear and software etc... When I set up my own tests I always rely on her for sound quality checks.
Cheers
A.
But I have a secret weapon... My wife!
She is a singer and pianist and simply does not care about or have tech knowledge of gear and software etc... When I set up my own tests I always rely on her for sound quality checks.
Cheers
A.
If you want fun, you need a 4 way sistem. You can have a 3 way system but you need a very big sub box. Is all about low end, if you like nice chest punchy midbass, you need some efiicient drivers. A proper midbass driver have low mms and high fs like50hz 90+db spl1w/m. And then you need a sepparate sub.
If you want both you need a woofer with very low mms and very high vas(high cms for low fs) But you need a way bigger box.
Low end, box size, and efficiency. You can chose only 2, or little of each
Sorry for not reading everything
If you want both you need a woofer with very low mms and very high vas(high cms for low fs) But you need a way bigger box.
Low end, box size, and efficiency. You can chose only 2, or little of each
Sorry for not reading everything