Hi all,
The Cabasse 'coaxial' parts name is tc21 for the 2way, tc22 for the 3way. This is an interesting design an to my knowledge the only one of it's kind: this is a dome surrounded by an annular membrane ( or 2 for the Tc22).
The speaker i had at home was a Goelete 500 i checked. The TC 22 is used in their latest flagship speaker "la sphere" which is a four way with a woofer located inside the spherical box for low end. Filtering is dsp active. Nice looking speaker and largely inspired by Ellipson 'la boule' speaker speaker series from 70's.
I re read the thread quickly last night and had an other thoughts about some things. I hope you won t mind if i share them:
When Wae specified target for his plan he talked about price of high end monitors and was schocked by them. Well in some case (not all) i can be explained: first the market for this kind of product was small (and it is getting smaller and smaller) so brands do not sell a lot of speaker and r&d cost a lot so... price goes up quickly.
Secondly when you are looking to improve on design there is a non linear curve when it come to price, lets take an analogy with bicycle: when you want to increase performance with a bicycle you'll chase weight and rigidity. Halving the wheight and doubling rigidity is not to pricey from a standard product: you use aluminium in place of steel and geometry tricks and here you are. Price won t go up this much and you met your target.
But once you have halved the weight and played with geometry how do you upgrade? Well you ll need to use exotic material and welding technique, composite material and the like. And here to gain 100gr you ll more than double the price! Then once it is done if you want to gain 10gr same law apply and you double price again, etc,etc,...
This is one of the reason some of this monitor cost so much: they rarely use standard version of drivers and the details are pushed far away versus hifi speakers. In particular some cabinet design can be pricey to build and when the loudspeaker is passive the cost of components for filter skyrocket too as we are talking of kw hungry speakers ( kinoshita for eg).And i won t talk about price involved for drivers developments this seems crazy when you ask for a largely modded one in small series.
The last reason in my view is sneaky. If you have spent time in large commercial studios and know the owner you ll quickly realise those RICH men does almost always behave as 6yo kid: competitor studio invest in a new desk... they NEED the new one too and if possible with more options becausd they DESERVE something 'exclusive'...
So price is not really something that matter as long as quality is here. This the market Wae plan to attack. So there may be an issue in the high segment of the market for a commercial product aimed at pro control room with a budget minded product. Snob attitude.
In the lower segment you ll face an other issue: project or home studio doesn t have the space for big mains most of the time especially for inwall as cost to achieve this frighten most peoples (right or wrong).
Here budget is an issue so Wae plans seems good to me but you ll have to focus on a medium sized monitor for them as this is what they ll needs and this is probably the one he ll sell the most. You may face another issue with budget minded studios and choice of components if you go active: forget behringer for example... Why? Because musicians await something different than what they use at home when they pay for a service! This may seems crazy because this is the skills of operators which is important but that is a fact! And i know because i faced this multiple time: my first digital desk was acquired from a reputed studio in Paris owned by one of the best enginneer of the place which couldn t stand musicians explaining him how to achieve things... So he discarded the 02R for an Euphonix digital... and when i picked up the desk he explained me that as long as stayed digital chain there was ZERO difference in sound between the two desks. I was shocked but he was right!
I ve seen potential customers running away from me for the reason i ve had some Behringer units in my rack too... one spectre analyzer and a psychoacoustic effect which is as effective as the 'prestigious' one... So if you plan to sell some i would suggest BSS or DBX as the minimum allowable brands.
This shouldn t stop you however but if you invest money just be sure it won t be lost and that you ll have some money return on your investissment.
The Cabasse 'coaxial' parts name is tc21 for the 2way, tc22 for the 3way. This is an interesting design an to my knowledge the only one of it's kind: this is a dome surrounded by an annular membrane ( or 2 for the Tc22).
The speaker i had at home was a Goelete 500 i checked. The TC 22 is used in their latest flagship speaker "la sphere" which is a four way with a woofer located inside the spherical box for low end. Filtering is dsp active. Nice looking speaker and largely inspired by Ellipson 'la boule' speaker speaker series from 70's.
I re read the thread quickly last night and had an other thoughts about some things. I hope you won t mind if i share them:
When Wae specified target for his plan he talked about price of high end monitors and was schocked by them. Well in some case (not all) i can be explained: first the market for this kind of product was small (and it is getting smaller and smaller) so brands do not sell a lot of speaker and r&d cost a lot so... price goes up quickly.
Secondly when you are looking to improve on design there is a non linear curve when it come to price, lets take an analogy with bicycle: when you want to increase performance with a bicycle you'll chase weight and rigidity. Halving the wheight and doubling rigidity is not to pricey from a standard product: you use aluminium in place of steel and geometry tricks and here you are. Price won t go up this much and you met your target.
But once you have halved the weight and played with geometry how do you upgrade? Well you ll need to use exotic material and welding technique, composite material and the like. And here to gain 100gr you ll more than double the price! Then once it is done if you want to gain 10gr same law apply and you double price again, etc,etc,...
This is one of the reason some of this monitor cost so much: they rarely use standard version of drivers and the details are pushed far away versus hifi speakers. In particular some cabinet design can be pricey to build and when the loudspeaker is passive the cost of components for filter skyrocket too as we are talking of kw hungry speakers ( kinoshita for eg).And i won t talk about price involved for drivers developments this seems crazy when you ask for a largely modded one in small series.
The last reason in my view is sneaky. If you have spent time in large commercial studios and know the owner you ll quickly realise those RICH men does almost always behave as 6yo kid: competitor studio invest in a new desk... they NEED the new one too and if possible with more options becausd they DESERVE something 'exclusive'...
So price is not really something that matter as long as quality is here. This the market Wae plan to attack. So there may be an issue in the high segment of the market for a commercial product aimed at pro control room with a budget minded product. Snob attitude.
In the lower segment you ll face an other issue: project or home studio doesn t have the space for big mains most of the time especially for inwall as cost to achieve this frighten most peoples (right or wrong).
Here budget is an issue so Wae plans seems good to me but you ll have to focus on a medium sized monitor for them as this is what they ll needs and this is probably the one he ll sell the most. You may face another issue with budget minded studios and choice of components if you go active: forget behringer for example... Why? Because musicians await something different than what they use at home when they pay for a service! This may seems crazy because this is the skills of operators which is important but that is a fact! And i know because i faced this multiple time: my first digital desk was acquired from a reputed studio in Paris owned by one of the best enginneer of the place which couldn t stand musicians explaining him how to achieve things... So he discarded the 02R for an Euphonix digital... and when i picked up the desk he explained me that as long as stayed digital chain there was ZERO difference in sound between the two desks. I was shocked but he was right!
I ve seen potential customers running away from me for the reason i ve had some Behringer units in my rack too... one spectre analyzer and a psychoacoustic effect which is as effective as the 'prestigious' one... So if you plan to sell some i would suggest BSS or DBX as the minimum allowable brands.
This shouldn t stop you however but if you invest money just be sure it won t be lost and that you ll have some money return on your investissment.
Last edited:
Cabasse / budget
Did a quick peek, look worth diving into. See attached image as found example.
Thanks for the views you have on budget and "behaviour". You are spot on.
My R&D is quit simple;
- People: I have 3 employees (ME, Myself and I), no kidding it's just one. My time is for free, it's my spare time from my consultancy business in ICT. I'm not really making money on this, it's a hobby that pays for it self. MAYBE that will change, but for now WAE is a small company. Growing gives all kind of <PEEP> and I'm not into that right now.
- KNowledge: I read and read and read.
- Skills: Woodworking is OK, Electronics so-so, Metalworking- outsourced
- Material: Before I start to build, I design on paper (No Autocad or other expensive Software) but just the simple MS Office Excel and PowerPoint (To make drawings, sometimes a pain in the ***). Once I feel confident, I start with the first moc-up with cheap wood (so I now it all fits should work as planned). Next I make the real model and start measuring (REW on MacBook Pro). Once fixed all issues, I have the production versions made by a local company that uses dis-abled people (low cost, but deliver PRIME Quality). My Portable PA top-cabinet (4" x 5" x 17,5" holding four 4" drivers) took me 5 versions, of which 2 where product improvement. Estimated R&D cost for this product around 500 Euro's. Sold 26 sets, made a small profit
- The world as a source of knowledge: I am always curious and willing to listen to idea's and input from others. That's what this website is also intended for. THAT is a valuable source which comes for free, where I think it's fair for me to publish what I'v learned and pass it on. Including drawings, schema's, settings, products used, the lot.
So I think I have it under control so far. That's by the way, why I am looking at affordable components for a start. Low cost R&D to proof the concept, low cost end product for affordability.
Bye the way, I've been working in a company that was mainly R&D, so i've seen the danger of being sucked dry on theory's. Also, R&D tends to fly "all over the place" with again new improvements & other products & new theories, etc. If we can explain it and measure it, doesn't mean we will be able to here it.
So I come up with a concept, and unless it is absolute rubbish, stick with it and sole the issues.
Just read the other day on a company that has invented another way of mounting speakers to the cabinet. They change the material of the front of the cabinet. Ooh really.. ?? Yep, also found a fancy name for it. That must have cost some serious pennies I guess. I don't go that route for a foreseeable period.
So much for R&D, thanks for pointing out.
Active:
I've set my eyes on the Hypex try-amp plate amplifier. It comes with a DSP and software, goes around 350 euro's. THEY have done all R&D and have a good name in the market.
Speeking of - Market:
Not shure yet, it's a strange world. Since I intend to do a series (from 6" to 12" with a coaxial) this one is the first of the set, because I already have some parts of the 12" laying around (limited investment). Would be nice to make a 2x 6" with a 4" coaxial (does exist!) as a near field....
I started entering the broadcasting studio market just recently. I will do the live performance sound engineering for a local radio, and use MY monitors to work with. That way, they will hear another affordable but very good product. Hope to sell them my first studio monitor.
I guess it will not be viable to expect a large commercial studio to ditch their PMC BB3 set and by a set of mine, but wonders do occur now and than (just dreaming and time flies).
OK, I'm off, got to start soldering 16 signal cables this afternoon.
kind regards.
Did a quick peek, look worth diving into. See attached image as found example.
Thanks for the views you have on budget and "behaviour". You are spot on.
My R&D is quit simple;
- People: I have 3 employees (ME, Myself and I), no kidding it's just one. My time is for free, it's my spare time from my consultancy business in ICT. I'm not really making money on this, it's a hobby that pays for it self. MAYBE that will change, but for now WAE is a small company. Growing gives all kind of <PEEP> and I'm not into that right now.
- KNowledge: I read and read and read.
- Skills: Woodworking is OK, Electronics so-so, Metalworking- outsourced
- Material: Before I start to build, I design on paper (No Autocad or other expensive Software) but just the simple MS Office Excel and PowerPoint (To make drawings, sometimes a pain in the ***). Once I feel confident, I start with the first moc-up with cheap wood (so I now it all fits should work as planned). Next I make the real model and start measuring (REW on MacBook Pro). Once fixed all issues, I have the production versions made by a local company that uses dis-abled people (low cost, but deliver PRIME Quality). My Portable PA top-cabinet (4" x 5" x 17,5" holding four 4" drivers) took me 5 versions, of which 2 where product improvement. Estimated R&D cost for this product around 500 Euro's. Sold 26 sets, made a small profit
- The world as a source of knowledge: I am always curious and willing to listen to idea's and input from others. That's what this website is also intended for. THAT is a valuable source which comes for free, where I think it's fair for me to publish what I'v learned and pass it on. Including drawings, schema's, settings, products used, the lot.
So I think I have it under control so far. That's by the way, why I am looking at affordable components for a start. Low cost R&D to proof the concept, low cost end product for affordability.
Bye the way, I've been working in a company that was mainly R&D, so i've seen the danger of being sucked dry on theory's. Also, R&D tends to fly "all over the place" with again new improvements & other products & new theories, etc. If we can explain it and measure it, doesn't mean we will be able to here it.
So I come up with a concept, and unless it is absolute rubbish, stick with it and sole the issues.
Just read the other day on a company that has invented another way of mounting speakers to the cabinet. They change the material of the front of the cabinet. Ooh really.. ?? Yep, also found a fancy name for it. That must have cost some serious pennies I guess. I don't go that route for a foreseeable period. So much for R&D, thanks for pointing out.
Active:
I've set my eyes on the Hypex try-amp plate amplifier. It comes with a DSP and software, goes around 350 euro's. THEY have done all R&D and have a good name in the market.
Speeking of - Market:
Not shure yet, it's a strange world. Since I intend to do a series (from 6" to 12" with a coaxial) this one is the first of the set, because I already have some parts of the 12" laying around (limited investment). Would be nice to make a 2x 6" with a 4" coaxial (does exist!) as a near field....
I started entering the broadcasting studio market just recently. I will do the live performance sound engineering for a local radio, and use MY monitors to work with. That way, they will hear another affordable but very good product. Hope to sell them my first studio monitor.
I guess it will not be viable to expect a large commercial studio to ditch their PMC BB3 set and by a set of mine, but wonders do occur now and than (just dreaming and time flies).
OK, I'm off, got to start soldering 16 signal cables this afternoon.
kind regards.
Attachments
If Hypex DSP has FIR filters you could examine the tricks Fulcrum does with FIR filters on their speakers, at some point. I found some papers from the designer Gunness. EAW NT series was perhaps first that had the processign, then he started Fulcrum Acoustic and later on helped Presonus to implement similar DSP processing on their coaxial speakers. I think it was the EAW NT paper that said they use their own software to measure timedomain related issues from a speaker frequency response and fix those separate to non timedomain frequency response issues. Horn honk and splashing and other acoustic impedance change related issues was mentioned, also for the woofers.
They design the speakers so that DSP is included from the get go to reduce compromises on a design. Some issues which can be fixed with DSP are left for it so physical design can be optimized for other things such as directivity.
Very much reading needed to compete them but, since you like reading I thought to hint.
They design the speakers so that DSP is included from the get go to reduce compromises on a design. Some issues which can be fixed with DSP are left for it so physical design can be optimized for other things such as directivity.
Very much reading needed to compete them
but, since you like reading I thought to hint.
Spot on Tmuikku!
They licenced dsp for the Spectre range to Presonus.
I didn t know of this Cabasse driver you linked Wae.
The one i was refering to are this one: Cabasse tc21 - Google Search
And this one: Cabasse tc22 - Google Search
Cabasse tc22 - Google Search
They may be very interesting to use but it won t happen as Cabasse don t sell them to end user. ( i know i asked to them!)
And here is la sphere: cabasse la sphere - Google Search
cabasse la sphere - Google Search
Thanks tmuikku,
Never heard of Fulcrum, but shure fits in this approach. I also did find that they use B&C, but most likely have their own version with a horn added for the HF. Have to check whether Hypex supports FIR filters. I did see a FIR config file for ICEpower amps.
SO, also did have a quick look at B&C (I use some of their components) and found a nice LF and Coaxial 8" versions form B&C.
MOSTlikely, I will start with the 8" PM-DMT, 2x LF and 1 Coax for 270 euro's. The work very well in a transmission line, where the Coaxial 8" need 3 liters to sit in the Transmissionline. ALSO, the required LF offset for the B&C 8" leaves enough room for the COAX to sit between the LF and the terminus of the TL
I'm happy.
kind regards.
They licenced dsp for the Spectre range to Presonus.
I didn t know of this Cabasse driver you linked Wae.
The one i was refering to are this one: Cabasse tc21 - Google Search
And this one: Cabasse tc22 - Google Search
Cabasse tc22 - Google Search
They may be very interesting to use but it won t happen as Cabasse don t sell them to end user. ( i know i asked to them!)
And here is la sphere: cabasse la sphere - Google Search
cabasse la sphere - Google Search
Thanks tmuikku,
Never heard of Fulcrum, but shure fits in this approach. I also did find that they use B&C, but most likely have their own version with a horn added for the HF. Have to check whether Hypex supports FIR filters. I did see a FIR config file for ICEpower amps.
SO, also did have a quick look at B&C (I use some of their components) and found a nice LF and Coaxial 8" versions form B&C.
MOSTlikely, I will start with the 8" PM-DMT, 2x LF and 1 Coax for 270 euro's. The work very well in a transmission line, where the Coaxial 8" need 3 liters to sit in the Transmissionline. ALSO, the required LF offset for the B&C 8" leaves enough room for the COAX to sit between the LF and the terminus of the TL
I'm happy.
kind regards.
I Wae,
Your plan seems like a good one to me ( in fact it is not very disimilar to what i plan for my next loudspeaker which is itself influenced by a commercial product availlable at the moment! ).
About FIR, well in the Fulcrum/Presonus approach it does make sense to counteract the effect of horn in the area just below crossover point. This is the most interesting point in their strategy in my view and could renew the interest in the Urei approach of coax.
Not sure it can be as interesting for a Tannoy approach coax. Once you hipassed the medium you really limit the most objectionable point (lf mod of the cd through the cone) of the principle. The second objectionable point is the smoothness of transition between horn and cone of medium which induce some issues in the high.
Maybe FIR could help about that but i doubt it. If you plan to use FIR only for filter duty it may help with potential break up issue with the woofer or protect the CD but nothing more to be gained imho. Tannoy gave some nice evidence that asym not too steep xover slope are efficient if the woofer is well behaved. Abit of delay to compensate for driver offset and this could be all needed for a quasi perfect point source.
Don t take me wrong though i use a FIR filter with my present loudspeakers and plan to keep/use it with my next speakers which are going to use coaxial.
Last thing, have you seen the latest TAD range of speakers? It seems there is a coax revival at the moment.
Your plan seems like a good one to me ( in fact it is not very disimilar to what i plan for my next loudspeaker which is itself influenced by a commercial product availlable at the moment!
).About FIR, well in the Fulcrum/Presonus approach it does make sense to counteract the effect of horn in the area just below crossover point. This is the most interesting point in their strategy in my view and could renew the interest in the Urei approach of coax.
Not sure it can be as interesting for a Tannoy approach coax. Once you hipassed the medium you really limit the most objectionable point (lf mod of the cd through the cone) of the principle. The second objectionable point is the smoothness of transition between horn and cone of medium which induce some issues in the high.
Maybe FIR could help about that but i doubt it. If you plan to use FIR only for filter duty it may help with potential break up issue with the woofer or protect the CD but nothing more to be gained imho. Tannoy gave some nice evidence that asym not too steep xover slope are efficient if the woofer is well behaved. Abit of delay to compensate for driver offset and this could be all needed for a quasi perfect point source.
Don t take me wrong though i use a FIR filter with my present loudspeakers and plan to keep/use it with my next speakers which are going to use coaxial.
Last thing, have you seen the latest TAD range of speakers? It seems there is a coax revival at the moment.
Last edited:
Yeah, it seems a clever thing with fulcrum is they measure frequency response and somehow differentiate timedomain related anomalies from non td anomalies. Fix those with FIR and the rest maybe with IIR so they can provide DSP presets for their speakers for multiple DSP products, some of which don't support FIR. I think they think that using IIR filters for TD related frequency response issues don't make better sound. When recording, disable FIR filters and enjoy almost as good sound (flattish fr) with no delay.
Hi Charles,
Yes but in my case and with the hardware i use at the frequency of xover between woofer and CD this is not more that around 1ms for very steep slope (70db/oct<). For sub to woofer xover i agree forget about it for realtime events!*
Hi Tmuikku, this is what i understood too. At the moment i have the solution you give: two preset one with FIR the other one with IIR. Save a lot of headache when recording.
*edit: in my case it could be possible to filter up to 48db/oct linear phase down to 250hz with 1,25ms latency of treatment. But i'm not sure Wae could afford to use Labgruppen amp for the whole system if he wish to stay competitive.
Yes but in my case and with the hardware i use at the frequency of xover between woofer and CD this is not more that around 1ms for very steep slope (70db/oct<). For sub to woofer xover i agree forget about it for realtime events!*
Hi Tmuikku, this is what i understood too. At the moment i have the solution you give: two preset one with FIR the other one with IIR. Save a lot of headache when recording.
*edit: in my case it could be possible to filter up to 48db/oct linear phase down to 250hz with 1,25ms latency of treatment. But i'm not sure Wae could afford to use Labgruppen amp for the whole system if he wish to stay competitive.
Last edited:
I haven't heard a Sceptre from Presonus but would like to. As i see it in the case of the horn FIR could be of help because the artefact is 'fixed'. In the case of a Tannoy coax the transition of horn/woofer is not by definition so i suppose the same kind of approach could not be used as artefact is dynamic. What do you think about that, maybe i'm wrong about it?
Ok so we arrive to same conclusion. This does make sense their dsp treatment can't fix distortion. But as far as i understand it can counteract diffraction in a way. Is this seen as distortion?
Implementing it to Tannoy coax would need to have some kind of predictive (and delayed like some pro digital limiter sidechain if you see what i mean) action and modelisation. It should be possible but not something i would do by myself even if i had knowledge and gear to do it.
Edit: i was probably drunk when i said i could cut at 250hz with 8order linear filters! With 1,25ms lower limit is 1khz, 2,5ms is 500hz, 5ms is 250hz, 10ms is 125hz 20ms is 62,5 and 40ms is 31,25hz... make smuch more sense and i could have lived with only one preset if that was possible.
Implementing it to Tannoy coax would need to have some kind of predictive (and delayed like some pro digital limiter sidechain if you see what i mean) action and modelisation. It should be possible but not something i would do by myself even if i had knowledge and gear to do it.
Edit: i was probably drunk when i said i could cut at 250hz with 8order linear filters! With 1,25ms lower limit is 1khz, 2,5ms is 500hz, 5ms is 250hz, 10ms is 125hz 20ms is 62,5 and 40ms is 31,25hz... make smuch more sense and i could have lived with only one preset if that was possible.
Last edited:
sorry krivium, I'm not sure if I replied all your questions. I i'm in a hurry and as i said i'm not pro, just trying to get my head around what the fulcrum does since it is fresh thinking and implementation for old and mature technology they are doing
By distortion i ment nonlinear effect the woofer has as a waveguide depending on excursion and frequency and its own nonlinearities. Diffraction from a static boundary seems to be somewhat addressable by FIR filters from the little hints I found by Googling, especially when bounced sound wave comes back to the driver membrane it can be soaked there and prevent it ever reaching the listener. This is what i've understood
Anyway, lets keep on topic.
By distortion i ment nonlinear effect the woofer has as a waveguide depending on excursion and frequency and its own nonlinearities. Diffraction from a static boundary seems to be somewhat addressable by FIR filters from the little hints I found by Googling, especially when bounced sound wave comes back to the driver membrane it can be soaked there and prevent it ever reaching the listener. This is what i've understood
Anyway, lets keep on topic.
That is fine Tmuikku, when i asked about distortion this was related to the diffraction which in a way could be seen as a distortion ( well this is how i see it). The horn being fixed this is much less problematic to counteract than a system which is highly non linear and more or less unpredictable ( the Tannoy way).
I think we agreed.
You are right let's stay on topic.
I think we agreed.
You are right let's stay on topic.
Horn Coax or Cone-waveguide COAX
Thanks guys for you thoughts. It won't be easy to figure this out (at least not to me). I think whatever choice of Coax we make, we still need to solve a number of issues.
Started reading the Fulcrum approach. Partly I believe, they try to fix the issues of Waves bouncing back into the start of the HF unit as a result of the impacted waves at the coupling of the small (Tannoy design) horn to the COAX-LF cone Then these waves radiate out gain, of which some will bounce back again. This way, they claim that this is a continual ongoing proces. These waves are "delayed" in time because of this (for one note, it will die out in a number of bounces, which will cause delay of part of the waves), which they refer to as "Smearing". The complete set of waves (the initial followed by the bounced back ones) will reach our ears over a period of time. NOT having all facts at hand when writing this, WE humans sample sound. As long as this set of waves arrive within one sound sample in our brain, we experience it as ONE wave. The higher the frequency, the more waves we will "catch" in one sample. SO, WHO cares
In theorie we catch a number of waves over time, but "hear" one, so WE can't distinguish these separate wave set. DOES this call for Higher crossing of the COAX-HF so Humans can't here this artefact?
They use their named TQ technology combined with FIR filters to correct this (and a number of other) issue(s). Although, when reading some parts of the theorie make sense, but also other parts of this "new" approach make me scratch my head and wonder if it all ads up. Did not really understand it, which maybe 99% lack of knowledge from my side. More questions then answers. As an example (next to previously described, what I just can't get) is: I can understand that waves will be impacted / disrupted as a result of flowing over the coupling area of the HF horn and the LF cone. BUT ANY frequency(from the High Pass freq. upwards of the Coax-HF) will be partly disrupted at the coupling area, so I would expect ALL frequencies (from the HF only?) need to be treated with this fancy technology.
INFACT I would expect that the bounced back waves need to be cancelled out entirely at first instance. How would you do that with DSP and filters? 180 phase turning is not an option as far as I can see. The original wave is long gone, so what "anti wave" need to be applied and where is it coming from. One would need to "remember" this audio signal applied to the HF, wait for the round trip time and apply an anti wave to cancel out the bounced back signal. All this without interfering with the next fresh initial signal. Hhhmm, some magic going on here...
AND if NOT attack the whole range, which part of the waves / frequency range requires treatment, based on what formula or theory? IS a FIR filter fixing this? Are we looking at a mystery here???
Added to this mystery, I SEE a horn implementation, SO WHAT bounce back...... There is no disruption of waves, at least NOT until they are combined (only the cross-over mutually used frequencies) at exit of the horn. I can see time-alignment here and phasing issues.
Our hearing (ears / brain combination) work in a very specific way. What can be explained and even proven in theorie and also measured with the correct equipment does not mean we can hear it.
ANYWAY, this calls for more reading.
MEanwhile, I think I've decided on the test cabinet. I've opted for B&C, also because I can get them on company conditions. It will be 2 transmission lines in one box combined with a "cone-wave guide" Coax (Tannoy style) in a 2,5 liter closed box. Resulting in 1,25 liters obstacle sitting in each TL, just ignored it, the modelling software kind of supports it, i added a segment that was 1,25 liter smaller in the TL line. NO effect to bee seen / noticeable.
Driver will be 2 x 8NDL51 (which feel very much at home in a TL, according to the modelling software, and the 8CX21 Coaxial. The coax-LF will be high passed somewhere around 300 Hz, where the max excursion is around 0,3 mm (also according the modelling software). Shure that is a bigger "movement" than a fixed horn to load the coax-HF, would (the distortion) be audible remains to be seen. The HF needs to be crossed around 2.2 kHz or higher, to be decided.
I will post the design shortly, but this little version of the PM-DMT will produce around 110 dB with the Hypex modules. Also did read somewhere that the mini-DSP support FIR filters (IF that would be a route to follow). Hypex and FIR haven't found it but will ask Hypex support. If just DSP is to be applied, we could go full blown with Hypex DSP, which is very flexible.
Also will try to make an overview of called issues for a Coax, to see IF and WHAT we need to fix.
kind regards.
Thanks guys for you thoughts. It won't be easy to figure this out (at least not to me). I think whatever choice of Coax we make, we still need to solve a number of issues.
Started reading the Fulcrum approach. Partly I believe, they try to fix the issues of Waves bouncing back into the start of the HF unit as a result of the impacted waves at the coupling of the small (Tannoy design) horn to the COAX-LF cone Then these waves radiate out gain, of which some will bounce back again. This way, they claim that this is a continual ongoing proces. These waves are "delayed" in time because of this (for one note, it will die out in a number of bounces, which will cause delay of part of the waves), which they refer to as "Smearing". The complete set of waves (the initial followed by the bounced back ones) will reach our ears over a period of time. NOT having all facts at hand when writing this, WE humans sample sound. As long as this set of waves arrive within one sound sample in our brain, we experience it as ONE wave. The higher the frequency, the more waves we will "catch" in one sample. SO, WHO cares
In theorie we catch a number of waves over time, but "hear" one, so WE can't distinguish these separate wave set. DOES this call for Higher crossing of the COAX-HF so Humans can't here this artefact?
They use their named TQ technology combined with FIR filters to correct this (and a number of other) issue(s). Although, when reading some parts of the theorie make sense, but also other parts of this "new" approach make me scratch my head and wonder if it all ads up. Did not really understand it, which maybe 99% lack of knowledge from my side. More questions then answers.
As an example (next to previously described, what I just can't get) is: I can understand that waves will be impacted / disrupted as a result of flowing over the coupling area of the HF horn and the LF cone. BUT ANY frequency(from the High Pass freq. upwards of the Coax-HF) will be partly disrupted at the coupling area, so I would expect ALL frequencies (from the HF only?) need to be treated with this fancy technology. INFACT I would expect that the bounced back waves need to be cancelled out entirely at first instance. How would you do that with DSP and filters? 180 phase turning is not an option as far as I can see. The original wave is long gone, so what "anti wave" need to be applied and where is it coming from. One would need to "remember" this audio signal applied to the HF, wait for the round trip time and apply an anti wave to cancel out the bounced back signal. All this without interfering with the next fresh initial signal. Hhhmm, some magic going on here...
AND if NOT attack the whole range, which part of the waves / frequency range requires treatment, based on what formula or theory? IS a FIR filter fixing this? Are we looking at a mystery here???
Added to this mystery, I SEE a horn implementation, SO WHAT bounce back...... There is no disruption of waves, at least NOT until they are combined (only the cross-over mutually used frequencies) at exit of the horn. I can see time-alignment here and phasing issues.
Our hearing (ears / brain combination) work in a very specific way. What can be explained and even proven in theorie and also measured with the correct equipment does not mean we can hear it.
ANYWAY, this calls for more reading.
MEanwhile, I think I've decided on the test cabinet. I've opted for B&C, also because I can get them on company conditions. It will be 2 transmission lines in one box combined with a "cone-wave guide" Coax (Tannoy style) in a 2,5 liter closed box. Resulting in 1,25 liters obstacle sitting in each TL, just ignored it, the modelling software kind of supports it, i added a segment that was 1,25 liter smaller in the TL line. NO effect to bee seen / noticeable.
Driver will be 2 x 8NDL51 (which feel very much at home in a TL, according to the modelling software, and the 8CX21 Coaxial. The coax-LF will be high passed somewhere around 300 Hz, where the max excursion is around 0,3 mm (also according the modelling software). Shure that is a bigger "movement" than a fixed horn to load the coax-HF, would (the distortion) be audible remains to be seen. The HF needs to be crossed around 2.2 kHz or higher, to be decided.
I will post the design shortly, but this little version of the PM-DMT will produce around 110 dB with the Hypex modules. Also did read somewhere that the mini-DSP support FIR filters (IF that would be a route to follow). Hypex and FIR haven't found it but will ask Hypex support. If just DSP is to be applied, we could go full blown with Hypex DSP, which is very flexible.
Also will try to make an overview of called issues for a Coax, to see IF and WHAT we need to fix.
kind regards.
Last edited:
Which Coaxial... B&C OR Faital Pro
Did not spent much time on the project last weeks, BUT this one question keeps running around in my head:
"Which of the 2 coaxials are we going to use in the test version of the PM-DMT???"
B&C 8CX21 OR Faital Pro 8HX200.
The idea was to use the Coaxial in a sealed box within the dual TL box and HP the Coaxial LF (= our MF) around 300 Hz, then cross the MF to HF around 3 kHz (2k2 or higher is specified). Which means I "don't care" about the behaviour of the MF and HF OUTSIDE those working areas.
- Price wise, the B&C wins with around 100 euro's cheaper.
- Since we will use a sealed box, the LF part (being our MF in this concept) is not that critical. ALTHOUGH the B&C wins here, being slightly flatter, see "MID AREA".
- The HF of the B&C is VERY rough, as can be seen in the frequency plot attached. Looks like the Faital pro wins here. Having said that, I could correct this with the DSP. Which brings me to my doubts here, is a flat frequency response by design of the driver (it's behaviour) "the same" as a DSP corrected originally "non flat driver response", resulting in a flat response.
The DSP processes the signal anyway being a 3-way system, but for the B&C we need to correct at least 4 !! areas using VERY high Q-factor filter settings, where the Faital could be passed almost straight through, with a gain reduction of 1 area.
I don't care that much in spending some time figuring out the optimal DSP settings.
- Dispersion wise, I don't have enough data at this point in time. The Faital seems to do well with around 100 degrees@2kHz to around 60 degrees @ 20kHz (-3 dB). I don't have detail on the B&C at this point in time. Provides enough sweet spot area when placed correctly for a good sound stage around the mixing desk.
- HF impedance behaviour more or less the same in the range we will use them.
- Since we use a DSP anyway, filter wise it isn't the complicating factor. Passive filtering is another ball game I suspect.
MY feeling: Although more expensive, Faital 8HX200 wins as far as the HF is concerned.
YOUR thoughts ??
kind regards,
Frans
Did not spent much time on the project last weeks, BUT this one question keeps running around in my head:
"Which of the 2 coaxials are we going to use in the test version of the PM-DMT???"
B&C 8CX21 OR Faital Pro 8HX200.
The idea was to use the Coaxial in a sealed box within the dual TL box and HP the Coaxial LF (= our MF) around 300 Hz, then cross the MF to HF around 3 kHz (2k2 or higher is specified). Which means I "don't care" about the behaviour of the MF and HF OUTSIDE those working areas.
- Price wise, the B&C wins with around 100 euro's cheaper.
- Since we will use a sealed box, the LF part (being our MF in this concept) is not that critical. ALTHOUGH the B&C wins here, being slightly flatter, see "MID AREA".
- The HF of the B&C is VERY rough, as can be seen in the frequency plot attached. Looks like the Faital pro wins here. Having said that, I could correct this with the DSP. Which brings me to my doubts here, is a flat frequency response by design of the driver (it's behaviour) "the same" as a DSP corrected originally "non flat driver response", resulting in a flat response.
The DSP processes the signal anyway being a 3-way system, but for the B&C we need to correct at least 4 !! areas using VERY high Q-factor filter settings, where the Faital could be passed almost straight through, with a gain reduction of 1 area.
I don't care that much in spending some time figuring out the optimal DSP settings.
- Dispersion wise, I don't have enough data at this point in time. The Faital seems to do well with around 100 degrees@2kHz to around 60 degrees @ 20kHz (-3 dB). I don't have detail on the B&C at this point in time. Provides enough sweet spot area when placed correctly for a good sound stage around the mixing desk.
- HF impedance behaviour more or less the same in the range we will use them.
- Since we use a DSP anyway, filter wise it isn't the complicating factor. Passive filtering is another ball game I suspect.
MY feeling: Although more expensive, Faital 8HX200 wins as far as the HF is concerned.
YOUR thoughts ??
kind regards,
Frans
Attachments
Last edited:
- Status
- This old topic is closed. If you want to reopen this topic, contact a moderator using the "Report Post" button.