Anyone, Snaxo vs Exakt ?

Huge posted:

In theory, the Linn exact system should be better in that it can be precisely adapted to the specific drive units in the individual speakers you have.  However, that comes at the risk of more noticeable mathematical artefacts being created, particularly as the drivers age.

On the other hand with analogue crossovers, although there are likely to be more 'artefacts' due to the lower precision of the matching, these are of a more 'natural' sounding form as the analogue systems can't have such sharp frequency response / phase response features.  With a well designed speaker systems these artefacts will still be less than +/-2.5dB anyway; and, in practice, non-idealities in the room response will usually be much larger than this.

Now we're talking, A Gentleman that introduced me to Linn, about thirty years ago, who happens to be employed by a Linn Dealership, shared that he wasn't all that awed by Linn's  Aktiv Exakt. Paraphrasing, to much of a burden on the dacs, incorporated within the Exakt. 

As you say, in theory it sounds nice, capture and preserve the digital signal early in the game, then convert the digital signal to analog, late in the game! 

Although @ that price it should be a killer! 

Let's take a look at a potential rear end again, from Linn perspective! 

Exakt ~ 10K GBP

3 x 500 ~ 54K

Kudos ~ Titans 20K

That's 84K, not including Linn's Digital Front End! 

I owned my briks nearly thirty years, but I know they aren't no Kudos Titans, or for the Matter DB's Ovators.

Imagine if we could do that comparison with the same rear end, 3 x 500s/ Titan 808

as opposed to Briks, which appears the only Speaker Available that will work with Both Linn & Naim. 

Hope Someone is Listening! 

Linn's Digital Front end vs Naims Analog Front end! 

Room & Speakers held constant! 

Theory wise, I know Hugh, looks like you and my Friend, gives the edge to Snaxo's tested and tried analog approach. 

Mind you, My Friend has been affiliated with Linn for Decades, and also very familiar with Naim. 

I think we are getting Close, Bristol show, Naim Amps with Titans, and Berhinger X0. 

Linn is forcing Focal/Naim hand, by reaching out to KUDOS, PMC, B&W, KEF, AND JBL.

Now its time for Focal/Naim to call ! 

If one wants to increase their share of the Active Market, They must reach out to Third Party Speaker Manufactures.

The Merger of Naim & Focal should start to pay dividends, allowing Naim to focus on electronics, and allowing Focal to focus on Speakers, compatible with Naim's electronics.

INCLUDING THE SNAXO! 

Just some thoughts! 

Allante93!

ALLANTE93 Posted

"Now we're talking, A Gentleman that introduced me to Linn, about thirty years ago, who happens to be employed by a Linn Dealership, shared that he wasn't all that awed by Linn's  Aktiv Exakt. Paraphrasing, to much of a burden on the dacs, incorporated within the Exakt. "

Tha's an odd view - the DACs in Exakt have no more to do than the DACs in a DS or DSM. In fact, as there is a DAC per driver, you could say they have less to do, but that's really a bit meaningless as they have to process the same number of same length words per second as they do in a DS - they just output a much smaller analogue frequency range.  The Exakt processing is done in the "Exakt Engine" (which I assume is an FPGA) not in the DAC.  As mentioned in an earlier post, there seems to be a gap in the education programme.

In principle, I agree with the 'same amount of work in the DACs', however one could perhaps argue that there may be a greater dependency on the tracking of all the DACs in each of the channels driving a particular loudspeaker assembly.  However do I think that, in practice, tracking accuracy will easily be sufficient for the job, so I don't see a real problem here.

Just a minor point though, why assume a FPGA?  Although I don't know the internal technology Linn use, I would have thought that one or more DSP chips would be a more likely solution for the digital signal processing.

sunbeamgls posted:

ALLANTE93 Posted

"Now we're talking, A Gentleman that introduced me to Linn, about thirty years ago, who happens to be employed by a Linn Dealership, shared that he wasn't all that awed by Linn's  Aktiv Exakt. Paraphrasing, to much of a burden on the dacs, incorporated within the Exakt. "

Tha's an odd view - the DACs in Exakt have no more to do than the DACs in a DS or DSM. In fact, as there is a DAC per driver, you could say they have less to do, but that's really a bit meaningless as they have to process the same number of same length words per second as they do in a DS - they just output a much smaller analogue frequency range.  The Exakt processing is done in the "Exakt Engine" (which I assume is an FPGA) not in the DAC.  As mentioned in an earlier post, there seems to be a gap in the education programme.

Hugh posted:

" Just a minor point though, why assume a FPGA?  Although I don't know the internal technology Linn use, I would have thought that one or more DSP chips would be a more likely solution for the digital signal processing."

There you have it, 3 opinions expressed by three technical gents.

Gentleman one, experienced in Linn gear for Nearly three decades, and very familiar with Naim gear.

In fact, I looked up some of his post, before I joined the Forum.

 Gentleman two, Sunbeamgls, I don't know his credentials, but I was only able to regurgitate 75% of his reply, so I would assume he knows his stuff.

 Gentleman three, Hugh!

Now there's one thing that all Forum members have in common, with these gents, EARS.

Time for another show, presenting Naim's Snaxo technologies, with 3rd party Speaker Manufacturer that can be used with both Linn & Naim.

Exakt/Common Rear End/Ears

Vs

Snaxo/Common Rear End/Ears

 Mind you, this is Hi End !

I did the Math, Either way:

S1 @ 57K

Nds/2 x 555PS/552

Snaxo @ 1.4K

Klimax Exakt XO @ 10K

Klimax DMS @ 15K

But Remember, Market share.

Three Active post  this very moment, one with Credos active 172, I think, another with 252/SBLs, and  DB's 500 DR Upgrade, not to mention my 282/Briks in a the workings.

That's right the trickled down technologies, and Market Share.

 Keep an eye for the next show, who knows what Focal/Naim have up their sleeves!

Allante93!

ALLANTE93: 30 years of experience with Linn doesn't indicate an understanding of a new technology they launched 2.5 years ago - but it doesn't preclude it either of course.

Huge - I assumed the Exakt Engine uses an FPGA because they're doing something unique that won't be available in most DSPs (even if you don't want to recognise this), and therefore they need something they can programme.  I have just checked and they're using this FPGA family: (Editted to remove hyperlink as I'd forgotten they are not allowed - just search for Xilinx Spartan 6)

sunbeamgls posted:

ALLANTE93: 30 years of experience with Linn doesn't indicate an understanding of a new technology they launched 2.5 years ago - but it doesn't preclude it either of course.

True, Sunbeamgls, but the Gentleman is headed to work right now!

You got, an Linn Dealership in the USA!

He might know a little bit, about these technologies you guys are discussing!

Allante93!

Unless they've invented an entirely new class of mathematical algorithms (and why do that when the current ones work and cover the required operations), then, mathematically it's nothing particularly special, just applying a bit of general DSP maths (multipole FIR or IIR filters, delays etc).  This is possible in a variety of different DSPs or FPGAs.   I still don't understand why you believe the mathematical processes involved are unique.

The FPGA route is more expensive in initial development/debug time, but more flexible in future development.  Perhaps, initially, they were unsure of the hardware route they wanted to follow - that would be a very sound reason for developing on a FPGA platform.

And yes I know of the Spartan product line.

Allante93 posted:

sunbeamgls posted:

ALLANTE93: 30 years of experience with Linn doesn't indicate an understanding of a new technology they launched 2.5 years ago - but it doesn't preclude it either of course.

True, Sunbeamgls, but the Gentleman is headed to work right now!

You got, an Linn Dealership in the USA!

He might know a little bit, about these technologies you guys are discussing!

Allante93!

He's not reading about this new technology, guessing which dacs are incorporated within the Exakt, applying theories,  but using what we all use, his EARS, on a daily basis!

Allante93!

Exakt/Common Rear End/Ears

Vs

Snaxo/Common Rear End/Ears

The Big Picture, whose Active technologies dominates the 21st Century!

Snaxo vs Exakt!

Clearly the Speakers, is the missing link, that would allow for a fair comparison.

Then our EARS, can handle it from there!

Allante93!

Huge posted:

Unless they've invented an entirely new class of mathematical algorithms (and why do that when the current ones work and cover the required operations), then, mathematically it's nothing particularly special, just applying a bit of general DSP maths (multipole FIR or IIR filters, delays etc).  This is possible in a variety of different DSPs or FPGAs.   I still don't understand why you believe the mathematical processes involved are unique.

The FPGA route is more expensive in initial development/debug time, but more flexible in future development.  Perhaps, initially, they were unsure of the hardware route they wanted to follow - that would be a very sound reason for developing on a FPGA platform.

And yes I know of the Spartan product line.

Computer Engineering, I regurgitated 65% of the above, continue you guy s are having fun, that's what's the Forum is about!

Allante93, when you come down to it... +1, both your last posts (does that make it +2 or stereo?   ).

Huge posted:

The FPGA route is more expensive in initial development/debug time, but more flexible in future development.  Perhaps, initially, they were unsure of the hardware route they wanted to follow - that would be a very sound reason for developing on a FPGA platform.

Huge - are you sure - I used to use FPGAs and PLAs their predecessor   because they were quick, easy and cheap to develop on as opposed to designing and building discrete hardware... Most have quite good simulators and development environments and even in the the late 80s PLA dev environments were quite easy to use although complexity was significantly less back then  .... Clearly PIC structures are easier to use for many because they are more oriented to traditional CPU assembley language and traditional instruction set programming - but they tend to be more generalised and slower.. Although my final year undergraduate project in the late 80s I used a PIC and discrete logic hybrid design - as I required speed from the discrete logic to do some high speed calculations across a video frame store and a PIC to do the DSP processing. I was not aware of a PLA/FPGA at that time that could do those functions... it would be different now 

S

Huge posted:

Unless they've invented an entirely new class of mathematical algorithms (and why do that when the current ones work and cover the required operations), then, mathematically it's nothing particularly special, just applying a bit of general DSP maths (multipole FIR or IIR filters, delays etc).  This is possible in a variety of different DSPs or FPGAs.   I still don't understand why you believe the mathematical processes involved are unique.

The FPGA route is more expensive in initial development/debug time, but more flexible in future development.  Perhaps, initially, they were unsure of the hardware route they wanted to follow - that would be a very sound reason for developing on a FPGA platform.

And yes I know of the Spartan product line.

I never mentioned mathematical processes at all, but as you've brought it up, it's pretty clear to me that they're still learning and therefore off the shelf or existing algorithms don't do what they wanted to achieve - when Exakt was launched it was absolutely underwhelming IMHO, but 3 software and filter updates in the period November 15 to May 16 (approx) have transformed its capabilities - may be it was launched too soon to coincide with a significant historical date. Their designer's ambition was to remove phase distortions from speakers (not just at crossover points) and there doesn't seem to be any harm in that.

I just think their implementation is unique in the market and I've seen nothing here to change that view.  If its "just this" or "just that" then we'd have lots of companies delivering software driven, easy to deploy, digital crossovers with very low jitter, phase distortion removing algorithms, room optimization, frequency response tailored to the exact profile of individually measured drive units to match speaker pairs to within hundredths of a dB.  But we don't seem to be seeing that.

Of course, you don't have to be impressed, just go and listen.  If you don't like it that's fine, but maybe you will.

Allante93 posted:

Allante93 posted:

sunbeamgls posted:

ALLANTE93: 30 years of experience with Linn doesn't indicate an understanding of a new technology they launched 2.5 years ago - but it doesn't preclude it either of course.

True, Sunbeamgls, but the Gentleman is headed to work right now!

You got, an Linn Dealership in the USA!

He might know a little bit, about these technologies you guys are discussing!

Allante93!

He's not reading about this new technology, guessing which dacs are incorporated within the Exakt, applying theories,  but using what we all use, his EARS, on a daily basis!

Allante93!

Allante93, I'm not arguing a point about someone's opinion on how Exakt sounds - that would be very arrogant, as opinion is entirely personal.  I was just pointing out that the reason given doesn't seem very likely and is a symptom of the apparently less than ideal dealer education.  Exakt processing doesn't happen in the DACs and as such they are not under additional stress.

sunbeamgls posted:

Allante93 posted:

Allante93 posted:

sunbeamgls posted:

ALLANTE93: 30 years of experience with Linn doesn't indicate an understanding of a new technology they launched 2.5 years ago - but it doesn't preclude it either of course.

True, Sunbeamgls, but the Gentleman is headed to work right now!

You got, an Linn Dealership in the USA!

He might know a little bit, about these technologies you guys are discussing!

Allante93!

He's not reading about this new technology, guessing which dacs are incorporated within the Exakt, applying theories,  but using what we all use, his EARS, on a daily basis!

Allante93!

Allante93, I'm not arguing a point about someone's opinion on how Exakt sounds - that would be very arrogant, as opinion is entirely personal.  I was just pointing out that the reason given doesn't seem very likely and is a symptom of the apparently less than ideal dealer education.  Exakt processing doesn't happen in the DACs and as such they are not under additional stress.

I agree, I don't have a clue!

But, bet you three gents, would have a nice conversation.

 But, you do get my point, and Title?

Simon, it's the very rapid increase in complexity that's the problem - FPGAs used to be very simple with, for instance, low risk of unforeseen block/block interactions.  Yes they're easier than discrete logic, but not easier than assembling from well known and well designed ASICs (if they cover all the functionality you need).

In this case we're comparing to software in a DSP, not to hardware.  The programming interfaces to DSPs are similar to FPGAs in that you assemble the programmes / applications from predefined application blocks, but with DSPs the config parameters are loaded at runtime and stored dynamically in RAM.  This allows for easier optimisation in real-time using a UI, without re-programming an FPGA or recompiling the DSP code.  Later for production much of the config data in dynamic RAM (and the code) are transferred to flash ROM.

ASICs go out of production and you're left with the need to build a new physical design.  Software in an FPGA is more portable to newer hardware.  Linn (and I dare say a good few other small volume manufacturers across a number of industries) have been burned before by the need to make expensive hardware and manufacturing changes due to ASICs going out of production.

ASICs go out of production
Transistors go out of production
Capacitors go out of production
FPGAs go out of production
OEM speaker drive units go out of production
Transformers go out of production
Car engine management components go out of production
Jet engines go out of production
...

Yes Huge - ASICs really are not in vogue these days too limited and expensive to upgrade or replace. In my line of business we see  ASIC devices replaced by PIC and FPGA devices increasingly - and they cheaper and easier to update. Also very rarely see PROM or even EPROM- more FLASH these days (i.e. firmware) so one can undertake in field updates and repairs. DSP in custom FPGA devices tend  to be very modular too with the FPGAs being really a collection of programmable blocks - although these days I don't program them - I certainly use and work with the end product.... and very useful it is that they are field upgradable - it would be a disaster otherwise.....

Even in audio land - the SHARC processor used by Naim is aa rather nice FPGA of configurable DSP blocks with a nice development envionment  - the interesting development recently however IMO has been the very much improved processor/power ratio - and in closed system audio products that is useful... as I said above that would have been great to have had that power in such small devices back in the late 80s when I was quite involved with  a lot of this stuff.. it was an interesting time when DSP and audio was starting to take off and get popular ... and a lot of very new devices were coming onto the market to develop with.

S

Don't know what I'm looking at, but looks very expensive!

Upgrades & Accessories

EXAKT

Klimax Exakt 350 Modules 21,500   

Exakt Akubarik Modules 14,250 

Klimax Exaktbox (package price for 2) 16,000

The Big Picture, who dominates the Active technologies during the 21st Century?

Allante93!

Huge posted:

ASICs go out of production
Transistors go out of production
Capacitors go out of production
FPGAs go out of production
OEM speaker drive units go out of production
Transformers go out of production
Car engine management components go out of production
Jet engines go out of production
...

Starting to border on trolling now H.

Allante93 posted:

Don't know what I'm looking at, but looks very expensive!

Upgrades & Accessories

EXAKT

Klimax Exakt 350 Modules 21,500 =  10 power amps, 10 DACs and 2x Exakt engines, dealer install

Exakt Akubarik Modules 14,250 = 10 power amps, 10 DACs and 2x Exakt engines, dealer install

Klimax Exaktbox (package price for 2) 16,000 = 2x Exakt engines and 12 DACs (and 2 very nice lumps of alloy), dealer install

The Big Picture, who dominates the Active technologies during the 21st Century?

Allante93!

As above.  Not saying it is good VFM, but just letting you know what's in there.

p.s. Exakt Engine = an FPGA (the low cost hardware) and the software (the bit that consumes large amount of expensive brain power, research and testing)

Allante93 posted:

Klimax Exaktbox (package price for 2) 16,000

Where did you find this little gem please Allante93?

Allante93 posted:

 

The Big Picture, who dominates the Active technologies during the 21st Century?

I don't know if such a small market like active systems lends itself to the term domination. Active systems (in the sense of Naim/Linn) are a niche in the niche of high end audio and even high end audio is shrinking. So active systems will probably remain small scattered islands in the ocean of passive crossovers.

But since you asked: Linn !

sunbeamgls posted:

Huge posted:

ASICs go out of production
Transistors go out of production
Capacitors go out of production
FPGAs go out of production
OEM speaker drive units go out of production
Transformers go out of production
Car engine management components go out of production
Jet engines go out of production
...

Starting to border on trolling now H.

No, since you were progressively clutching at smaller and smaller straws to justify the uniqueness of the techniques in use in the Exakt system, this was just 'reductio ad absurdam'.

While it's true that Linn have put together features that previously haven't been assembled in one place, individually they're well known.