The best way possible to build an active system.

[Nyal Mellor]

Yes obviously if your going to build a top notch state of the art system, the knowledge and equipment to do so will be a requirement. But the purpose of this thread is to talk about the best technology possible to use in such a system. If you setup a nice purpose built for audio PC, it can be every bit as good as what Trinnov has done, or even better.

You must have missed where I posted the review of the Merging NADAC on my first post of my thread:


http://positive-feedback.com/audio-discourse/impressions-the-merging-technology-nadac-mc-8-dsd-dac/


It's light years above the Exasound E28. In fact the best DAC David Robinson has ever reviewed at any price. I'm not sure where the measurement data or the feedback you heard about this DAC came from. I haven't heard any feedback from anyone who has actually spent some time with one who wasn't blown away.


I think the closing paragraph to that review sums things up pretty good:

"If you have the budget to play in this range, there's simply no question that you need to put the NADAC at the top of your list of DSD DACs. If you can't get to a show or another site that's equipped with one to hear it for yourself, I'm going to go out on a limb and tell you that I would recommend a purchase, regardless. I'm confident that it's really that good…it's the best that I've heard in DSD and high-resolution PCM/DXD DACs to date!

Period.

Run. Do not walk. Get what I'm experiencing for yourself!"

And I'm quite familiar with Exasound. George and his team make excellent products. Here's my master clock modded E20 with a matching custom ultra low noise linear supply I built for it:

View attachment 22595
View attachment 22596

The specs on the Exasound E22 are better than the NADAC. Check out the measurements published on both sites. Whether that translates to SQ differences may be a moot point as the performance of both is superb. Reviewer hyperbole aside I don't see anything in Merging's DAC implementation that suggests significantly better performance than the Exasound. Seems pretty bog standard to me except for the network interface. Personally I just think you are focusing on the wrong thing. The best possible way to build an active system is to design the speakers first, then work out what DSP you need, then figure out how to implement said DSP (and whether you need to write your own code or can use something like accourate). The speaker and how it interfaces with the room is the biggest sound quality hurdle facing us in the present day, not the differences between two DACs that are within a hair nose of each other performance wise.

 

[dallasjustice]

DJ, what are you currently running for a total system? Did you actually go fully active yet?

I use symmetrical 2 octave digital crossovers for 2 subs and my full range speakers. So it's a four channel system for stereo playback. I own and have used both Audiolense and Acourate. I've also used Dirac in the past, but don't use it anymore.

The issues you will run into have to do with hardware/software ASIO interface. I'm not going into it here because it's not important. But if you are really are serious, talk to Uli and you will get very good info about the requirements needed.

Btw, there are differing opinions about whether speakers can be measured in normal rooms quasi anechoic. I think Uli would give you a different opinion than me. But if we are talking "best practices", the actual driver measurements will be lower resolution in a normal room because you will need to window the impulse before the first reflection due to lack of real estate.

Please understand, I'm not talking about the measurement at seated position. That's a different type of measurement and different filters are needed for those types of measurements. I'm just talking about driver linearization. Driver linearization and crossovers have to be done in an optimal way which takes into account the driver polar response over frequency range and cabinet/baffle effects (eg. Baffle step effect and diffraction).

 

[Blizzard]

The specs on the Exasound E22 are better than the NADAC. Check out the measurements published on both sites. Whether that translates to SQ differences may be a moot point as the performance of both is superb. Reviewer hyperbole aside I don't see anything in Merging's DAC implementation that suggests significantly better performance than the Exasound. Seems pretty bog standard to me except for the network interface. Personally I just think you are focusing on the wrong thing. The best possible way to build an active system is to design the speakers first, then work out what DSP you need, then figure out how to implement said DSP (and whether you need to write your own code or can use something like accourate). The speaker and how it interfaces with the room is the biggest sound quality hurdle facing us in the present day, not the differences between two DACs that are within a hair nose of each other performance wise.

I don't even see a single published spec on the manufacturers website that even uses the same measurement standards to even compare. Besides, you can't just judge a DAC based on specs and the chip used. There's $100 DAC's that use Sabre chips and $15000 DAC's that use Sabre chips. There's much more to the total implementation than the DAC chip.

I'm quite aware how to build speakers as I've built hundreds of sets over the last 20 years. Here's the unit's I'm currently using to test the electronics:


And the difference between the ADC/ DAC sections used in high end stand alone ADC/ DAC's, and the ADC/ DAC sections used in the all in one DSP boxes are not a nose hair apart. Try running one of your DEQX boxes in unity mode comparing the DAC section to the Exasound E22 in a 2 channel passive system. After that try the same with the Trinnov. Share the results here.

 

[Blizzard]

I use symmetrical 2 octave digital crossovers for 2 subs and my full range speakers. So it's a four channel system for stereo playback. I own and have used both Audiolense and Acourate. I've also used Dirac in the past, but don't use it anymore.

The issues you will run into have to do with hardware/software ASIO interface. I'm not going into it here because it's not important. But if you are really are serious, talk to Uli and you will get very good info about the requirements needed.

Btw, there are differing opinions about whether speakers can be measured in normal rooms quasi anechoic. I think Uli would give you a different opinion than me. But if we are talking "best practices", the actual driver measurements will be lower resolution in a normal room because you will need to window the impulse before the first reflection due to lack of real estate.

Please understand, I'm not talking about the measurement at seated position. That's a different type of measurement and different filters are needed for those types of measurements. I'm just talking about driver linearization. Driver linearization and crossovers have to be done in an optimal way which takes into account the driver polar response over frequency range and cabinet/baffle effects (eg. Baffle step effect and diffraction).

What I mean't with my question is what component's are you using in your system? Amps, DAC's , Speakers, cables, turn table, etc.

 

[Nyal Mellor]

I don't even see a single published spec on the manufacturers website that even uses the same measurement stands to even compare. Besides, you can't just judge a DAC based on specs and the chip used. There's $100 DAC's that use Sabre chips and $15000 DAC's that use Sabre chips. There's much more to the total implementation than the DAC chip.

I'm quite aware how to build speakers as I've built hundreds of sets over the last 20 years. Here's the unit's I'm currently using to test the electronics:

View attachment 22601

And the difference between the ADC/ DAC sections used in high end stand alone ADC/ DAC's, and the ADC/ DAC sections used in the all in one DSP boxes are not a nose hair apart. Try running one of your DEQX boxes in unity mode comparing the DAC section to the Exasound E22 in a 2 channel passive system. After that try the same with the Trinnov. Share the results here.

Your thread was titled "the best possible way to build an active system". In that regard I stand by my assertion that you should start with the speaker and work outward. IMO you are getting way too hung up on the electronics. It's like saying the only thing that matters about non-active speaker design is how the passive crossover is done and nothing else matters.

Again you misread my post. I said that the Exasound and NADAC were probably a hair nose apart since they both have exceptional measurements. I have not auditioned the NADAC.

As for DEQX, Trinnov. I've never asserted they are SOTA and that their DAC performance could not be bettered. DEQX PreMATE does not have SOTA DAC performance, but it is very good. I know what the DEQX sounds like as a DAC and how it compares to other products at the same price point like the ExaSound and also those at higher price points like the Bricasti M1. I believe they are getting 19 bits of resolution and all other measurements are good (as per Stereophile review). I don't think doing a crossover on a Windows/OSX computer as you seem to be wanting to do is SOTA either. I'd think you'd want something like what Trinnov has done which is a Linux box with an Intel processor running custom DSP software. You could then feed into a multichannel DAC.

 

[dallasjustice]

Linkwitz LX521

Your thread was titled "the best possible way to build an active system". In that regard I stand by my assertion that you should start with the speaker and work outward. IMO you are getting way too hung up on the electronics. It's like saying the only thing that matters about non-active speaker design is how the passive crossover is done and nothing else matters.

Again you misread my post. I said that the Exasound and NADAC were probably a hair nose apart since they both have exceptional measurements. I have not auditioned the NADAC.

As for DEQX, Trinnov. I've never asserted they are SOTA and that their DAC performance could not be bettered. DEQX PreMATE does not have SOTA DAC performance, but it is very good. I know what the DEQX sounds like as a DAC and how it compares to other products at the same price point like the ExaSound and also those at higher price points like the Bricasti M1. I believe they are getting 19 bits of resolution and all other measurements are good (as per Stereophile review). I don't think doing a crossover on a Windows/OSX computer as you seem to be wanting to do is SOTA either. I'd think you'd want something like what Trinnov has done which is a Linux box with an Intel processor running custom DSP software. You could then feed into a multichannel DAC.

I agree with Nyal. Just to add another data point which reinforces this perspective. I heard a Linkwitz LX521 at RMAF last weekend. It was one of the best systems at the show. Many others who heard it agreed. That is a well designed active speaker for sure. It didn't sound like other omnidirectional speakers I've heard. The soundstage was very precise and totally effortless and natural. BUT, the electronics were the least audiophile of the show. The amp was a MCH emotiva and the digital/DAC was a cheap MiniDSP unit which retails for $499.
http://www.minidsp.com/products/minidsp-in-a-box/minidsp-4x10-hd

As long as one doesn't listen with their eyes, I doubt many folks would disagree if they heard the LX521. It's all about implementation and the speaker/room is 90% of what you actually hear. The rest of it is just icing on the cake, if that. Just my opinion, of course.

 

[Blizzard]

Your thread was titled "the best possible way to build an active system". In that regard I stand by my assertion that you should start with the speaker and work outward. IMO you are getting way too hung up on the electronics. It's like saying the only thing that matters about non-active speaker design is how the passive crossover is done and nothing else matters.

Again you misread my post. I said that the Exasound and NADAC were probably a hair nose apart since they both have exceptional measurements. I have not auditioned the NADAC.

As for DEQX, Trinnov. I've never asserted they are SOTA and that their DAC performance could not be bettered. DEQX PreMATE does not have SOTA DAC performance, but it is very good. I know what the DEQX sounds like as a DAC and how it compares to other products at the same price point like the ExaSound and also those at higher price points like the Bricasti M1. I believe they are getting 19 bits of resolution and all other measurements are good (as per Stereophile review). I don't think doing a crossover on a Windows/OSX computer as you seem to be wanting to do is SOTA either. I'd think you'd want something like what Trinnov has done which is a Linux box with an Intel processor running custom DSP software. You could then feed into a multichannel DAC.

The electronics are much more important in an active DSP based system, than the passive components in a passive system. This is because the DAC, preamp and everything before the amps are all an integral part of the system. Do you not think that choice of DAC is important in a 2 channel passive system?

As I said in an earlier post, the complete system can only be as good as the weakest link. I just posted pictures of the speakers I'm using for testing, so obviously that end of things is complete. I'm on to electronics testing now. Once I find the best way this can be done, the same system can be applied to various active designs.

Actually you said: "The specs on the Exasound E22 are better than the NADAC. Check out the measurements published on both sites"


I've explained the system I'll be using in prior posts. The reason I brought up Acourate running on Windows is because I feel that paired with a good multichannel DAC will kill any stand alone DSP box.

I was kind of looking for feedback from people who have actually compared the configurations, instead of basing on assumptions. As I haven't compared every DSP box solution myself.

 

[Blizzard]

I agree with Nyal. Just to add another data point which reinforces this perspective. I heard a Linkwitz LX521 at RMAF last weekend. It was one of the best systems at the show. Many others who heard it agreed. That is a well designed active speaker for sure. It didn't sound like other omnidirectional speakers I've heard. The soundstage was very precise and totally effortless and natural. BUT, the electronics were the least audiophile of the show. The amp was a MCH emotiva and the digital/DAC was a cheap MiniDSP unit which retails for $499.
http://www.minidsp.com/products/minidsp-in-a-box/minidsp-4x10-hd

As long as one doesn't listen with their eyes, I doubt many folks would disagree if they heard the LX521. It's all about implementation and the speaker/room is 90% of what you actually hear. The rest of it is just icing on the cake, if that. Just my opinion, of course.

I agree the LX521 is an excellent design. In fact there's a guy out there who used a PC DSP based setup with them and said nothing comes close having tried DEQX, Minidsp etc. And he was just using DIY Buffalo DAC's.

 

[Blizzard]

I have compared optical to spdif out with the SB touch and they sound identical to me.

I will try another way .. cue up 2 squeezeboxes (I have a few) , sych them , one plays thru the DSP box , the other into the amp .. both would use SPDIF...

I found this thread on the minidsp forum of a guy who was using a squeezebox as well with his MiniSHARC/IO board combo. Same thing you have but not in a box. Not the friendliest fellow in the world. I think he was just expecting much more for the price.

Did you try out your new test yet?


http://www.minidsp.com/forum/hardware-support/9260-minisharc-sound-review-issues

 

[Keith_W]

Your thread was titled "the best possible way to build an active system". In that regard I stand by my assertion that you should start with the speaker and work outward. IMO you are getting way too hung up on the electronics.

With respect Nyal, my impression is that he is after the best possible digital crossover, cost no object (see first post). I am not sure how much experience he has with other digital crossovers, but I have experience with DEQX and MiniDSP. My feeling is that both of them can be improved upon, as far as sound quality goes. Incidentally, I had a conversation with an amp designer in Melbourne recently. He said that the problem with consumer level digital crossovers is that they are way below what the pros are using. Professional DSP units use 64 bit floating point processors. He felt that if these consumer DSP products were to use pro grade ADC, internal processing, and DAC ... they would have a fighting chance of being transparent. But at the moment, if your priority is transparency, an analog active crossover is superior (of course you lose all the other benefits of DSP). You then have to weigh up what you wish to sacrifice. His opinion, and mine.

This is why I totally support what Blizzard is doing. I would be particularly interested to see him compare his solution to something like a DEQX or a MiniDSP.

May I ask which iteration of DEQX you are using? I have a HDP-3. I am curious what the HDP-5 sounds like, I might contact a dealer and arrange for a loaner.

 

[Blizzard]

With respect Nyal, my impression is that he is after the best possible digital crossover, cost no object (see first post). I am not sure how much experience he has with other digital crossovers, but I have experience with DEQX and MiniDSP. My feeling is that both of them can be improved upon, as far as sound quality goes. Incidentally, I had a conversation with an amp designer in Melbourne recently. He said that the problem with consumer level digital crossovers is that they are way below what the pros are using. Professional DSP units use 64 bit floating point processors. He felt that if these consumer DSP products were to use pro grade ADC, internal processing, and DAC ... they would have a fighting chance of being transparent. But at the moment, if your priority is transparency, an analog active crossover is superior (of course you lose all the other benefits of DSP). You then have to weigh up what you wish to sacrifice. His opinion, and mine.

This is why I totally support what Blizzard is doing. I would be particularly interested to see him compare his solution to something like a DEQX or a MiniDSP.

May I ask which iteration of DEQX you are using? I have a HDP-3. I am curious what the HDP-5 sounds like, I might contact a dealer and arrange for a loaner.

Exactly. The best possible way means compromised DAC's absolutely can not be used. My DAC reference for this system is my Resonessence Labs Mirus playing tracks from the internal SD card reader. The Mirus was designed 100% by Dustin Forman (the man who designed the ESS Sabre DAC chip) It's his version of the ultimate implementation of the ESS 9018 chip. It uses dual ESS 9018 chips in monoblock mode for the highest performance possible. The SD card transport is probably the lowest jitter transport in the industry at any cost. When you put the SD card in, the FPGA (which is synchronously clocked to the DAC master) buffers the tracks into a ram chip. From there, straight over to the DAC chip's. The only jitter in this system is the jitter of the master clock itself (which I also had factory upgraded). It's also the only DAC that uses the Sabre's internal async feature the way it was intended to be used.

The quality of this DAC even through the USB port, is far and above the Exasound E22. Night and day. And via the SD card transport, it's in another world. So for one of these DSP boxes to meet the minimum requirements for my system, the first test will be to put in unity mode, and compare with the Mirus playing tracks via the SD card. If it's not at least as good as the Mirus, it's absolutely unacceptable for this project.

 

[Barry]

I have experience with DEQX and MiniDSP. My feeling is that both of them can be improved upon, as far as sound quality goes. The problem with consumer level digital crossovers is that they are way below what the pros are using. Professional DSP units use 64 bit floating point processors. If these consumer DSP products were to use pro grade ADC, internal processing, and DAC ... they would have a fighting chance of being transparent. But at the moment, if your priority is transparency, an analog active crossover is superior (of course you lose all the other benefits of DSP).

This is very interesting as I've always wondered about the transparency of DSP products. You've convinced me that I shouldn't ditch my Pass Labs XVR-1-3 way active analog crossover. I do have DSP on the subs only, after the crossover. Even with a high-pass XO Fq of only 80 Hz and 24dB/octave slope, there is still plenty of usable bandwidth to correct my room problems up to almost 300 Hz.

 

[DaveC]

The best speaker I have heard (Lotus Granada) was active but used pretty basic DSP (parametric EQ, delays, etc) and A/D, D/A converters. I guess the message is that the digital side is much lower priority than the electro-acoustical speaker engineering part.

Agreed, the Granada is definitely in the top few speakers that I have ever heard and certainly the only dipole I really like.

 

[Blizzard]

This is very interesting as I've always wondered about the transparency of DSP products. You've convinced me that I shouldn't ditch my Pass Labs XVR-1-3 way active analog crossover. I do have DSP on the subs only, after the crossover. Even with a high-pass XO Fq of only 80 Hz and 24dB/octave slope, there is still plenty of usable bandwidth to correct my room problems up to almost 300 Hz.

Would be interesting to compare the XVR-1 with an Intel broadwell I7 NUC running Acourate/Jriver combined with Exasound E28. With analog Xovers you still need a 2 channel DAC for digital sources, and some sort or transport (computer). So it can get expensive. With the NUC/Acourate/Jriver/Exasound E28 solution, all your missing is the amps and speakers. And we are talking around $5000 total cost.

 

[anders]

Blizzard I look forward to seeing what you come up with, you said you were going with custom designed/built DACs?

After hearing the full blown Kyron system I am convinced that full digital with all digital crossovers and completely multi amped is the reference for high fidelity. Absolutely nothing else comes close for realism.

Personally I would not be so tied down to just DSD, IMHO all the reviews/debates about DSD vs PCM are done on old fashioned passive speakers. On a fully active digital system even 'lowly' 16/44 classical of a well recorded piano sonata sounded like the real thing (timber, detail, soundstage, and it's physical presence).

 

[Blizzard]

Blizzard I look forward to seeing what you come up with, you said you were going with custom designed/built DACs?

After hearing the full blown Kyron system I am convinced that full digital with all digital crossovers and completely multi amped is the reference for high fidelity. Absolutely nothing else comes close for realism.

Personally I would not be so tied down to just DSD, IMHO all the reviews/debates about DSD vs PCM are done on old fashioned passive speakers. On a fully active digital system even 'lowly' 16/44 classical of a well recorded piano sonata sounded like the real thing (timber, detail, soundstage, and it's physical presence).

I won't be tied down to just DSD. My system will be PCM compatible up to 32/768. Yes this will require a custom DAC solution. I hear what your saying about even 16/44 PCM sounding awesome fully active. Now just imagine DSD 256. The same improvements heard between 16/44 PCM and DSD 256 on a passive system, also apply to active.

 

[anders]

Are Resonessence building you a custom multichannel DAC or is DIY? If it's the latter I have heard the ESS DAC with a Class A (I think modified Nelson Pass Zen) output stage and it was very good. There are some threads about it on DIYAudio. I think the route will be taking is a bit different from yours since I will be digital only and don't need the A/D conversion for vinyl. After speaking with Uli at Accourate I think ultimately I am going to do room and crossovers computer side. It seems like that would be the most flexible and best sounding solution since there is no A->D process; everything is done digitally and only the DAC (where the quality is in my control) performs a conversion to analog.

Also most commercial multi-channel DACs are limited to roughly 8 channels. For now that suits my needs, but if I plan to get truly insane with it with distributed sub networks (ala Geddes) that will require even more channels.

No argument from me that higher resolution digital sounds better. But the vast majority of the SQ is already determined by the recording and/or mixing phase, so I don't worry myself about resolution much these days. Instead trying to find the most neutral, least harmful mastering of a particular album/classical piece.

 

[Blizzard]

Are Resonessence building you a custom multichannel DAC or is DIY? If it's the latter I have heard the ESS DAC with a Class A (I think modified Nelson Pass Zen) output stage and it was very good. There are some threads about it on DIYAudio. I think the route will be taking is a bit different from yours since I will be digital only and don't need the A/D conversion for vinyl. After speaking with Uli at Accourate I think ultimately I am going to do room and crossovers computer side. It seems like that would be the most flexible and best sounding solution since there is no A->D process; everything is done digitally and only the DAC (where the quality is in my control) performs a conversion to analog.

Also most commercial multi-channel DACs are limited to roughly 8 channels. For now that suits my needs, but if I plan to get truly insane with it with distributed sub networks (ala Geddes) that will require even more channels.

No argument from me that higher resolution digital sounds better. But the vast majority of the SQ is already determined by the recording and/or mixing phase, so I don't worry myself about resolution much these days. Instead trying to find the most neutral, least harmful mastering of a particular album/classical piece.

Resonessence was going to do my DAC's at one point. Dustin was over at my place a few times listening to my system. We had several meetings with Mark and Martin Mallinson. But turned out that Dustin is just too bogged down running the Sabre chip team at ESS to take on such an elaborate project. So he offered his support where he can, and I found someone else who is an extremely competent DAC designer for the task.

The system I'm working on will be scalable up to 48 channels of DSD 256. So the sky's the limit if your bank account can handle it.

As far as A/D, my system is going to be digital source only. If one was going to use it with analog sources, I would reccommend making DSD 256 rips with a Merging Hapi. Or just trust that the guys who have the master tapes at the studio, are capable of making better digital copies than you will be able to from vinyl at home. In some cases, yes might be the case, but not all. I've heard some exceptional vinyl rips.

 

[Keith_W]

This is very interesting as I've always wondered about the transparency of DSP products. You've convinced me that I shouldn't ditch my Pass Labs XVR-1-3 way active analog crossover. I do have DSP on the subs only, after the crossover. Even with a high-pass XO Fq of only 80 Hz and 24dB/octave slope, there is still plenty of usable bandwidth to correct my room problems up to almost 300 Hz.

I wouldn't dismiss DSP altogether. As I have mentioned elsewhere, DSP might work for you. In some applications, it may actually be superior to what you are currently running. If your system has severe issues with frequency response, or time alignment, or phase ... you may gain more than you lose. It just so happens that with my system, what I gained was not worth what I lost. But just because it did not work for me, won't mean that it won't work for you

To the other poster who mentioned the Kyron, I have heard them and I wasn't particularly impressed. They sounded really good with your typical audiophile guff (female solos, small jazz ensembles, etc). I was in the room, really impressed at the coherence and impact of the thing. Then, I asked if I could play a disc I brought along ...

Unfortunately, if you throw complex classical music at it, it somehow falls flat. I heard on that system what I hear in mine with the DEQX controlling the highs - the texture in strings goes missing, and it seems to sound flat and lifeless. And this was without analog to digital conversion, their transport was feeding digital into the DEQX. This was what convinced me years ago that it wasn't ready for prime time, and made me wonder what Mike Fremer was listening to if he could not hear its obvious flaws.

 

[Blizzard]

I wouldn't dismiss DSP altogether. As I have mentioned elsewhere, DSP might work for you. In some applications, it may actually be superior to what you are currently running. If your system has severe issues with frequency response, or time alignment, or phase ... you may gain more than you lose. It just so happens that with my system, what I gained was not worth what I lost. But just because it did not work for me, won't mean that it won't work for you

To the other poster who mentioned the Kyron, I have heard them and I wasn't particularly impressed. They sounded really good with your typical audiophile guff (female solos, small jazz ensembles, etc). I was in the room, really impressed at the coherence and impact of the thing. Then, I asked if I could play a disc I brought along ...

Unfortunately, if you throw complex classical music at it, it somehow falls flat. I heard on that system what I hear in mine with the DEQX controlling the highs - the texture in strings goes missing, and it seems to sound flat and lifeless. And this was without analog to digital conversion, their transport was feeding digital into the DEQX. This was what convinced me years ago that it wasn't ready for prime time, and made me wonder what Mike Fremer was listening to if he could not hear its obvious flaws.

My exact point that any system can only be as good as the weakest link. When people hear an active system for the first time, or are just used to passive speakers, the advantages of directly coupling the amps to the drivers pop out right away. It's easy to just focus on the positive attributes that your not used to. But then after a while you start to realize that great DAC's were created for a reason. And when the role of a great DAC is replaced by a cheap DAC, the more transparent the system is, the more the flaws shine through.