Multi-Channel, Multi-Amplifier Audio System Using Software Crossover and Multichannel-DAC

[dualazmak]

Hello mikessi and gene_stl,

Thank you so much for your invaluable various information and discussions which are really exciting and suggestive not only for myself but I believe also for many people visiting this thread. You two are quite exceptional DIY audio enthusiasts with much knowledges and actual profound experiences in speaker building and driving.

The amazing "Mikessi DIY Audio Studio" you kindly shared the photos with us seems to be the ultimate "secret base" for audio explorers, including myself. Such a DIY audio studio with many machines for wood processing and devices for acoustic measurements would be also my ultimate dream.

Your nice communications are seriously pushing me to start considering my way forward with new SP drivers with keeping the Yamaha Be-dome at the center of the SP system, and I already have some rough idea and/or sketch of my next step-forward in multichannel multi-amplifier project (to be hopefully covered by separate new thread).

I'm cautiously hesitant, however, to get into the action right away, since, in my coming new steps, the (provisionally) completed present system should be the "reference sound", so for the time being, I need to intensively listen, to understand, and to know the "current" sound and system. My two new amps, i.e. Yamaha A-S3000 and Sony TA-A1ES, would need to be burnt-in for several months or so. Yes, I will slowly and steadily consider and draw my plan for next steps while fully enjoying my music library using the (provisionally) completed present system.

Your continuing interests on this thread and kind discussions and suggestions, therefore, will be much appreciated.

P.S. -- gene-stl -- Thank you for your kind reminder on Igor Kirkwood's wonderful system with NS-1000X. I am now very carefully comparing and investigating the shapes of my Fq response curves (post #318 and post #321) with the amazing system Fq response curves Igor kindly shared in his posts #185 and #188...

The major slight difference would be in 8 kHz -20 kHz high Fq region where I use Be-TW plus Super-Tweeter T925A. It may be highly possible that, with ordinary "aging" of myself, my ear sensitivity would have become a little bit lowered in 7 kHz - 20 kHz so that my "preferable" Fq curves measured by "healthy (young ear)" microphone shows a little bit higher dB compared to those of Igor's.

It is a kind of surprise for me that the Fq response in low region, i.e. 20 Hz - 200 Hz covered by sub-woofer and woofer, seems to be quite similar for my system and Igor's. As far as only judged from the Fq shapes, in sub-woofer's 25 Hz - 45 Hz area, mine would be very slightly better than Igor's while this can be easily adjusted by sub-woofer's volume/gain.

 

[mikessi]

Dualmazak, you are too generous. It is no secret ultimate base! My wood (mostly) workshop is modified from an old firewood storage shed, and the studio was the simplest/cheapest way to add a "clean" room to this building so that I don't have to haul the speakers 60 meters back to the house to hear them. But I build many more things like rough tool sheds, outdoor furniture and work tables than audio gear. (I live on 6 acres on Salt Spring Island, a rural island 50km from Vancouver.)

This will be sacrilege to some here: What I would ideally like to do with the Yamaha BE mid dome is to use it in a 4-way, but modify it for use with a wooden horn CNC built for it to have constant directivity from <500 Hz to >5 kHz. This means getting the entire faceplate off the mid dome so that it can be mated properly to the horn. Removing the faceplate without damaging the BE dome will be a challenge, tho.

The horn design I'm thinking of is something like this ES-290 by Joseph Crowe, maybe smaller. The horn might allow the mid dome to go another hundred Hertz lower, which would be great. The whole idea is to get controlled constant directivity up & down the frequency range, maybe in a 90 or 120 degree arc, with as little diffraction as possible. Even tho some say they can't hear diffraction, it's actually measurable and it certainly can't hurt to minimize it.

A modern BE tweeter for the top, perhaps horn loaded as well?, maybe a high efficiency 12" pro driver in aperiodic box or short horn for 60-400 Hz. Separate big sealed subs for lower. Some kind of PC-based software for the crossover, OKTO DAC8 Pro (or similar). The mini-DSP 4x10 HD & OpenDRC DA8 I have are very functional & useful but rather too noisy for use with high efficiency drivers like horns.

 

[dualazmak]

Hello friends,

Throughout my long years of enthusiasm in classical music and Hi-Fi audio, I have always deeply loved lute music.

Stunning recording of lute music, therefore, is one of the many major targets for this multichannel multi-amplifier project. As you may agree, nice recordings of lute music always have wide-spread and well balanced frequency spectrum over ca. 25 Hz - 25 kHz, and which is always a kind of nice challenge to our Hi-Fi audio system.

While fully enjoying wonderful sound of lute music with my (provisionally) completed multichannel multi-amplifier system, I decided to start a new thread entitled "Lute Music: Stunning Recordings We Love" even though I fully understand it would be quite a small niche field in classical music. Your visit and participation will be much welcome, if you are also interested in lute music...

Edit: on August 25, 2022
The title of that thread has been modified into: Lute Music and Other Early Music: Stunning Recordings We Love.

 

[dualazmak]

Hello friends,

It looks that amirm will soon test, measure and evaluate the extraordinary expensive Trinnov Altitude 32 DSP processor, and an interesting thread has been started arizing/summarizing "what and how amirm should measure Trinnov".

Since I also have seriously considered and web-reviewed Trinnov Altitude 32 just before starting this thread and my project in early 2020, yesterday I posted my current feeling and thoughts on Trinnov in that thread, and the OP Spocko replied with nice comment, also followed by Scott Borduin.

In any way, soon we will see what amirm may find with Trinnov...

 

[dualazmak]

Hello friends, just for your reference and interests (if any),,,

Again(!), endless (and non-productive?) heated-up discussions are going-on on the thread entitled "Things that cannot be measured".
I posted one cynical joke comment, and also posted my personal monologue on the thread.

I found today, however, one interesting post regarding amplifier's possible higher Fq harmonic distortion(s), and one reply to it.

Then I posted my thoughts from the viewpoint of multichannel multi-driver multi-amplifier stereo audio system like my project and system shared in this thread.

I am looking forward to hearing your thoughts and/or comments, if any, here in this thread...

 

[dualazmak]

Hello again, friends,

Just for your reference and interests,,,

@Harmonie in France, who kindly joins my new thread on lute music, has started his thread entitled "How do you sort (organise) your CDs, LPs, K7s, Tapes; physical medias".

I started participating in his thread with sharing my policy and actual operations in organizing the physical LPs, CDs, DVDs, etc. as well as the LP-digitized, ripped (CD, DVD, SACD) and downloaded music and video libraries.

 

[dualazmak]

Hello friends,

Today, into my "audio sampler playlist", I added two tracks:

track-03 "Adagio"
and
track-04 "Dat Mistige Rooie Beest"

from album "Last Tango in Paris" (MYCJ-30404 M&I Jazz) by Karel Boehlee Trio (Jazz), Karel Boehlee (piano), Hein Van de Geyn (bass), Hans Van Oosterhout (drams);

The ripped CD tracks are very nice performance and wonderful recording quality. This CD and other CDs by Karel Boehlee Trio (I have all of them) are really nice for enjoyments and also as reference sound samplers for my (our) audio system.

Even though I highly recommend you to get the original CD, you can listen to YouTube clips of both the tracks...
track-03 "Adagio";

and track-04 "Dat Mistige Rooie Beest";

 

[TheBatsEar]

After skimming through the thread i have a few questions:

What is the point of the super tweeter? Shouldn't the NS-1000 tweeter reach above 20 kHz with sufficient SPL by itself?

This means, at least for my ears and brain, I can now hear much more clearly the difference in "musicality" between the amplifiers, and also the "musicality" in the combination(s) of amplifiers in multi-amp configuration.

My ears must resolve with less resolution than yours, for i always had the impression that, if the speaker load is easy enough, the Yamaha A-1020, AX-10, A-S 300, A-S 1000 and A-S 1200 sound exactly alike. Except for the A-S 1000, which i sold, i still have them around and really can't make out any difference in sound. Obviously i like the A-S 1200 best, for it's build quality and VU meters.

I suggest to organize a blind testing with a few buddies. It might change your views.

My philosophy is this: Ideally, an amp behaves like a wire with gain, without changing the frequency response at all, no coloration, no musicality, without taking or adding to the source signal. Most amps, even the cheaper ones, can do that (unless stressed). All the musicality and emotion should come from the musicians playing their tunes.

I suspect this viewpoint might look radical and maybe even crude to you ;-).

I envy you for those NS-1000, i don't think they where sold in Germany at all. A few NS-1000M exist, but you need to fork serious cash to get a pair.

 

[TheBatsEar]

View attachment 114903

Stunning speakers indeed. I seriously hope the tiny porcelain plates on top aren't some family heirloom, they are basically one Tchaikovsky away from falling down and breaking. ;-)

 

[mikessi]

Shouldn't the NS-1000 tweeter reach above 20 kHz with sufficient SPL by itself?

On paper yes, but in reality, probably not. I measured the tweeters in a pair of NS1000s I experimented with for 18 months. Their FR was only good to maybe 14~15 kHz, beyond which output dropped steeply. Not really audible to me, but...

 

[dualazmak]

....
What is the point of the super tweeter? Shouldn't the NS-1000 tweeter reach above 20 kHz with sufficient SPL by itself?
....

On paper yes, but in reality, probably not. I measured the tweeters in a pair of NS1000s I experimented with for 18 months. Their FR was only good to maybe 14~15 kHz, beyond which output dropped steeply. Not really audible to me, but...

Hello TheBatsEar and mikessi,

I essentially agree with you. YAMAHA NS-1000's Be-tweeter is still very nice. I add FOSTEX T925A super-tweeter, therefore, in relatively low gain at above ca. 8800 Hz which is well overlapped with Be-tweeter.

One another important reason for having T925A in the quite unique alignment/position under the NS1000's cabinet enabling excellent 3D perspectives of high-range sound over ca. 9000 Hz. As for this unique alignment/positioning of T925A, please refer to my post #27 on this thread.

 

[dualazmak]

....
My ears must resolve with less resolution than yours, for i always had the impression that, if the speaker load is easy enough, the Yamaha A-1020, AX-10, A-S 300, A-S 1000 and A-S 1200 sound exactly alike. Except for the A-S 1000, which i sold, i still have them around and really can't make out any difference in sound. Obviously i like the A-S 1200 best, for it's build quality and VU meters.

I suggest to organize a blind testing with a few buddies. It might change your views.

My philosophy is this: Ideally, an amp behaves like a wire with gain, without changing the frequency response at all, no coloration, no musicality, without taking or adding to the source signal. Most amps, even the cheaper ones, can do that (unless stressed). All the musicality and emotion should come from the musicians playing their tunes.

I suspect this viewpoint might look radical and maybe even crude to you ;-).
....

Hello TheBatsEar,

I do not feel your points radical and/or crude, since in my project I also tentatively used the budget side YAMAHA A-S301 as you may find in my post #249 through #291, and even A-S301 was proved to be quite nice in direct driving Be-tweeters plus super-tweeters without LC-crossover coils and capacitors.

The lack of XLR balanced input with A-S301, however, was the main drawback in my project since I like to use OKTO DAC8PRO's 8-channel balanced XLR analog outputs into XLR balanced input of my "integrated" amplifiers.

High quality XLR balanced input and nice VU meters, together with enough power reserve and excellent S/N as well as extremely low distortion characteristics with A-S3000 highly encouraged me selecting A-S3000 to dedicatedly directly drive woofers in my multi-driver system.

 

[dualazmak]

Stunning speakers indeed. I seriously hope the tiny porcelain plates on top aren't some family heirloom, they are basically one Tchaikovsky away from falling down and breaking. ;-)

Thank you for your kind notice on the tiny porcelain plates on the L&R NS-1000; they are not so "family heirloom" ones, but my wife really loves these tiny Wedgewood plates which I have purchased in London about 30 years ago. We usually keep them in safer cabinet, and I placed them on NS-1000 just for the photo sessions...

 

[phoenixdogfan]

I also purchased one license of DePhonica;
https://dephonica.com/
but so far I could not find enough time to test it, especially the I/O routing. I know DePhonica provides its own ASIO driver for input of any sound into it, so the input routing should be easier than the that for EKIO. I would highly appreciate if some of you using DePhonica may join in this thread and share the configurations including how we can allocate output channels of DePhonica (max. 8 channels) into single DAC, and also into multichannel DAC like DAC8PRO.

Now I am very happy to invite any post fit for the thread title of "Multi-Channel, Multi-Amplifier Audio System Using Software Crossover and Multichannel-DAC".

I downloaded full dephonica. It seems to be complete freeware.

 

[dualazmak]

Hello friends,

Let me just share with you interesting info on dspNexus 2/8 and dspNexus 2/16, 2-in 8-out, 2-in 16-out, new onebox DSP units; please visit here and here...

The introduction web site is;
https://danvillesignal.com/dspnexus-dsp-audio-processor

 

[gene_stl]

At AXPONA 2019 I paused for quite a while in their suite. They shared a suite with a guy who had a speaker kit using their crossover and a GR servo sub in an open baffle enclosure. They were playing some music which sneaked in a very large organ pipe. Both sets of product were very impressive.
Danville and the speaker guy. ( I have forgotten his company name but have his card in the Axpona pile.) The Danville proprietor was an electrical engineer. Very impressive guy.

I don't remember how much his crossovers were. They were a little pricier, than I was interested in, but good to know about, for reference.

 

[dualazmak]

Hello friends,

I assume it would be somewhat worthwhile sharing my most update playlist of 65 tracks I currently use for check and fine-tuning of my system;

Please note that in my multichannel multi-driver multi-amplifier system and configurations, I send all the signals in 192 kHz 24 bit stereo using JRiver MC's "DSP output format" (for on-the-fly up-sampling or down-sampling ) into software crossover EKIO.

 

[mikessi]

There is really no audible benefit to playback beyond 24/96 sampling, especially with any recordings other that those done with the most advanced high res gear and high fidelity values. There's potential for lowering of measurable distortion, however, tho audibility of this may be another matter. This comment is probably going to be seen as opening a can of worms, but the standard assumption of more bits and higher sampling is always better isn't borne out by any serious studies I know of. Just saying.

Here is one of the many articles I've read on the matter: http://archimago.blogspot.com/2016/06/musings-analysis-is-there-any-value-to.html?m=1

 

[dualazmak]

Hello mikessi,

Thank you for your nice post again.

As shared in my very first post #1, I am still keeping my single DAC (OPPO SONICA DAC)+single amplifier+LC network system configuration for bit-perfect DSD native stream playback up to DSD256 (4x), i.e. 11.2 MHz 1 bit DSD format. DSD512(8x), i.e. DSD 22.6 MHz 1bit, and up to PCM 768 kHz 32 bit.

And just before starting my multichannel multi-driver multi-amplifier project, I very carefully considered and investigated (on various papers, books and web sites) about the pros and cons of diving into all 192 kHz 24 bit processing with complete elimination of LC-network (inductors and capacitors) by dedicated and direct wiring from amps to SP drivers.

Even at present, I sometimes compare the sound of current multichannel system (192 kHz 24 bit) with the still-keeping DSD native playback system using the DSD files.

Very fortunately, as exactly I expected, the pros of complete elimination of LC-network considerably exceeded the native bit perfect DSD playback with LC-network in all aspects of sound quality including 3D-perspective, phase accuracy, transparency, cleanliness, delay features, etc.

I fully agree with you, therefore, at least in my project, "192 kHz 24 bit processing all the way" is just good (or perfect) enough in achieving really stunning sound quality in my multichannel multi-driver multi-amplifier system.

Also I am very happy that I can fully trust JRiver MC's nice on-the-fly DSP up-sampling/down-sampling functionality into 192 kHz 24 bit for all kinds and all sampling rates of music and video formats.

 

[Descartes]

So much equipment! I wonder what the room acoustics are like?

I am looking forward to getting an OKTO DAC8PRO!