How much can a computer do?

This is my current system. The computer is the center of it all.

Digital Crossover: 8 channel convolver (stereo 3way XO 64 bit resolution processing)

Players: Media Center, Media Player, FooBar, VLC, NeroHD (blu-ray)

Screen shot of it action. Notice that the digital crossover in in the computer.

First off, you could explain what all the gubbins in the diagrams does. Then you could tell us all what you think of it. That'd be cool.:rocker:

Ok,

First of all everything above the speakers and amplifiers is the computer.
1. The picture of the little speakers up in the top left corner represents the windows direct sound interface. This is the interface that all windows sound applications gain access to sound hardware.

2. The middle mixer strip (Playback) is the ASIO access to the sound hardware.

3. The top mixer strip (Hardware in) is the line-in interface.

4. The bottom mixer strip (Hardware out) is the actual hardware line-out interface.

5. These three mixer strips together are called TotalMix and are provided by the RME Hammerfall DSP drivers for the Multiface II professional audio interface. This Interface uses top quality AD/DA converters.

6. The Art Teknika Console in the upper right corner is a standalone modular VST host that allows you to draw lines between inputs and outputs of devices to connect them the way you want.

7. The “Waves Linear Phase EQ”'s are VST plug-ins that are being hosted by console.

This is how it works:

a. 5.1 Sound comes in via the first 6 channels of the middle mixer strip (playback)

b. The first 6 channels are then routed to output Hardware channels 11-16.

c. Output hardware channels are internally looped back to input hardware channels 11-16.

d. Input hardware channels 11-16 are sent to Console for EQ and Xover. This creates 9 output channels (left lo, left hi, right lo, right hi, sub derived from left/right, sub proper, center, left surround, right surround)

e. The 9 channels coming out of Console are received via the playback interface and are routed to the output hardware channels. derived sub and sub proper are combined at this point.

f. Hardware outputs do the D/A and feed the amplifiers.

What do I think? I think it’s totally cool.

ALL sound processing and routing is done in the digital domain with at least 40 bit resolution. The EQ processing is similar to the Dolby Lake processor.

The output DAC’s are top of the line.

How does it sound? In a word amazing!

My system has never sounded better. I will be selling my heavily modified Behringer DCX24/96 XO because this system sounds much better and is no longer needed.

EDIT: I almost forgot. TotalMix is MIDI controllable so you can us an external control surface to manage all the faders. I don't have one so I wrote a small utility that watches the Windows Master volume and translates that to MIDI and sends it to TotalMix. That utility is freely downloadable at the RME forum.

There is more to come...

Soon I will be using Audiolense to capture sound measurements and combine these measurements with crossover curves to create a convolvable impulse that can then be convolved in real time via ConvolverVST (hosted by Console). This will be doing Crossover, time alignment and room correction all at the same time.

http://www.juicehifi.com/index.html

Totally awesome. This kind of power can not be achieved outside the computer for less than $5k.

I will be posting my complete system with pictures soon. In the mean time here is a picture showing what I use to control everything.

Scott, you're probably blowing the minds of all non-techies...

I'll make a few points, for the moment.

1) My PC makes a noise. So does my laptop. Even though I use my laptop a lot to feed my Tri-Vista DAC's Toslink input via an M-Audio USB converter, I find the noise annoying. I should find a better, more quiet solution. Have you bothered to try and keep the PC/laptop noise down? Even with an expensive PSU, and fans. sinks etc on my quad core I can still hear my PC in my lounge.

2) Your totally right - PC hardware costs a lot less with impressive DACS etc. A while ago, either Asus or ABit released a mobo with a tube output stage. Must have a look and see if they are still available before my Tri-Vista breaks down totally. Probably not, I expect.

3) When I have had a few beers, I don't want to wait for my laptop to boot up, fire up and manage complex s/w etc.

4) Lots of audio processing never seems to give great results. Depends on the quality of the programming, though. I haven't experimented much in the hi-fi realm, but there is a lot of guitar processing s/w out there that gives truly bad results. Also, SS guitar amps that try to model valve sound and different amp types/cabinets. They mever manage to pulll it off, though some are better than others.

Anyway - I look forward to the pics.

Here is a closeup of the 8ch bass mgmt, 8ch convolver and 1/6 octave EQ

User211 said:

Scott, you're probably blowing the minds of all non-techies...

Click to expand...

sorry. just thought some might enjoy seeing just what is possible these days.

1) My PC makes a noise. So does my laptop. Even though I use my laptop a lot to feed my Tri-Vista DAC's Toslink input via an M-Audio USB converter, I find the noise annoying. I should find a better, more quiet solution. Have you bothered to try and keep the PC/laptop noise down? Even with an expensive PSU, and fans. sinks etc on my quad core I can still hear my PC in my lounge.

Click to expand...

My PC is verry quiet. All fans are rated less than 20 DB.

2) Your totally right - PC hardware costs a lot less with impressive DACS etc. A while ago, either Asus or ABit released a mobo with a tube output stage. Must have a look and see if they are still available before my Tri-Vista breaks down totally. Probably not, I expect.

Click to expand...

Motherboards do a good job hosting CPU's but I haven't heard sound from one that I find anywhere near acceptable. The RME OTOH has externally hosted DACs and jitter is less than two nanoseconds. It just sounds pristine.

3) When I have had a few beers, I don't want to wait for my laptop to boot up, fire up and manage complex s/w etc.

Click to expand...

My system is on all the time. Even when it does fire up everything is set to go. When it does fire up it starts Media Center so my wive and kids can drive it. I usually exit Media Center and use FooBar 2000 to send CD and DTS to ASIO out.

4) Lots of audio processing never seems to give great results. Depends on the quality of the programming, though. I haven't experimented much in the hi-fi realm, but there is a lot of guitar processing s/w out there that gives truly bad results. Also, SS guitar amps that try to model valve sound and different amp types/cabinets. They mever manage to pulll it off, though some are better than others.

Click to expand...

Waves makes some of the best processing in the industry. These plugins are used on computer based recording rigs like Pro Tools. They are very transparent, have close to zero phase distortion and signal to noise that is better than 140db.

Anyway - I look forward to the pics.

Click to expand...

As soon as I get everything back together and cleaned up...

Here's the interface the wife and kids see:

I remember that tube mobo got some good reviews in hi-fi mags. It wasn't a lemon.

What X-over freq are you using? I presume you have messed around with it a bit? Did you just end up setting it to pretty much the same freq. as ML rate the panel to?

Hey Scott,

I like it.... not necessarily a traditional straight wire with gain approach but impressive.

And the file format is.... ?

:rocker: Respect :rocker:

Psychoacoustic

:Summits, Perreaux 750 Mono's, PS Audio Dac, Krell pre, JL Sub & some etc. etc. etc.

I run a home theater/home audio HTPC as well, but I'm not doing any crazy time alignment or EQing in the software or anything.

Mine is a little Mac Mini, they are really great. About the size of 4-5 CD's stacked on each other, and the thing is completely quiet (as in cannot hear...does not make noise)

You can run Windows or Linux on it, but I run MacOSX and just use the free built in Front Row software to control it all. It gives my wife and kid an easy to use front end to play all the music, movies, tv shows, etc. My only gripe with it (actually a gripe at my TV) is I can't seem to get the overscan settings correct.

Scott,

VERY cool. If you could package that......

Hi Scott,

I've always been of the "simpler is better" school regarding sound reproduction in a home. Then again, based on your picture, I could be your father if age was the sole criteria.

The underlying rational, as I'm sure you know, is the less processing between the original signal, and the signal that you hear from the transducer, minimizes any electronically induced artifacts within the overall signal path, thus providing a "cleaner" listening window to the music.

So, with all due respect, and on the assumption that you've tried the "purist" route and prefer the "processed" route, why is the latter superior?

Best regards,

GG

PS: Congrats for using your full name on this website. I think you and I are the only two individuals that do that.

No slam intended for the others. I understand the underlying rational.

Actually I consider myself a purest. Let me explain:


In my original system pictures below the signal went something like this:

CD(digital music)->Cable->Yammy(->(+processing maybe)DAC [or ADC then processing then DAC]->Pre amp (switching and volume)->amp)-wire->passive xover->speakers.

Long story short here I am now:

digital music->digital xover->RME DAC->wire->amps->wire->speakers.

or another way to see it:

(computer)-DAC->wire->amps->wire->speakers

The signal actually goes through less 'processing' because everything is done in the digital realm with out the use of cables. Processing is essentially lossless since it being done at, at least 40 bits (64 in the xovers which is almost 3 times the resolution of most standalone XO's). Jitter is also kept to an absolute minimum, the only piece in the whole system that has any effect on it is the RME and it has been tested at 2ns. The real potential for problems arises in the algorithms that are being used but since they are in software they can easily be changed. And since I have access to the latest stuff (I know some pro audio people) I can change out the processing until I find the right 'fit'. As seen way below some of the best XO's are free.

Since I removed the digital XO that has limiters in the output stage I will be trying a light bulb in the chain in an effort to protect the panels (in case of computer crash or some other accident: http://www.eden-electronics.com/info...weetertext.htm

(computer)-DAC->wire->amps->wire->bulb->wire->speakers

I have serious doubts about the effect it may have on sound quality however and it may not be there very long.

If my source were vinyl it would be a completely different story but since it starts as digital I want to keep it that way. I also want to use the most accurate processing available until the last possible point in the signal chain.

This picture gives some idea of the stuff that is no longer being used.

Now for some pictures

First my old system and play room(s):

The main speakers are Vandersteen.
Vandersteen Center (behing speaker grill above TV)
The 4 ceiling mounted speakers are Sound Dynamics RTS-3's.
Components are sunken into the wall. They're easier to see in the band shot just left of center. Access was via the studio (hidden behind a small door next to, and incorporated into, the closet) it consisted of (from top down):

Counterpoint SA100
Onkyo t9090
Yamaha DSP-A1
Sony DVP-S7000
Sony Top-of-line VCR (I don't remember the number)
(some dvds)
2 Sony 200 disk carousels (I can remember those numbers. One was master and the other slave and I controlled both via an slink-e)
Original Pronto universal remote (custom programmed) controlled everything + lights.

Phillips 55" TV sunk in back wall.

Here is where I am today (as you can see we've downsized quite a bit since then, at least the TV got bigger):

Mains - Martin Logan Aerius I (passive crossovers bypassed)
Center - Martin Logan Cinema
Surround - Sound Dynamics RTS-3

RME Multiface II (DAC)
Monster Reference HTS-5000 power.
Sony DVP-S7000
Custom built computer with (bla, bla, bla)
Sunfire TGA 5200
B&K ST2120 (modified: silver internal wiring)

Soon to be sold (maybe):
Behringer SRC2496 (AD/DA and format converter)
Behringer DCX2496 digital xover (heavily modified)

Mitsubishi WD-73734 73" 1090p DLP.

<small>(Humm... dusty chess board... wow the flash really brings it out I guess I need to clean-up. Just shows how much I don't play any more)</small>

Hopefully I can get Tom to move this whole thread to member systems.

Tom?

User211 said:

I remember that tube mobo got some good reviews in hi-fi mags. It wasn't a lemon.

What X-over freq are you using? I presume you have messed around with it a bit? Did you just end up setting it to pretty much the same freq. as ML rate the panel to?

Click to expand...

I have tried everything from about 200 to 700. The passive is at 400 (I think) but I prefer about 250 to 350. The sound seems more open at that point. As the XO point is raised the sound gets fuller but the cone becomes more noticeable and starts to get a tiny bit honky with a slight nasal like quality (to me anyway). I think 400 was probably chosen by ML because it is a good trade off between open vs full. Open is better in my room probably due to acoustics and I will likely change it when the room changes.

All the following images show internally how the XO is performing at 258hz (the top lines in each image is peak hold, left=low / right=high).

This image shows pink

This one white

This one shows sine log sweep from 20 to 20k

This one shows sine sweep at 258

notice that exactly at the crossover point there is almost no phase deviation. That's what I call linear.