pc based audio

[Jim R.]

Another option, which I haven't heard myself, but is still somewhat in development, and based on a linux single board computer is a:

http://www.auraliti.com

I am building a similar box, but using a single board alix computer running MPD, and connected to a NAS running NFS.  This and another one similar to it will be used for my living room and headphone rigs.

[glynnw]

Well, I put my foot in it and asked over on the Asylum Computer Audio Forum if anyone had done a PC vs Mac comparison and if not, would anyone want to.  YEEOW!  All I got were comments that it was pointless for a myriad of reasons.  I can't help but think that the main negative reason would be that it would end many of their daily posting arguments if someone were to produce a real comparison with valid results saying one format sounded better.  Every time I post elsewhere I am reminded of what a great group of people we have here at Bottlehead.

[ironbut]

For the most part, I've given up on most of the online computer audio forums.
That's not to say that they aren't very helpful in getting a computer rig put together from square one or finding out the latest stuff that you might want to try.
But when it comes down to refining the sound that comes out of it, I can't say I've ever seen so much confusion!

[denti alligator]

Got the Juli@ and installed easily. However, it doesn't seem to recognize samples rates over 44.1. I can play 24/96 files in foobar or J.River Music, but they only appear as 44.1 in Juli@. When I force the sample rate to 96k they sound all sped up and wrong.

Any clues as to what I've done wrong?

[John Roman]

Have you checked their website www.ESI-audio.com for tech support?
Also check www.computeraudiophile.com.
Let me know what you figure out.
John

[denti alligator]

Hm, well, I recovered from a crash due to fooling with the settings and now it seems to be fine.

Oh, how glorious this stuff will sound with a Bottlehead DAC and my SEX!

[Doc B.]

We are working on some aspects of the DAC circuit today.

[dstrimbu]

Hi everybody,

I'm fairly new to the BH family - I just built a Crack over Xmas break - but I have some ideas here.  I'm an engineer, and I've spent 22 years with Autodesk, Inc. in various roles.  I'm primarily a music lover, and I'm fascinated with the concept of music reproduction.  I recently (two days ago) changed "The #1 Wish in my Life" from driving a factory Porsche GT3 at the Nuerburgring (which was never going to happen) to getting my hands on a production BH DAC... pretty messed up, eh?

Regardless of my weak qualifications, I have spent a couple of years building several pc-audio systems, and I think that I've learned a lot.  I know nothing about vacuum tube circuit design, but I think I understand PCs.  So, here's a synopsis:

1.) Analog audio out of a PC is... bad. You can imagine all of the schytt that's going on inside of that box. PCs are Ghz-frequency digital devices, and - amazingly enough - the noise/RFI generated inside of the PC is of little consequence to the digital circuitry inside.  It's either a zero, or a one... and it's fairly easy to make that determination in the digital domain.  In the analog world, however - all that RFI is just radiating, looking for an "antenna".  Your analog audio wiring inside the box looks a lot like an antenna, BTW.  I know that you're all aware of the limitations of analog audio from a PC-source, as you are all Bottleheads... so enough on that subject.

2.) USB was NOT designed to carry music.  I am not an EE, I was trained as a ME... however, with 20+ years in the industry, I understand the difference between data and music.  Data is, just that - get my zero'es and one's resolved, and I'm fine - I will have data integrity.  On the other hand, digitized music is a slave to the 4th dimension - time - and, if you don't pay attention to that requirement, even your digital streams are going to sound like they've originated from a PC.  This timing issue can be ameliorated, but it takes either a.) a schytt-load of cash or b.) some ingenious software.  More on that to come...

3.) Harry Nyquist (and Claude Shannon) was (were) wrong - at least as far as music is concerned.  Their sampling theorem might hold true for data - in that case, the concept of the data being "completely determined" by a [max_freq]*2 sampling rate is valid and has been empirically proven.  But shuffling bits from one place to another has nothing to do with music, and that's why my loving-yet-tone-deaf wife can easily tell the difference between a CD and a high-rez (24/96 or higher) bit-stream, without being prompted...

So, as any engineer would approach this.... "how to fix?"  Indeed, it's not as hard as they (the audio press) would want you to believe. I have nothing but respect for the editors of Stereophile (I've been subscribing for.... 15+ years?) and The Abso!lute Sound...but the true solution to the problem is not based on throwing money at it.  It's about being smart, and working to understand the issues at hand.

[BTW, my friends:  it's Friday night, I've had a coupla brews, and I'm listening to Janet Jackson's "Rhythm Nation 1841" via a FLAC stream from my HP workstation... bitperfect (via ASIO) digital out to a marvel of price/performance:  the Pro-Ject USB Box.  Analogue voltage swings exit the Pro-ject via AudioQuest Black Mamba II cables (BTW, the concept of 'cables' will be the subject of another rant, RSN) to my beloved, hand-forged Crack.  :-)  The modulated audio signal causes current to cascade across that workhorse 6080 (and it's such a pretty tube!), inducing voltage-swings that are transmitted via the (marginal) stock cable, into the drivers of my 1,000+ hour Sennheiser HD-600s... did I say that 'Life is Good'? I'm 52 years old, and I'm chair-dancing to Michael Jackson's sister's worst album... auch du lieber, Augustine!]

But I digress.  Doc, I promise to get my 'script refilled tomorrow... I promise.

I'll be really honest:  if $$$ were unlimited, I'd buy all my music in highrez from HDTracks.com - OMG, I've always loved Alison Krauss, but "A Hundred Miles or More - A Collection" will stop your heart... and no, Doc - I'm not talking about two hands in the B+ power supply.  I'm talking about plain 'ol LOVE.  The girl is an angel...

And then... I'd listen to those 48k/24-bit (or higher) bitstreams via a dCS Puccini U-Clock, because it's the most incredibly-stable DAC out there... [Doc, Paul... I know that you guys can do better...  ;-) ]   But, I have two kids in college and a wife.  So, no U-Clock, and... what the heck to do with these 800+ Nyquist-theorem-crippled 44.1khz/16-bit CDs?

I have three systems in the house.  Two are purely digital in terms of input; both are driven by PCs.   Both sound amazing, IHMO.   I promise to tell you guys more about what I did, and how I did it, if you're interested.  Let me know, OK?

And Doc... how's that DAC coming?  I'm on the order list as soon as you announce...  :-)

Thanks for putting up with me,

-don

[John Roman]

Sounds like you'll fit right in around here. But chair dancing to Janet Jackson, now that is cause for concern. You steadied the ship with Alison Krauss though.  BH is a true value in this all to often over priced audiophile endeavor. None the less I'm along for the ride and BH has yet to disappoint. I'm also interested in a pc based system and am currently looking at building a music server. I've rolled it over so many times I'm gettin nauseous. Time to bust a move on it soon but for now,
Ich mus eine Heineken trinken
John

[Jim R.]

Hi don,

I don't mean to sound snarky, but there are lots of myths, half-truths, misinformation, and even some dysinformation concerning computer-based digital audio  out there.  Also, this is in no way meant to criticize bottlehead as hardly anybody outside BH has heard their dac so far.  That said, a few things...

usb, when ddone right (and that does not mean async necessarily) is perfectly suitable to superb digitial playback.  Secondly, there are a huge variety of asio drivers, and none of them can gurantee bit-perfect playback as there is much much too much going on inside the PC, even when carefully stripped downwith nLite) to guarantee bit-perfect playback.  I also don't agree that it takes tons of money to get superb playback, but software can be a very important part.  After loads and loads of experiments by myself an friends, it is clear that a windows based transport is in third place witha single-board linux box in second place and a tweaked Mac Mini, clearly ahead of everything out there atthis time.

Also, spdif is full of compromises and is far more difficult to get a quiet, clean data stream out of.  How do I know this?  Well, like you, I am 52, an EE/CE and a 40 year audiophile, but all that aside, it is because I am listening to a usb-only, 16/44.1 only dac connected to a modified Mac Mini as transport, and for the firstime in my life I am hearing the real promise of digital in a completely, grain-free, glare-free, dynamic, extended, detailed,and completely enchanting way.  The dac in question is less than $2k, is a NOS chip, has no async usb interface, nojitter reclocking circuitry, no tubes -- just a power indicator light, an iec power inlet, a usb B data connector, and two rca outs, and cost less than $k (currently, but going up at the end of the month.)

There is absolutely nofatigue at all with this dac, and I have listened (to plain-jane 16/44.1 cds for 10 hours straight and could have kept going.

This dac asympotitically approaches vinyl, but has overall dynamics, and maye even more importantly, dynamic contrass to die for and that you'd have to spend an executive's yearly salary on an analog setup to even start to come close to.

I've heard many many dacs of all kinds, with all kinds of transports, and at all price level imagineable, and to my mind, this is the one to beat.

I too am looking forward to the BH dac, but it's going to take some amazing engineering to outclass this one I'm talking about.

This is the dac that shattered all the myths about usb and the limitations of 16/44.1 and affordability.  It really is the game changer to pay attention to if you are considering and/or designing a new dac.

-- Jim


P.S. -- welcome!

[denti alligator]

and that DAC is ... the HRT MusicStreamer II+ ? That's less than the cost of the Seduction.

Listen, I understand next to nothing of the technical talk. This is not my field. But I care deeply about getting the best out of my music, and feel that I can hear the differences.

With digital it gets tricky. I have a lot of FLAC that I've downloaded and a lot of my own collection ripped to FLAC. Loaded in J.Rivers Music Console these files are easy to access and sound relatively good. I'm running Vista and have a Juli@ installed. It is doing the dac, so that I'm outputting analog. Now I know it would be ideal to get an external DAC, but I don't have the funds yet.

I listen to my CDs and SACDs on a Denon 2930CI. I haven't yet compared the two, but I suspect the Denon sounds better. Once I do the dac externally, however, they should sound identical, right? The catch is SACD, which I don't think can be output digitally (on this machine). So I let the Denon do the dac, but then I'm short-changing myself, or?

(Addendum: actually the Denon does output SACD digitally through the Denon Link, only compatible with Denon receivers. It would be great if this signal could somehow be converted into one recognizable by a standard DAC.)

[Jim R.]

Well, I didn't really want to mention it on the BH forum, but since somebody asked, I'll mention it this once...

it's the db audio labs Tranquility SE.

I also am generally really rserved when it comes to talking about audio products as I know there are such huge variables in our systems, rooms, tastes, and budgets, but I'm definitely not a wealthy person, but music and audio are my number one priority after my wife and our critters, and as a blind person,it is probably more important to me than most as it is the most powerful way I connect with the emotional/sensory world.  I'm also not easily impressed.

-- Jim

[dstrimbu]

Hi Sam - I don't think that anyone has come up with a consumer-friendly way to get the digital bits off a SACD.  Unlike CD, the SACD data is stored on disc, and delivered to the player's digital processing chain, as an 1-bit Direct Stream Digital bitstream at 2.822 Mhz.  There are several pretty heady physical copy protection features that prevent this data from being ripped, including an encryption key that's located on the disk in a physically inaccessable area - except to a licensed SACD device.

And even if you could get the bits off the disk, conventional DACs can't decode the encrypted, delta-sigma modulated data.

So, I would say: don't stress on which DAC might be better; we have an older Denon DVD/SACD player (DVD-2200) in our surround system downstairs; it sends multichannel SACD audio out to a Sherbourne pre-pro via six unbalanced analog (RCA) outputs (5.1 channel), and it sounds fantastic.  I know that SACD sounds wonderful - my desire is to get as close to SACD sound as I can with the 800+ CD discs that are hanging around the house.

I would say - enjoy those SACDs as played by the Denon; I know that mine sounds really great - and as you've said, probably better than your music server.

Thoughts?

-don

[dstrimbu]

Hi don,

I don't mean to sound snarky, but there are lots of myths, half-truths, misinformation, and even some dysinformation concerning computer-based digital audio  out there.  Also, this is in no way meant to criticize bottlehead as hardly anybody outside BH has heard their dac so far.  That said, a few things...

usb, when ddone right (and that does not mean async necessarily) is perfectly suitable to superb digitial playback.  Secondly, there are a huge variety of asio drivers, and none of them can gurantee bit-perfect playback as there is much much too much going on inside the PC, even when carefully stripped downwith nLite) to guarantee bit-perfect playback.  I also don't agree that it takes tons of money to get superb playback, but software can be a very important part.  After loads and loads of experiments by myself an friends, it is clear that a windows based transport is in third place witha single-board linux box in second place and a tweaked Mac Mini, clearly ahead of everything out there atthis time.

Hi Jim,

I don't think that you have a snarky bone in your body, sir.  I realize now, after re-reading my email from last night, that I could have been a bit overenthusiastic (read: inebriated) regarding friend's hardware that I have heard, and the virtues of high resolution audio files.  Thank you for your patience with me, I'm new here.  And thanks for the welcome, too.

Regarding USB for audio, I actually downloaded and read a fair amount of the USB Device Class Definition for Audio Devices today (yeah, I'm a bit weird).  Also interesting that Release 1 of that doc was dated 18 March 1998... not nearly as old as the CD-DA standard, for sure.  And you are 110% correct - USB _was_ designed with high quality audio in mind.  From the Management Overview section, pp17:

An essential issue in audio is synchronization of the data streams. Indeed, the smallest artifacts are easily
detected by the human ear. Therefore, a robust synchronization scheme on isochronous transfers has been
developed and incorporated in the USB Specification. The Audio Device Class definition adheres to this
synchronization scheme to transport audio data reliably over the bus.

So what I'm reading here - at the risk of oversimplification - is that the USB standard supports high-quality audio, but very few have built software that takes advantage of this capability.

Hmmmmm.  Sounds like an opportunity.  In the meantime, those of us who don't have the right chops will be stuck with S/PDIF, I think.  I totally welcome a USB audio implementation that sounds correct.  If someone could build it at the "right" price-performance point, they'd be overwhelmed with orders.

That being said, Jim - I still think that the 16-bit / 44.1khz standard for CD-DA is flawed - but then, the Red Book was first published in 1980.  I guess that my real issue (and I know that you got this) is what I said in italics near the bottom of my rambling:  "...what the heck to do with these 800+ Nyquist-theorem-crippled 44.1khz/16-bit CDs?"  I think that the solution looks like this:

1.) bitperfect FLAC rips of CDs (with a tool like dbpoweramp)
2.) bitperfect output from the pc to the DAC - (USB, S/PDIF... don't care - as long as it's correct)
3.) time-domain coherency in the reconstruction of the audio waveform and
4.) enough amplifier power to drive the transducer of choice

I think we got #1 and #4 nailed.  Still waiting on #2 & #3. 

I've been applying "gearhead techniques" to this problem here at our Casa - mainly because I'm a gearhead, and that's all that I know.  I'll have to rely on folks like Doc, Paul J, Paul B, Grainger and others to come up with an elegant solution to the task of converting 1s & 0s to smooth, spacious, grain-free sine waves at a price point that I can afford.  And, if it's a kit that I can actually build, it'll be awesome.

Thanks again, Jim.  Great to meet you!

-Don

[Paul Joppa]

...I'll have to rely on folks like Doc, Paul J, Paul B, Grainger and others to come up with an elegant solution ...

Don't leave out John Swenson, who has been spearheading the digital portion of the DAC we're working on. I've played with digital acoustic data for more than 30 years professionally, but I turned down that job - John is way more qualified, knowledgeable, and talented. I'll just work on the tube/audio parts.