Getting the music out of the "box"

[rickcr42]

More than one way to skin THIS cat so what is YOUR personal preference ?

1-Internal Sound Card
2-Using a PCI Card extender to get the soundcard "out of the box" and use a local power supply to islate the card from computer power supply generated noise
3-Galvanic Isolation to an external DAC (again to isolate the computer power supply and computer ground noise)-Playback Only
4-Externbal USB CODEC (one of the TI PCM29XXX chip based solutions)
5-External USB DAC-Play only
6-External USB "Ramp" to an external PCM DAC
7-"Insert your method here"

On the output side (music playback) I am currently running a USB Ramp with a galvanically isolated SPDIF output into a PCM DAC with "the future" of this hookup being either a reclocker located between the two (Still a SPDIF stream),a reclocker with an I2S output to the PCM DAC or modifying the PCM DAC itself.

On the input (record) side I build a little "box" that has both microphone and buffered line inputs and a line matching transformer output which in turn goes directly to the sound card line input.not a perfect solution but A solution until i can get my DIY ADC project working to my satisfaction

[charleyphogg]

Rick,

What do you mean by Galvanic isolation? Sounds like Optical only, but I think you may mean true 75ohm. If so whats the best, simplest;est and last but not least, cheapest way for a halfwit to accomplish it?

I've read your stuff on it before, but few minutes of thinking before approaching nuclear melt down gets used up rather quickly.

[Spirit]

Not having much clue what half of your listed options are, I'm going to hang my head a little and say #1 =P.

Auzentech X-Meridian/LM4562 with the dc filtering caps bypassed.
This directly feeds a pair of HD600s (or will do again when I get a new cable)
(Yes, I will be getting a kit Novo as soon as they're available).
Also drives a MF X-150 (slightly modded up too) to a pair of Dali Ikon 7s.

The Well Tempered Computer offers a fantastic overview of pc audio .

I'll see how it sounds when I get the Novo, but the more I read, the more it seems I should be using an external dac. X-Meridian uses a great dac chip (AK4396), but obviously being inside a pc, conditions are less than ideal, and I'm doubtful too much thought has gone into the implementation.

[rickcr42]

Rick,What do you mean by Galvanic isolation?

the techy version :

en.wikipedia.org/wiki/Galvanic_isolation

in short you are providing both line matching AND a ground breaker there being no actual connection between the transformer primary and transformer secondary so while you are still passing a high bandwidth signal (mHz range-radio !) if done right you will reject much of the out of band RFI.

More effective on the "computer generated noise rejection" front would be to have another transformer in the audio signal path.This one being designed to handle audio rather than radio frequencies means any out of band signal will be rejected and since again we have no actual input/output electrical connection even digital-ground-noise will be eliminated while the audio signal will pass through untouched.

It is my personal belief come to emprically that ALL digitl sources should have such isolation at some point in the signal path with the preferred/most effective being a simple 1:1 audio line transformer between the sound card (or CDP or DVD Player or DAC analog out) and the preamp if one exists or amp if one does not.

Once you hear "digital black",that is the elimination of all the stray noise inherent on the digital ground bus and that is most noticeable between music selections or during low level music passages,you will never look back.

[rickcr42]

I'll see how it sounds when I get the Novo, but the more I read, the more it seems I should be using an external dac. X-Meridian uses a great dac chip (AK4396), but obviously being inside a pc, conditions are less than ideal, and I'm doubtful too much thought has gone into the implementation.

that is not a "given" since I personally have heard some very good internal sound card based systems and some terrible versions of all the other choices so it comes down to implementing the idea more than choice in topology.

having said that my personal best is so far is with everythything "out of the box" and with local linear power supplies though the USB sections does use bus power (decoupled and regulated down to 3.3VDC).

As originally stated there are many ways to get to the same place and all are valid if done right

[Deleted]

that is not a "given" since I personally have heard some very good internal sound card based systems and some terrible versions of all the other choices so it comes down to implementing the idea more than choice in topology.

having said that my personal best is so far is with everythything "out of the box" and with local linear power supplies though the USB sections does use bus power (decoupled and regulated down to 3.3VDC).

As originally stated there are many ways to get to the same place and all are valid if done right - Rick

Rick
I would go along with that. I have not tried your preferred option, as I am using the SC HA from the analgue output of my mildly modified X-FI soundcard for convenience.
I can say, that the results with the Microsoft Mixer bypassed, and using E.A.C. sourced .wav files from the HDD, can be stunning, and superior to that from a normal CD/DVD player and good DAC combination.
I suspect that the majority of affordable CD/DVD players have a long way to go with their optical read areas, and processing sections/clocking. This is evident, if you burn a copy of a good quality CD to a 24K Gold CD-R using .wav files as above. The clone usually sounds noticeably better than the original !
Alex

[Spirit]

Sept 30, 2008 22:33:41 GMT rickcr42 said:

that is not a "given"

I'm fairly sure it'll be good enough to stop me spending $300 on a dac . Mate has a keces of some kind that we'll do comparisons with when he gets his little dot back, and I get my Novo - can't wait.

Having said that, the engineers decided to run the outputs at more than 2V (louder is better, right?) and anything above about 60% on the software volume control and it clips.
Makes me wonder...

[Will]

I had a look at 'PC as source' after being dissatisfied with using my works laptop as a glorified mp3 player.
The line out from the laptop (Dell D610) is hideous, with huge amounts of noise. To give a better quality signal to my amp, I bought the Alien DAC as a kit. I know it's not the greatest thing but I've been pleasantly surprised at the end result, especially for £35, once I'd built it. This is of course a USB DAC, based on TI PCM2702 chip. For the money, it's certainly good enough for the office. It also introduced me to how scary SMD soldering

It really improved in SQ after installing ASIO4ALL and using Foobar as the music player. All music files are FLAC. Always bypass the windows mixer.

For the future, and especially after reading Ricks posts on computer based music systems in the past, I'm considering a PC especially for music. Time will tell.

Two DACs that interest me at the moment are the Buffalo DAC, based on the new ESS chipset www.twistedpearaudio.com/digital/buffalo.aspx, which I think Leo is getting, and also the y1 gamma dac, based on the Wolfson WM8501 (or WM8759) chipset. headwize.com/ubb/showpage.php?fnum=3&tid=7469 The y1 may start to suffer 'design by consensus' but may be an interesting mobile solution.

[rickcr42]

Two DACs that interest me at the moment are the Buffalo DAC, based on the new ESS chipset www.twistedpearaudio.com/digital/buffalo.aspx, which I think Leo is getting,

All the evidence suggests the ESS chipset is the real deal and considering it is damn near a self contained DAC would be mighty hard to screw up (though some will try in the name of "improving" it ) so should be a good choice for one considering a new DAC.

It has been a while since I visited the ESS data files but from memory I beleive I would just toss it in a box with a good power source and using the voltage output go direct to the out jacks via a 1:1 line transformer in BAL to SE mode keeping the essence of the design unchanged rather than try and reinvent the wheel

[rickcr42]

BTW-I should add that in my own case I have a max. 15K Ohms available on the inputs of my passive line controller (AKA Rick's Line Master ) which would require an active buffer after the line TX unless the ESS DAC is capable of driving a 600 ohm line directly in which case a 600 Ohm bridging TX would suffice by itself (600 ohm input/600 ohm drive).

It has been my personal experience that it is always best to use only what is needed and no more in the signal path for best sonic purity (read sonic butchering since ALL inline devices have an effect and that effect never better than if it were not there unless levels are too low or there is an impedance mismatch ) so if given a choice i choose a passive option if I can and active only if I must