There is a lot of discussion about Computer/DAC setups and Computer/DAC versus Transport/DAC as separates or as CD players, including sidelines in threads on this forum. I find it a compelling discussion, and since we have the ZDAC, the choice is there for Decware users. It seems like a subject that could use its own thread, so I am starting this one. In fact the following is basically a fleshed out response to a side discussion in the Zen Ultra thread of the Decware News
http://www.decware.com/cgi-bin/yabb22/YaBB.pl?num=1301477558/58#58
….please join in with your experiences and opinions.

Here is my simple (haha) take on the question of relevance of computer/DAC versus transport/DAC.




No matter what circuits or transport mechanism we use to decode and convey CD data, in the beginning, it is just that....data. Our most common choices for turning the data to music with the least compromise seem to be CD players (combined transport and DAC in one box), separate transports and DACS (offering mix and match potential that CD players don't offer), and Computer/DAC combinations. Each individual or combined setup is all about converting the data and making it sound good according to the tastes of the designers and users.

As far as I have been able to tell, the main technical differences between properly done error corrected, uncompressed Computer data taken from a CD, versus CD data through a Transport are as follows. All transport/DACs bring with them the issues are error correction and jitter, issues inherent to transports and the CDs themselves. So the technical difficulty for transports is extracting and conveying the data with as little error and jitter as possible. Then the DAC stuff happens, trying to correct or mask inherent errors and jitter, while converting the data into listenable analog music. Then comes the output section with its technical and sound choices. Choosing circuits and circuit paths brings a designers preferred tonal attributes to the stuff that shows up at the output jacks.

Properly done, with a good computer, error correction software (for extracting and correcting the CD data), and an optimized setup for transparent computer to DAC music serving, the issues of errors and jitter are virtually eliminated. This solves a bunch of the challenge transports bring to the picture. In this we theoretically have as good a beginning as is presently possible to get the CD data ready for the data-to-music conversion. As far as I know there is no technical question here. I can't conceive of any problems with the efficacy of eliminating (for the most part) the errors and jitter, thus simplifying the work of the DAC, and at the same time simplifying the circuit path.

Whether we get good sound or not is dependent on the proper data, equipment design, optimized systems components, system setup, and finally synergy. These are where the technology blends with tastes, both for design and listening.

To make this happen requires allowing our technological biases to shift to an open mind in order to deal elegantly with the details and issues of turning music into data, and then data to music. As with the transition from LPs to CDs, we are a recalcitrant lot, and once we find something we like, we tend to believe it is the end-all, so the digital transition is still not as refined as it should be. A big part of the problem was that we had pretty good sound, especially in the most refined systems with great tables. And most of the rest of the problem was that gear was designed so that tape and LPs sounded at their best.

So when the 44k CD data hit the scene, poorly derived from tape, and played on gear designed for tape and LP, it sounded awful. All the gear from mics to mixers, compressors, monitors, reverbs, and the rest in the studio, as well as what we listened with at home, was made to deal with the triumphs and deficiencies of the original mediums. Tape and LPs have native "warmth"  and compression, but also, they have a seemingly infinite "bit rate," a very particular set of criteria when recorded and reproduced.

CDs are mostly limited to 44K, not enough bits for it to sound like music, and are made and played from a deficient and still developing knowledge base......bizarrely, we still argue over how to work well with 1s and 0s. And our original digital recordings are also put down in bits, usually more than 44k now, but also very, very clean with no inherent compression and relatively noiseless.... This is a very different thing than tape and LP, and requires very different approaches to make it come out like beautiful music. Our heritage of many decades of tape/LP based gear development does not make shifting to digital an invalid pursuit though....As a medium to efficiently capture and convey music as accurately as possible it has great potential.



So even today, after several decades of digital development, if your desire is to convert to a computer beginning, the general knowledge is surprisingly immature. There are still many questions around the bit data and its conveyance. I have heard the difference between "lossless" compression files and uncompressed files. I have heard the difference between different USB cables. Both of these, some would argue, are delusional perspectives. The word on the net tends to favor the Mac Mini over other computers for a server. Many believe that using an external drive helps the sound. There is debate over output quality….USB, coax, toslink, and SPDIF….Some prefer one error correction software over others……and many more things. The point is that though data is 1s and 0s, how we move that data around and extract and convert it to music is still in relatively young state of development.



But I have found that if I dig for information, and experiment with an open mind, there is plenty of information to put together a great sounding rig. And in my experience, like with analog, there is as much need for proper setup, gear choices and the resulting synergy with Computer/DAC setups as there are for Transport/DACs… And since our recent technological legacy is relatively entrenched in making Transport/DACs sound good, the computer setup needs its own considerations and implementations to realize its potential.

And finally, it is always the tastes of the DAC designers that gives us the quality and flavor of the analog out…..the caps and resistors used, chips and clocks, the choice of NOS chips and Upsampling chips (and which ones), outputs sections, and so on. In this, the refinements and synergy of how this is all dealt with is determined by DAC makers tastes. Then, how the preferences of designers and listeners come together puts it in our court.

To me, in the question of transports reading and conveying CD data versus correctly implemented computer extraction and conveyance, there is little doubt that computers have greater potential. With it the DAC gets the best possible data, and there is for the most part, one less link in the chain. This in turn opens the door for the DAC to be designed without having to deal with the longterm issues of Transport/DAC design...jitter and errors.

But it is relatively new stuff, whereas transport design, and how they work with DACs has been refined with much attention over several decades, so it is no surprise that this arrangement can produce amazing music. But thankfully, in recent years, Computer/DAC is getting a lot more attention, and the inherent advantages of this method and ways of optimizing these advantages for making music are slowly being exploited.

And it is not all about the first stage. With a computer there is much potential to go further toward personally refining a system to taste. For example, the ability to apply Digital Signal Processing to the data pre-DAC, a refining potential that can be as clean as the DSP applied to the masters of all those CDs we have, but we can "master" our whole system with it so to speak.



So I made the leap, partially because I think this is where the cutting edge design is going, but mostly, because I prefer my Mini/ZDAC1 sound and flexibility to my Rega Apollo (a well designed CD player that retails for just a couple hundred less than my Mini/ZDAC1). But I definitely resisted the transition. And, very importantly, I only began to truly appreciate it after I tuned the system to the new sound this brought.

Are there Transport/DACs in this price range that sound this good? There may be, but I don't understand how the transport section can be better than a computer at its main job....the job of getting the CD data to the DAC. What causes quality transports to be good is more likely the refinement of the designers choices in making the problems inherent to transports relatively indiscernible to the listener.

In the end, they are two different things though, and the issues the Computer can eliminate from the matrix, creates a different starting point. So applying a computer front end to a system that had a transport will sound different and perhaps colder and too clean.....but this does not mean the computer is the problem. It more likely means that the computer system, files and cabling are not optimally implemented, or that the system has not been tuned to the source, or a combination.

This ramble is not by any means to say that this is the whole story. I am no expert. It is just the way I have been thinking today as I write this.



Any thoughts?????

I'm going the 'puter route right now. I'm figuring that I can use my Ipod - Wadia for kick around the house days but my laptop for when I want to really concentrate into the music. I still haven't compleatly figured out how it will be done but I will get there somehow.
I'm open to suggestions also, what is the "right " way. I'm currently just using FLAC files played through Winamp out my USB into my ZDAC-1. Is it the best way? I doubt it. But so far it sounds pretty good to me.

Will, like you, I have a Mac rig, and here is how my office listening rig is set up:

1. Mac mini [2008 1.66gHz Intel Core Duo; 2gb RAM] dedicated to music playback running Fidelia, with iTunes as the back-end library manager.

2. WireWorld Ultraviolet USB out to Audiophileo Model 1 USB-S/PDIF bridge;

3. PS Audio Digital Link III DAC w/ Cullen Stage IV Mods

4. Decware ICs out to Taboo; Taboo to Audez'e LCD-2 cans.

At home, I have the Oppo BDP-95 for BluRay, DVD-A surround, and SACD playback, plus a server with the same files as at the office [another advantage of the server approach is the ability to mirror storage in physically disconnected locations].

I started out ripping ages ago, before I realized how destuctive lossy rips were, and then went back and cranked up to ALC, and then went back and re-ripped, as close to bit perfect as I could all discs into AIFF. For my DVD-Audio, I rip on my iMac running VMWare using DVD Audio Extractor [a really nice program, and worth the fee]. That creates WAV files. Those files, as well as hi-res WAV or FLAC files that I purchase from Linn, HD Tracks, Reference Recordings, etc., I convert to AIFF using Fidelia.

The Audiophileo has a bit perfect test mode, and so far, this setup is delivering that most of the time [I don't stare at the display, but every time I test, the chain delivers bit perfect].

This setup makes for some very sweet music. I have nothing against transports whatsoever, but I find that having 1+ terabytes of music, that a server is one way to keep me exploring it. I can set up playlists that cause me to re-doscover music that otherwise would be sitting in a jewel box. That, plus the convenience, are what drove me to the computer side.

Folks interested in server based setups should check out http://www.computeraudiophile.com . It is an excellent resource. The publisher, Chris, is very knowledgeable, and there are some great FAQs.

Greg

Will, you have a nice setup and workflow. So, are you playing out of iTunes? Looks to me like you are all set for some even higher-res stuff, and I suspect you would enjoy it. The HDTracks and iTrax sites are great sources. The new Stones hi -res releases are very revealing. The ZDAC can certainly handle it.

P.S. Nice USB cable you have there.

You should give a listen to Fidelia. I have Amarra, Pure Music, and Fidelia. Each has its advantages, but I prefer Fidelia overall, and it is easily the cheapest.

I've been using a computer as transport from the beginning of my obsession with this hobby. A computer can transport all forms of digital. Having one machine do it all, and being able to store all your digital media and cue it up quickly is just awesome.

It seems the crux of the argument for 'not all digital transports are the same' is to say that digital data is merely data while it is being stored on a disc or on a computer hard drive. However, as soon as you attempt playback we are dealing with not only the accurate conveyance of digital bits (this is the easy part), but the timing through which those bits arrive. Jitter is the measurement of time related distortion and is measured in very precise increments. So the name of the game is to provide digital bits to the DAC with zero errors, but also with respect to accurate clocking (low jitter).

I'm with the camp that advocates as little manipulation to that original digital data as possible. No oversampling, no compression. Absolutely agreed that hard drive space is cheap so there's no sense in using "lossless" compression. It's important to rip data from discs with few errors so that the ripped data mirrors the data on the disc as closely as possible.

My computer is Windows based and uses Foobar for playback. Different media players have different sonic characteristics, which is rather dismaying when trying to hold onto the belief that everything digital coming out of your computer is the same. Even within Foobar as you try different output ASIOs for kernal streaming (effectively bypassing all Windows OS manipulations) it results in different sound. This, even though supposedly the ASIO (Audio Stream Input/Output) is outputting bit perfect. I prefer the DirectSound 2.0 ASIO. To me it sounds absolutely neutral and clean.

The great thing about proper USB is that the digital data packets are transported bit perfect without the necessity for clocking. This is why I have a hard time with special USB cables. A standard USB cable can transfer data without error, or else there would be total chaos in the computer world. All USB DACs have a master clock which takes the incoming USB data and then processes it with respect for timing. It is logical to conclude that it's is only after this crucial process of clocking that we need be careful with the conveyance of the digital signal.

USB from my computer is sent to an Empirical Audio Off-Ramp which then clocks the signal using an Audiocom Super Clock 4, and coverts to SPDIF digital output for my DAC to receive.

SPDIF is considered a flawed standard of digital conveyance due to it's method of converting the data bits and clock into a sort of bit clock infusion with the purpose of serving the combined signal as a single bus.

The SPDIF Receiver Chip in the DAC unit then has to receive this infusion and convert it back to the separate data bit and clock streams as it was before the SPDIF convertion. This is because DAC chips generally require that the data bits and clock be delivered through separate buses.

And all this hopefully done without altering the timing of the master clock. There is a lot of opportunity for jitter to creep in in the process of converting to and from SPDIF. The AES/EBU standard is the same process only using balanced interconnection cables. So while an improvement, it is still not ideal.

I²S is the best means of delivering the digital signal after being clocked since it maintains separate streams for data bits and for clock timing. And it maintains this separation throughout and can be delivered directly to the DAC chip without the need for any conversion to and from SPDIF therefore bypassing at least two chips in the process. By taking out these additional manipulations you have lower jitter and fewer data errors if any.

The Off Ramp has an I²S output so my next project is to ad a daughter board to my DAC allowing for I²S input directly to the DAC chip. My DAC doesn't input I²S natively so a little modification is needed.

Example here http://www.pavouk.org/hw/modulardac/en_ad1865i2s.html

Many DACs have I²S inputs now.

http://www.empiricalaudio.com/ for Off Ramp info and using computer as transport related info.

Once the DAC does it's thing and you are dealing with an analog signal then it can be handled just as any other analog signal.

The integrity of the digital signal really is such a delicate thing. Same thing goes for a low level analog signal. For a while now I've been thinking about how amplification in an audio playback system is a lot like multiplication in optics. Each amplification stage being like a lens in a really powerful microscope. In audio it is said that the most important component is the quality of the source be it CD or Vinyl. Just as the most important lens in a microscope is the first one. If there is any distortion in the optics of the first lens, then that distortion will be magnified several times by the subsequent lenses. A small imperfection early in the system will become a major problem later on in the chain. Same is true in audio. That's not to say that subsequent stages aren't equally important, but if the first stage isn't up to snuff than you wind up magnifying distortion.

So with this in mind, when we talk about the nuts and bolts of digital data and timing and getting it to the DAC chip in the most pure form, the outcome improvements can be substantial.

In the case of a USB DAC and also with an Off Ramp, the computer sees the component as a sound card and uses the USB cable as a bus through which to feed its data as needed. Audio files are read and data is provided to the sound card according to the time index of the audio file and so yes the rate at which the packets of data are transfered to be clocked by the sound card is probably within milliseconds of the ideal real time playback. And the signal is buffered so that a slight hiccup in data transfer will not result in an interruption of playback. However jitter is measured in picoseconds and there is one quintillion picoseconds within a millisecond. So by comparison, the job of the USB cable is really quite lax.

Another cool feature of the Off Ramp is that it takes the digital signal from the USB cable but operates on it's own power source rather than using the USB power tap. So all that dirty computer power naysayers point out is not an issue here.

Having said all that I have never tried a nice USB cable. And if it makes an audible improvement than my ears are open to it. Stranger things have happened. I'm just reluctant to pay out the money to find out.

I'm really happy with my digital playback. But I've got a long ways to go before being done. Crazy thing is how in depth you can get with it and then realize that it is just a small part of the overall system, each part needing equal attention. We go from talking about digital theory, to vacuum tube amplification theory which is a whole gamut of natural laws in practice. Audio transformers are so cool and mysterious. Then electrical signal to mechanical motion causing sound waves through speakers. And finally room acoustics which is a whole nother science in it's own right. This is why audio is the greatest hobby there is.

When I was running my system on an Apple TV, before I got hooked on hi-res files, I considered putting an Empirical Audio Pace Car re-clocker on The ATV. Info here: http://www.empiricalaudio.com/products/pace-car-reclocker

It was impressive the difference that re-clocking, and by extension, asynchronous USB, can make. Taking the computer "out of the clocking business" is huge. In the end, I decided not to invest that kind of money—let's be honest, most high-end audio product makers are not as focused on value as Steve—in a system tied to the ATV. And I am glad I didn't. The Audiophileo is a huge improvement for me, and to my ears, reminds me of the effect I heard when I listened to a Pace Car. Detail, clarity, and air all improved.

For my next DAC, the Empirical Audio USB DAC is on the short list with the W4S, Ayre, Metric Halo and Cosecant. All are asynchronous; never going back. That's he only thing that keeps the ZDAC off the list.

One interesting thing I noticed about the Pace Car is that Empirical says that using it means you can use a cheaper USB cable, presumably because the cable is less critical. Of course, that doesn't keep Empirical from selling multi-hundred dollar USB cables. I have 2WireWorld Ultraviolet cables, because I wanted a quality cable that I knew met USB spec without spending an arm and a leg. As it turns out, Audiophileo supplied the very same cable with their Model 1. I think the rub with USB cables is that, depending on their use, the role of power in the cable, and whether the cable itself meets USB spec (so many do not), and whether there is post-cable reclocking going on, the cable can make a difference.