[Ruppin]

Responding, in part, to Raylpc, I agree that I would like input that compares apples to apples. I have asked for such specific input on another thread in this formumhttp://www6.head-fi.org/forums/showthread.php?t=164381.

I haven't used squeezebox, but do use Airport Express (optical) with a Mac system, and it certainly seems to be one hell of a tool, expecially at the price. I upload my CDs to hard drive in Apple Lossless (with error correction), and have been delighted. I currently use a HeadRoom Micro Amp (desk module) and a Micro DAC. Like it so much, I ordered a Home Balanced to see how much better it can get. My guess is that it will take a high end CD player to be a better source. But how good a CD player is something I'd like to know.

 

[jriihi]

Well this discussion comes just right time. I am just building system to play music from PC. I wonder if squeezebox is good solution or if 2x cheaper m-audio transit is enough for good quality optical output to external dac?

All i want now is pc->flac->dac solution. Rest is already almost done.

 

[bordins]

Quote:

Originally Posted by stevesurf Maybe we can set up a couple of sources, do a rip of a complex passage at the National Meet and get a vote from attendees.

Hi stevesurf. I talked with a friend of mine who is an audiophile reviewer. His listening room is one of the "best" I've visited. It is designed to express any differences of audio components. Even a small difference can be detected, maybe also by trained ears, e.g. lifting the AC cord away from the floor affects the sonics, etc. This kind of listening test should be also conducted.

The ability to perceive is another thing. This means differences exist but many people can't hear them. We need both kinds of proof.

 

[sonance]

I wish we could see some real progress in high end source technology. We're still slaves to jitter (no pun intended) because of some inane standard that made sense 20+ years ago. Jitter is easy to eliminate (not just work around) if the current mode of uni-directionally streaming data with a clock slaved to the transports clock wasn't being used. Sure we have "jitter-immune" DACs that try all kinds of tricks to get around this issue, but as long as SPDIF/PCM data is being transmitted you're just taking someones word for it that your setup is really jitter free - and there will always be the thought of how a better transport or cable or whatever could improve your setup.

Someone should make a DAC that uses formats other than standard PCM / SPDIF type signals where clock jitter is no longer an issue caused by a lack of bi-directionality. A better system would be a signal that wasn't slaved to the transmitting device (transport). The playback device ("DAC") should request data from the storage device ("Transport") as it needs it, in bursts or packets independent of the rate at which it is being played back, with the ability to re-request data for whatever reason. From what I understand USB DACs could do something like this already but don't quite.

For starters, an encoded system such as the audio coming from a non-lossy compressed (or encoded) variable bit rate data stream is almost by definition independent of the clock of the transport since it is expanded to a playable data-stream independently of the source clock. For example some of lossless streams employed for the next generation discs (Dolby TrueHD) should fit this criteria if they are truely lossless. Rather than spend many thousands on a top notch transport system and digital cable, you'd simply need to switch the signal to a type that is designed to eliminate the very source of jitter.
If such a device existed, you should in theory be able to accurately rip your existing CD collection to this new container format that could be transmitted in another format without being converted to PCM first (think AC3 or DTS streams, but lossless.)

But a better solution would be a "smart" DAC that would simply use your PC or other HD storage server as its source of data and would be more of a Player than a DAC. You would do the browsing, song selection and all other user interaction on this player, and it would simply request data over a bi-directional link in chunks as it needed it. It would always play back audio from its own memory and not really use the idea of a clock - it just requests data - over a wireless link or a wired data/network link - as and when it needs it, in chunks. It's the quantum leap we need to be free of quality limitations brought about by transports / cables.

Of course the analog output of this device forward would still be subject to the same old audiophile issues we've always had, but at least the transport -> DAC source would be bulletproof and taken out of the equation as a limitation to a top notch system without having to throw tons of money to fix something as silly as jitter.

 

[bordins]

Hi Sonance. We need a new way of audio encoding in order to achieve zero-jitter playback. PCM isn't good enough. While Steve Nutgent claims zero-jitter PCM playback is doable in PC-audio, with a proper design and a powerful computing power. You mentioned about lossless AC3 or DTS streams. When'll a format be chosen ?

People are building the best PC for playing PCM audio. People are agreeing on the best (new) digital audio format. Is this topic the chicken and egg question ?

I also wonder whether DSD encoding of SACD fits into this context ? Some top-notch CD players convert PCM into DSD data first.

 

[sniks7]

It's been said once, but I'd like to add my support to the idea..

Could someone please do a comparative review?! There must be someone out there with a high-end CDP, off-board DAC and a computer. It would make very interesting reading.

 

[Loftprojection]

Quote:

Originally Posted by sniks7 It's been said once, but I'd like to add my support to the idea.. Could someone please do a comparative review?! There must be someone out there with a high-end CDP, off-board DAC and a computer. It would make very interesting reading.

I've done a similar test but it is not quite fair because my "DACs" are not the same level as my "high-end CDP". I have an Arcam CD23T which was selling new in the $2000+ a few years ago. My amp is a small tube amp that has a USB fed DAC inside and I also have an Echo Indigo DJ sound card. So I tried to feed my amp three different ways.

- The CDP feeding the amp with analog RCA cables
- The computer feeding the amp in USB using the amps internal DAC
- The computer feeding the amp in analog from the line out of the Indigo

The CDP was better then the other two for sure and I'm using no compression .wav files plus Foobar on my computer. The CDP provides better detail, more refined and involving voices,...

Now, I've never tested my CDP against a "good" external DAC like an Apogee MiniDAC, a Benchmark, a DAC-60,... I'm sure a setup like that would be very close, maybe better than my CDP or just a matter of taste.

 

[sonance]

I haven't investigated all the audio formats listed in the Blu-Ray and HD-DVD specs, but Dolby TruHD sounds like a good candidate for a jitter free next generation format/stream that can a) Support higher bitrate/samplerate (and multi-channel) lossless audio while being jitter free b) Is a standard that is likely to be supported. IIRC DTS also has something similar in the specs (DTS-HD?). Maybe AAC too. I will look into it soon.
Of course this sounds good in theory but it would take someone with access to the real scoop on this stuff to really make a judgement as to how good a replacement for PCM these format are. Then there's the issue of licensing for these formats - probably going to be cheap because it needs to be included in a cheap player which is already burdened with all sorts of other licensing. Also we'd need to see an audiophile level component that is roughly like a DAC - that's probably not going to happen this year at least - if the awesomely slow level of change in the audio industry in the past is any guide. Face it, everyone hates new standards, but in this case there's a real need for these new formats for people like us interested in eliminating jitter.
Some of these audiophile company made HD based players might meet this jitter-free need sooner than a TrueHD or DTS-HD "DAC" audiophile component and would probably be more flexible, but they're expensive and they dont use your PC for storage (big mistake in my opinion.) I don't object to a device like this having a HD, but it should let you use your PC, where you can backup your data, stream things from various sources, etc. Plus it would reduce the cost a little.
On the other hand I think we will see receivers and pre-processors with TrueHD and DTS-HD this year, which are pre-cursors to a DAC. Once ICs from the common companies that supply audiophile components (CS, AD, BurrBrown etc) start supporting these things they'll just start appearning in DACs soon enough. I'm guessing the connector for this type of audio will be HDMI (instead of optical or coax spdif.) Hopefully we'll see lossless encoders for these standards soon too.
Sooner or later, the reasons to look down ones nose at PC audio will go away.

 

[wakked1]

Direct CD playback is flawed simply because its an obsolete approach. Unbuffered unidirectional output is just plain dumb, it makes it impossible to retry in the event of an error.

The only true error correction is resend... trying to massage the bits back into shape after they've been read once is a guessing game at best.. and that's all the expensive anti-jitter stuff in CD players is.

Ripping a CD with a high quality setup (i.e. Exact Audio in paranoid mode) gurantees you retrieved all of the original data (assuming that is possible).

Deliver that 16bit / 44.1 data to a DAC via USB and you have GUARANTEED delivery of the original CD data into the DAC. Jitter simply does not exist in this model.

How it sounds is purely a function of the quality of the DAC, amp, speakers/headphones and interconnects.

Frankly the concept of spending thousands of dollars just to read bit-perfect data from a CD is simply insane.

 

[Akathriel]

wakked you made a lot of very broad comments in there that aren't necessarily untrue but aren't necessarily true either. With current SPDIF usb your claim is impossible, however IF the usb converts direct to I2S AND IF the usb is asynchronous not isynchronous, this is possible. Unfortunately, these technologies are not widely available today. The only things I can think of are the Wavelength Audio products which claim to meet both criteria (but there is no scientific evidence), the North Star Dac192 which has I2S inputs, as well as EA off-ramp and I2s modded products.

 

[slwiser]

What is the real deal about I2S and S/PDIF? Based on this post they have the same basic jitter specifications:

http://www.audioasylum.com/forums/pc...ages/9601.html

I2S has the same jitter spec as S/PDIF from this...and the ideal of zero jitter is a non-stater when you consider that even the specifications do not require it.

 

[sonance]

Here's a stab at a player (the Olive Opus) that should have no jitter besides whatever it picks up on the rip process (which is not happening while playing a disc, so it has the ability to retry, etc.) It also plays back audio using ethernet and wifi protocols which are not unidirectional (and do have a built-in retry on error methodology,) and since the audio wouldn't need to be converted to SPDIF before the DAC stage, it *theoretically* is jitter free. Doesn't mean that this actualy implementation is perfect, but in theory its only limitation should be its DAC stage, not the transport or connection between transport and DAC.

http://homeentertainment.engadget.co...s-audio-3-000/
and a direct link to the company making this:
http://www.olive.us/p_bin/?cid=01_07_opus

If this was in the price range of a DAC-1 / Lavry Black, and had similar audio fildelity I'd be all over this.

 

[wakked1]

Quote:

Originally Posted by Akathriel wakked you made a lot of very broad comments in there that aren't necessarily untrue but aren't necessarily true either. With current SPDIF usb your claim is impossible, however IF the usb converts direct to I2S AND IF the usb is asynchronous not isynchronous, this is possible. Unfortunately, these technologies are not widely available today. The only things I can think of are the Wavelength Audio products which claim to meet both criteria (but there is no scientific evidence), the North Star Dac192 which has I2S inputs, as well as EA off-ramp and I2s modded products.

Its inevitable. You could do it now using bulk transfer, all you need is a big enough buffer on the client device and the supporting error correction in the drivers. The only hard limitations are latency (host and transport) and throughput (bandwidth), both of which are more than adequate with USB 2 or Firewire.

The first generation of USB sound cards are understandably cheap, but implementing a jitter free USB sound card is feasible and inevitable, because there is nothing to prevent it. If I can trust my USB hard drive to correctly store several hundred gigs of data I can trust my USB card to play so measly music correctly (yes its a different protocol, but USB is just a transport for various protocols).

The whole USB error rate thing has been overblown anyway, I really can't hear any with either the Porta Corda III USB or the U24. Maybe in a RF noisy enviornment using a very crappy unshielded USB cable.