Hi,

I have a few questions regarding my audio setup and would be grateful for any help.

My music is stored on a computer hard drive in FLAC format. For those unaware of FLAC, it differes from MP3 in that the file is basically uncompressed and "lossless", i.e. the data is identical to that on a normal audio CD (http://flac.sourceforge.net/).

My questions are:

1. What is the best way to get the sound from the computer to an (approx) £20,000 hifi system ?
2. Will the sound be as good as played from a dedicated CD player ? If not, why not ?

Things I'm thinking about are sound cards, cabling, external DAC, etc. I'm not too bothered about "streaming" my audio to multiple rooms. Many thanks.

The two obvious options are (1) a high end analogue sound card and (2) a cheap sound card with bit perfect output streaming to an external DAC. The second one is preferable, and should give you better SNR (less noise), but with a good internal card this is likely to be a very small effect. I am sure another forum member will suggest a good sound card.
Yes, it will be as good as a dedicated CD player, provided things are set up correctly.

Does the 20K hifi have digital input?

The FLAC files might sound better than any currently available dedicated CD player if you used a secure ripper which can recover from errors (which any dedicated CD player cannot), even on unscratched CDs.

Thanks for your replies.

The preamp does not have digital input - RCA and XLR inputs only.

Re. the external DAC - which would be recommended ? Any how would the soundcard connect to the external DAC ?

Currently I use a Creative 7.1 X-FI Elite Pro sound card.

M-Audio Audiophile 2496


In theory, sure, however my Krell CD player sounds noticeably 'better' than my M-Audio 2496 soundcard and I can tell which is which blindfolded.

Oh, but can you ABX it?

Some people like Benchmark DAC1 USB, it is supposed to be good and has also XLR outputs if you wish to use them. I have never heard it though.

2tec - I wonder what difference you would hear if you used an external DAC playing FLAC format ?

Re. the DAC1 USB - is the audio data streamed digitally via a USB port & cable to this device ?
And would my current soundcard be suitable to connect to the external DAC ? How would they connect ?

USB dac's connect to the computer with USB cable and appear as another sound card. It has nothing to do with any possible other sound cards in the computer.

Would the external USB DAC option make more sense than connecting the output from my current soundcard to a non-USB DAC ?

It only makes less sense if your soundcard supports bit-perfect digital/optical output.

This depends where your PC/Server is.Limitation is the amount of usb cable length which is said to be effective till about 5 meters. Unless you add a hub which will extend that distance. I have a DAC1 connected to UD10 and the sound is as good as any CDplayer I have heard.

You can connect in the following ways:
Soundcard RCA (analogue out) to pre-amp analogue in, the sound card does the conversion (DAC) and amplification to line level.

Sound card SPDIF out via Toslink (optical) or Coax or XLR to outboard DAC. This is the conventional way to connect a transport and a DAC. Probably all cards do this bit perfect but might differ in jitter.

USB port via SPDIF converter to conventional DAC. The way to connect an outboard DAC without USB input.

USB port to USB input DAC. Benchmark is an example, they convert USB to I2S. In principle this is input jitter immune

Best seems to be the I2S protocol between PC and DAC (jitter free by design) but I don't know any PC with a I2S output and there are very few DACs with a I2S input (North West)

Another option depending on your computer/network setup is the Transporter. It has XLR outputs so it should integrate well with your preamp.

If I'm using USB port to USB input DAC, why do I need a PC with an I2S output ? Doesn't the DAC just plug into a USB port ? Is the data then being transmitted from the USB port to the DAC free of any interference ?
Also, do you need special software to tell the computer to stream audio out the USB port ?
As regards the transporter, would this be better than an external DAC like the Benchmark ? I'm interested mainly in audio quality and not streaming audio to different rooms in the house.

You don't

As much as any other

No,but most recommend foobar(freeware) because of its flexibility.

Better is very relative here...Transporter offers more flexibilty,screen,remote,ect.The DAC is just a Dac.
Since you have that expensive system already,you should try both and see which you prefer...

USB port is all that is needed for USB DAC, no need for I2S.

I have never had any interference problems with USB DACs, even when the signal has been routed through hubs.



If the software you use allows to choose from multiple sound cards, that should be all that is needed. The USB DAC will be just another sound card from the software's point of view.



Transporter is be like digital audio player with display remote control, the remote streaming is just how it has been implemented. It is more of a usability issue -- I have three Squeezeboxes and really like them since they give easy access to full music library.

If you like the approach of a Transporter but do not need all that functionality, a Squeezebox with external DAC might also be one option. The builtin DAC in Squeezebox is not too bad but it is not a great one either.

I can't comment the DAC quality in Transporter, but it is supposed to be good.

unless your PC is silent, I suggest one of these:

http://flac.sourceforge.net/links.html#hardware

I have personal experience with the squeezebox, transporter, and sonos and they are all very good devices. the digital and analog outs on the squeezebox and transporter are very good (see also)

Thanks for all this great info ....

If I use an external USB DAC, will the existing soundcard (X-Fi) affect the sound at all ? It tends to use effects such as crystallisation and EAX.

No, as mentioned, the usb dac will appear as another soundcard in your windows control panel and if the application has it's own output options. You simply choose the usb dac as the output device and that is where it goes through.

The DAC in the Transporter is close to state-of-the-art. It's certainly up there with the Benchmark. Most people who have compared them remark that the Benchmark is more "crystalline" and can be a bit in-your-face. I have a Transporter and IMHO it sounds fabulous.

Mind you, remember that to use a Squeezebox or Transporter you'll need to run the associated server (SqueezeCenter). If you are intending to just run whatever media player you feel like then the Squeezebox system isn't really an option.

I would bet that none of them have ABX results to back this claim up to HA standard, though.


Would I win that bet?

Probably :-)

If I'm going for the external USB DAC option, would the cables inside the PC have any impact on the audio quality ? For example. the SATA cable transferring FLAC data from the hard disc.

I find Monster SATA Deluxe cables to bring more warmth to the music. MSRP: $499.99.
Is a smiley really necessary here?

1 smiley only?

Your present PC audio system is good enough for to do the job. Even it supports unbalanced type connections only (analog path). Take the output from X-Fi's PCI card (X-Fi LineOut port 1 -> Hi-Fi Aux or Tape-Input RCA). You'll need to get a suitable cable or adapter for connection (2xRCA <---> Stereo MiniJack plug cable/adapter).

If you want some improvements then just get a new audio interface w/ balanced type output ports (as I understood that your Hi-Fi system has balanced XLR connectors for input available). If you choose an external DAC, you possibly (if the DAC supports S/PDIF) be able to route the audio signal from X-Fi optical output as bit-matched into DAC. I can't say if the optical path is better than what it would be through USB port. Remember use ASIO device drivers for output.

I have a SB PCI card and E-MU USB interface running simultaneously w/o issues you're afraid of ... same w/ the possible internal noising (IMO, the PCI card is maybe more sensitive for these). One weakness, to mention, for USB audio interface(s) are periodically CPU intensive processes/services as like a dongle drivers, meant for to protect software, which are periodically checking the status of possible running protected software ... this CPU 'spike' can even mix the sync for audio -> crackling/distortion/dropouts in playback.

Juha

Is this the monster cable you refer to:

http://www.beachaudio.com/Monster-Cable/Jp...-Hp6-p-416.html

Or do I detect a hint of sarcasm ? The reason I ask about cables is because audiophiles spend many ££s per metre on cable - and cables inside the PC are cables also !

LOL
Turns out one smiley wasn't enough. Yes, it was sarcasm. Yes, "audiophiles" who spend hundreds (if not thousands) of dollars on cables (no matter what kind) are utter idiots. Any SATA cable will do.

Lol..... I believe there would be many who beg to differ. Most people who own a good system, in fact !

Hi. I'm also running a very similar post, which compares ipod as source with cd players. It's called Ipod Playing lossless vs Cd player, http://www.hydrogenaudio.org/forums/index....showtopic=63450.

To danhackley, I'd say that if you already spent £20,000, the way to go is digital source to DAC, and then ultimate connection cables to amplifier. You could use the ipod and the wadia itransport as digital source to feed into the DAC.

Read my other thread and say what you think.

BTW, I thought FLAC is compressed.

I know it's lossless though.

Can someone give me an answer on this?

You can compress in 2 ways, like WinZip http://www.howstuffworks.com/file-compression.htm, this compresses a file without loss of information by using redunancy, or like MP3 by removing information there where our hearing is less sensitive. This gives a better compression at the expense of loosing information.

Most people usually have a hard time admitting they were wrong. Even harder to admit they spent thousands of dollars for nothing.

Ha well I can imagine that !

In my opinion, it would make no difference, as I believe there's simply no way to get a personal computer to sound as good as a dedicated high end transport. The Krell chassis, powersupply and the transport itself are each worth far more than my PC, and the Krell's design is specific to the reading and preamplification of a CD. From what I understand, it isn't that the digital signal is in any way different but the resulting conversion to analog and the amplification are accomplished in a much more accurate and precise manner, resulting in what to me seems like a 'cleaner' sound. In addition, the Krell has far more gain.

Ah yes, I'm now completely sure you're absolutely convinced that a cheap audio system sounds just as good as an expensive one! Please, don't let me in any way dissuade you, as you are, of course, perfectly free and able to carry right on with your own beliefs. I, of course, intend to do exactly the same. Oh, and by the way, I for one, am just so happy to now know just how easy it it to satisfy your lofty standards.

I almost choked on my popcorn, lol.


It would be neglectful to not mention that the Benchmark DAC1 (and a lot of other "high-end" DACs) re-samples (SRC, Sample Rate Conversion) all inputs at any sample rate to it's max sample rate (in this case 192kHz). Their marketing department calls it "UltraLock" and they are very careful even in the manual to dance around the fact that it's actually adaptive SRC.. Here's one such dance-move: Basically, this is cheating more quality out of these converters by having the filters run (mostly) way above human hearing, and unfortunately for many DACs - their adaptive SRC sounds better than the filters they use.

Around the 750-1000 USD range, the Apogee Mini-DAC is one of very few converters that do not re-sample from the input clock, but does re-clock the incoming clock using similar methods as the DAC1, except without the SRC. The outcome is the Mini-DAC, in my opinion, sounds more transparent & accurate than any DAC under $5000 that I've heard. As such I've sold my Benchmark DAC1 and my Mytek Stereo96 DAC... both truly outstanding sounding converters for the price - I'm not denying that. But imo Apogee stands in a league of it's own in the $1000-sub price range, and all I'm suggesting is that you have one auditioned on your system with a few other good DACs there as well.


Hahhah, I almost choked again... I really shouldn't be eating popcorn while reading some of these threads. I will reply to this one though.

"good enough to do the job" isn't what we're discussing here. If we were, a $10 Chinese usb converter with a "headphone output" would be adequate.


I'm not sure weather to seriously reply to this, since you've already used an M-Audio converter as your example of "a personal computer" not being able to sound as good as your Krell.

But that's an interesting assumption, that "a personal computer" can never sound better than a given thing, because it is capable of providing a digital-audio data stream that is MORE accurate than any CD player ever can be... simply because CDs have errors - period. C2 error correction can be quite good, but even so it's not 100% perfect. Software playback however can be perfect. Not all software is obviously, but some is. As far as CD ripping, that's not somthing I'm talking about, or wish to get in to. That's a whole different conversation.

I do however agree with you on your opinion that the resulting DAC is the weak-link for the most part. Maybe you should audition an Apogee Mini-DAC or Rosetta 200, as well as a few other good converters like the Benchmark DAC1, Mytek Stereo96 DAC, Grace Design m904, and a Prism Dream DA-2... for starters. If you're into hi-fi then none of these are really "bank breakers" for you, since you're in for the long haul. ...But I could start recommending converters in the $10,000+ range if you're interested.



[edit] oh and, i forgot to mention... not only are the stats listed for the Mini-DAC tested at +24dbu output... but the output knob goes up to 11. [/edit]

Ok, then bring ABX results to show this. Saying "it's just my opinion" is not going to fly. I would imagine what you are hearing is not a better DAC, but some sort of cognitive bias.

In which particular objective measurements would you say your Krell is better than a PC? Higher SNR? Flatter frequency response? Lower THD+N? Lower distortion in a particular band? Audio electronics are not powered by hope, dreams, magic fairies or the ghosts of old baseball legends - exactly the same laws of physics apply as with all other sorts of electronics. If you are going to claim that your Krell is "better" your are going to have to define "better" in an objective way.

Of course. The "audio is a science/technology and not some sort of magic powered by faeries" crowd are only saying that because we have ears made of base metals, not hewn by the gods from pure gold like most audiophiles. Or maybe we just honestly, honestly believe that claims need to be backed up with evidence. "It's good because I paid a lot for it" isn't evidence.

How do you reconcile this belief with the fact that a dedicated high-end CD transport *is* a computer (which happens to get its data from a spinning optical disc rather than a spinning magnetic disk). Or are you referring to the analogue stages in a high-end CD player?


Let me preface my response to this by declaring my total conversion to streaming audio. I own three Slim Devices players (a Transporter, a SB2 and a Duet). I sold my (very expensive) CD player a couple of years ago.

Nevertheless, your statement that "CDs have errors" is rather disingenuous. While it is true that there are some isolated cases when C2 error correction fails, the simple fact is that the overwhelming majority of non-faulty CD players deliver a 100% accurate data stream when playing CDs in reasonable condition - even mildly scratched CDs usually have no uncorrectable errors. Yes, a hard disk is ultimately less error prone than a CD, but for all practical purposes a CD transport is an equally reliable source of digital audio samples.

It wasn't my intent to say that all CDs always have some error *after* C2 error correction. My point was that C2 error correction isn't 100% perfect all of the time. Have you used EAC before? What percentage of even brand-new fresh out of the polywrap CDs would you say have entirely 100% quality on every track?

That would be some interesting research to check out for people that claim that CDs are not error-prone - and it's already been done from many different angles. It doesn't mean that CDs don't sound amazing, they can. The format itself isn't the bottle neck even in many high-end systems.

You could check out the CD's White Book cousins, like VCD. They don't have any error correction at all, and yet manage to come across pretty well for the format's specs & intended quality.

Cognative bias? Surely you aren't serious? Personally, I can tell you that if there is anyone with a bias here, it would the one dismissing the experiences of others without good cause. Let me ask you, have you ever listen to the difference between any PC based source and a dedicated transport? I have, and if you think they are the same in terms of audio output, all I can say is all you've communicated to me is your inability to be objective and open-minded. By the way, I and dozens of others, have listened to both sources and I've never had one person choose the PC, ever.

Ok, since you clearly don't understand my point, I will restate it. Yes, digital audio is bit perfect, however, digital to analog and analog amplification are affected by both design and quality. Are you seriously suggesting that the engineering of my M-Audio soundcard is comparable to the engineering of my Krell? It seems to me that what you're saying is that you expect as much from a $99 soundcard as a $5000 transport, and furthermore anyone whose ever bought a quality piece of audio equipment has entirely wasted their money?

No, it's not good because I paid so much, it's good because it's well engineered. Yes, there are dubious audio claims, however, there is also well designed and well built equipment. Have you never heard these expressions, "You get what you pay for!", or "There's no free lunch!"? If you want audio quality, you need quality audio! People shouldn't allow themselves to be completely deaf in regards to the potential possibilities of audio improvement. If you think you can't do better, chances are you won't. As for evidence, there's been tons of reviews and testing done of the Krell line, if you require paper proof, it's out there for people like you who seem to need that sort of information.

Futhermore, this particular discussion is hardly on topic, perhaps if you feel so strongly about the issue, you should start a new one.

What, you don't think people have a right to challenge you on your TOS #8 violation?

2tec,

I think you may have wondered into the wrong forum.

No one here is in the least bit interested in sighted listening tests carried out by Stereophile, or sighted listening tests carried out by audiophiles. Any listening test of reasonable quality equipment where you know what you are listening to is invalid.

Listening tests should be carried out double blind, and yield statistically significant results. Then we can start arguing about which sounds better.

Feel free to ABX your $5000 transport against a $99 sound card digital output.

If all the ones remain ones, and all the zeros remain zeros; and if the subsequent D/A equipment is able to accurately recover and stabilise the clock; and if (other than the desired result of each one and zero) the input signal has no other effect on the output signal (i.e. there's no way to pass noise from the digital input to the analogue output), then I can't see how either can sound different from the other. That's a lot of "if"s - very few D/As are that well built, but at the price point you're describing, one could be hopeful.

If you hear a difference between two digital streams carrying the same ones and zeros, you may conclude that there's some undesirable difference between those two digital streams, or you may conclude that your D/A isn't doing a very good job.

Cheers,
David.

If C2 correction can be applied, it's perfect. Period. (If it can't be applied, then that's an uncorrectable error and the usual course of action is to conceal the error with an interpolation). My point was that in the overwhelming majority of cases, there are zero uncorrectable errors on a CD. Here's a data point: when I was ripping my CD collection to hard disk a few years back, out of about 1100 CDs, i came across about half a dozen that had uncorrectable errors. So in my collection, 99.7% of the CDs were read with 100% accuracy.


Yes, I've used EAC (although it's not my ripper of choice). When the track quality figure is under 100%, it doesn't mean there were uncorrectable errors. What it means is that the drive reported some C2 errors and EAC re-read those sectors. In the context of playing a CD on an audio player (as opposed to ripping it), those C2 errors are routinely corrected. (Note: not concealed - *corrected*).


I've never touched a VCD, so I don't know what error correction it may have. But I'll say this: it cannot possibly have NO error correction at all. Every single CD triggers a constant stream of C1 corrections as it is read. Without C1 correction, you'd never get any intelligible data off any CD. Perhaps VCDs have no C2 correction? If they don't then they must be very vulnerable.

From the standpoint of ripping with EAC this isn't exactly true. I don't want to belabor the point, but CU errors will only directly trigger re-reading when C2 pointers are used, and the drive provides the correct pointer. If EAC isn't using C2 pointers, then CU errors will indirectly trigger re-reading so long as they produce inconsistent results between successive reads.

In the case of C2 errors being *corrected* this is done by both audio players and optical drives regardless of the ripping program.

This isn't to say that all optical drives perform error correction equally well, though I have a hard time believing this doesn't also apply to audio players.

Ok, then you should write a paper for a peer reviewed journal, because your findings will contradict the current state of the art in the field. Have you never heard of confirmation bias? If you bring ABX results, then I will start believing you that you can year a difference. I can't prove that I can't hear a difference - but you have a great opportunity to prove that you can.

And yes --- I have tried out some of the best equipment in the field --- both in the lab and with my ears.

Computer hardware is generally cheaper because of economies of scale. More boards means both lower production costs and NRE costs per board. Sure, it might not be as good - but then you have an opportunity to show that. Bring measurements. Bring ABX results. Bring circuit diagrams. Bring Spice simulations. But bring something more convincing than "it's better because I say it is". If that's all you have, I'll still be standing here with my null hypothesis.

I have a PC on my desk that outperforms what something that cost 100x the price only a few years back. Why is audio technology immune from the same forces that have advanced computing, storage, radar, radio astronomy, etc? Sure, there is no free lunch - but your $100 a course restaurant is serving the same food as McDonalds! At least if you want to say "no, this gourmet food is better than McDonalds", you could present scientific evidence that you can taste the difference between it and a Big Mac. Please point me to evidence - real evidence - that your Krell is audibly better than your M-Audio, and I will issue you with a full apology.

All I am saying to you is "bring your evidence". If he can, then I'll believe you. This is not my faith, merely what I believe based on the available evidence.

No, I was just suggesting that this had wandered off-topic and that it deserves a thread of its own. Furthermore, if you feel that I'm in violation of the terms of service, please, feel free to moderate this discussion as you see fit, however, it wasn't my intention to do so.


I have done so. However, comparing two different types of systems isn't the same as comparing different files through the same system. ABXing these two sources, to the best of my ability, yielded the results I posted. It is like comparing apples to oranges. I'm not contradicting science nor am I in disagreement with the philosphies of this forum, all I am saying is that engineering and component quality play an important role in the quality of audio reproduction.

I considered splitting the thread but looking at the original post, you can see that it is not off-topic:

Then I must have misunderstood Andre's FAQ, which I quote:

I've just done a test. My drive (a Samsung S202) does have C2 error reporting support. I checked with a known bad CD and EAC's "Drive C2 Feature Test" said that C2 errors are reported. So I then took a CD that I was confident is error-free and ran that through the same C2 Feature Test. As expected, EAC reported "No C2 errors found". I ripped a track in secure mode, and the reported quality was 99.8%. I ripped the same track again in burst mode, and did an fc /b on the secure and burst rips: no differences were found.

My conclusion from this test, and in the light of Andre's FAQ, is that when running in secure mode, EAC instructs the drive to report when C2 errors are present, whereas in burst mode it lets the drive get on with correcting the C2 errors. But I accept that I may have misinterpreted what's going on here.

If you're saying that a track quality of under 100% genuinely means that there were *uncorrectable* C2 errors (which EAC fixed by re-reading), then how come a burst mode rip gave identical results?