It's finally PC upgrading time for me, so along with that, it's time to ditch my Creative Audigy2 ZS and Cambridge Soundworks Megaworks 650 6.1 setup for something higher end. After reading around the forum, it seems like either the E-MU 0404 or M-Audio Revolution 5.1 are the soundcards I should be seriously looking at. I primarily listen to music on the computer and rarely play games. I do also watch DVDs on the computer too. I don't care about a soundcard recording/authoring abilities, since thats not anything I'll ever do.

Along with the soundcard upgrade, I'm planning on getting an A/V receiver. I've been looking at getting the Denon AVR-1507. Granted, I'm on a college budget, so I don't have the luxury of spending a lot. I plan on slowly acquiring my home theater setup over the next several years. I don't have the money for a complete speaker upgrade, so I'll be using most of the speakers that came with the Cambridge system for now. Based on all of that, I've got a few questions:

1) My master plan is to use the optical (or coax) SPDIF output from the soundcard to the A/V receiver, with the idea that the fewer analog<->digital conversions, the better the sound quality. Is this thinking flawed for my limited budget range? Yes, I realize there are high-end, tube driven, analog systems which still kick ass; those are not for me.

2) Since I plan on using the digital connection of the soundcard (primarily), it doesn't really matter as much about what DAC's the card has, right?

3) I've read the Revolution 5.1 is technically better/newer than the Revolution 7.1?

4) Does the E-MU 0404 & M-Audio Revolution 5.1 allow SPDIF passthrough of AC3/DTS? i.e. can I watch DVDs on my computer and properly pass the digital audio signal to the A/V receiver for proper surround sound?

I will try to answer a couple questions for you. Im no full Audiophile but I am running the type of configuration you are thinking of except only in stereo and not surround. Audigy 4 (optical)-> Sherwood rd6108 receiver -> pair of Mordaunt Short Avant 902i's.
The M-Audio 5.1 from what I can see doesnt support a Optical SPDIF only an Coaxial which is still great as long as the Receiver you are going to has a Coaxial input.

2, In regards to the question about the DAC's it shouldnt matter as the receivers DAC's will do the work. However there has been questions about the quality of the circuitry on the card allowing noise from the bus through but in the case I dont think any of these cards would have this isue as it would have been brought up before.

3, I dont know about the Revolution being better then the 7.1 version?

4, Both cards should allow PCM/AC3 Data through to your receiver to decode Standard and dolby/DTS signals. Just make sure that everything is pointed to SPDIF on your software and the Receiver should indicate wether it is processing Dolby/DTS.

Someone step in if Im wrong

When your only (?) goal is SPDIF output, why spend that money on expensive
soundcard ?

Any onboard Realtek 883/888 should allow bit-perfect output to your receiver

Thats true!
Is it still true that if the board that you are using that codec on is designed badly can there still be intereference be heard through the output?

I get interference on mine, it's a piece of crap.

That's a digital stream. Bit-perfect should mean bitperfect. ALL bits from your PCM stream MUST reach to the receiver. To achive that ASIO or Kernel streaming output should be used to bypass mixer.

How badly is the analog part implemented, does not affect at all. Digital part is there or is not there. 0 or 1

But be aware, not all on-board chips have possibility for bit-perfect !

Alright, time has passed and I've finally got the new computer built. I've briefly played around with the optical/analog outputs from the 0404. The sound is a LOT crisper coming out of the Denon receiver, instead of my old Cambridge soundworks setup. In an unscientific listening test, several family members report that the analog sound is better/more full than the digital one.

Anyways, one problem I've run into is the inability to pass Dolby/DTS signals out to the receiver while watching a movie with PowerDVD 7. I've made sure the receiver is set up to use the optical input. If I select SPDIF as the audio output in PowerDVD, no sound comes out of the receiver (nor do I see anything in the E-MU mixer). Yet, if I change PowerDVD to audio setting to 6 speakers, sound obvious comes out (since its via the Wave Out mixer).

What needs to be done to get DTS/Dolby signals passed from PowerDVD to the receiver, via the optical connection?

What OS are you running?
Im running vista. In the sound devices it may show 2 devices. Speakers & SPDIF Out. You should have SPDIF Out set as default. Make sure you go into proporties and make sure that you have ticked the DTS, Dolby Digital and also the sample formats that are compatible with your receiver.

Once that is done, you should be able to select the optical channel on your receiver and that's it. I havent used Power DVD but if you cant get sound working in windows via Optical out then get that working first.

-Ditch the soundcard. Get a cheap USB DAC (or spend a little more a get Scott Nixon's delicious, unfiltered, tube buffered version) and skip the bloody S/Pdif interface and all the jitter (and NEVER use toslink, unless you're a fan of needless optical/digital conversion (x2). Digital Coax (aka any 75ohm composite video cable) works fine. Spend some time on google and visit the empirical audio website for some good write-ups.

2) I've found Denon to be a bit on the bright/harsh side. NAD is quite good. Arcam even better (plus they don't break). If you're on a budget go with Outlaw Audio. Best bang for the buck hands down.

How come?

Lower bandwidth, more jitter, and as I said before, why on earth would you want to introduce two additional rounds of conversion into the signal path if you didn't have to?: See here.

With a decent shielded coax, emi and ground loops should not be an issue. Not to mention optical cables are much more fragile (kink or step on one and you'll find out).

TP

Good heavens. Can you *hear* these supposed problems?




Or maybe the AVRs are indistinguishable as amps, when they aren't driven to clip. You'd need a level-matched blind comparison to tell. I'd say, go for whichever one has the features you need, and power suitable for the room/speakers you're using.

Btw, citing Stereophile is not usually a confidence-builder here. Citing Robert Harley, who believes the crystal directionality of copper wire makes an audible difference, is downright worrisome.

The conversion is totally lossless, i.e. there is no data loss during the conversion and transmission. Also

And finally, audibility of jitter is yet to be seriously researched:

Fragility: Yes I can. I kinked a toslink cable once. The result was not pleasing to my ear
Conversion: Have ab'd it, so no. That the conversions take place is not in dispute. Perhaps Egor can post results that prove said conversions are lossless, produces no jitter/synchronization issues.



Assuming none of the AVRs under comparison are overloaded in current or amplitude, sure. We'll also assume that they're solid state (I can A/B my SET amp against my Arcam solid state in my sleep), and oh yeah, that each of the AVR's has a true analog bypass that allows us to isolate the amplifier stage from preamp/processing



I'm well aware of the audiophoolery in that publication and others, but he didn't generate those graphs, so we can safely assume that toslink produces more jitter, according to results shown (this finding has been replicated elsewhere, though I don't have the links handy). As for audibility, it has been researched form what what I've seen. A quick search of HA provides links to academic papers that evidence the audibility of sinusoidal jitter in the picosecond range, well within tolerances of modern digital equipment. I won't pretend to fully understand the science behind all of these phenomena, but I'm hard pressed to refute peer reviewed studies, as some of these papers seem to be. And Egor, as I stated, cross talk on coaxial cables is a non-issue, particularly over short runs.

PurpleHayz, can you explain me, why my setup works flawlessly ?

I have PC, soundcard Terratec EWX 24/96 stereo with optical digital out, non-resampling soundcard (ie., 44.1 kHz stays 44.1, and 48 stays 48), connected digitally with "optical thin and most cheapo cable, 5 m ," to my HK AVR 5000 amp/receiver/decoder.

Ok, you might argue, my ears could not be capable to detect the quality difference (as eg. the A/B comparison is missing).
But now my key question and argument:

I play sometimes CDDA (16 bit, 44.1 kHz PCM stereo redbook) containing 5.1 DTS signals inside the PCM container stream.
Because the DTS is data, and every single bit, every 0 and 1, has very special meaning (opposite to the bits inside normal stereo PCM, where single wrong bits don't affect the msuic experience in such a way),
every wrong bit would destroy the ability of the dts decoder, to decode the 5.1 music properly.
Errors, eg. scratches on the CD, lead very quickly to annoying pops in the 5.1 music,
whereas scratches of same quality on a normal stereo CDDA don#t lead to annoying pops, because the error correction of the CD drive is working more or less.

So, if jitter etc would have any influence in practical setups, it should show up clearly during transmitting 5.1 DTS.

I think, the jitter existing is way below the necessary threshold, signal to noise ratio, to change the "0 and 1" values being transmitted.

What's correct, is, that the optical cables should not be 100 m long without reinforcer (but 5, even 10 m is no problem),
they should not be crimped to curves with very narrow radii.

Regarding PC, HiFi setup, maybe additional satellite TV with digital audio connection coaxial to amp/decoder,
it is safer to use optical instead of coaxial, as otehrwise the possibility increases, that you get very annoying electrical loops in your system with buzzing because of different potentials of PC, HiFi, satellite cable etc etc.

Well, Toslink was desingned by Toshiba's engineers to stream data in digital domain without loss. If you believe there is loss, then your S/PDIF device is simply broken. Edit. Oh, and the other possible reason you've successfully abx'ed toslink from coax is psychic abilities. But this is kind of unscientific approach .

Do you wish to know how Toshiba's Toslink technology works with light and there is no data loss? Read and understand the reference document:
http://www.toshiba.com/taec/components/Pro...ide/toslink.pdf

I did a "quick search" of HA but couldn't find links to these papers. Please can you provide the actual links?

The one paper I am aware of which conducted a properly controlled test is this one:

Eric Benjamin and Benjamin Gannon, "Theoretical and Audible Effects of Jitter on Digital Audio Quality", Pre-print 4826 of the 105th AES Convention (San Francisco, September 1998).

Unfortunately, because it's an AES paper, you can't just download it for free, you have to pay something like $20 for it. However, it concluded that the lowest level of jitter which produced an audible difference was 10ns (that's 10,000 picoseconds). Moreover, that level was only audible on a high level 17kHz sine tone. On normal musical programme material, jitter was inaudible below 20ns.

Anyhow, even if you do believe that jitter is important, I have to take issue with something else. You recommended getting a "cheap USB DAC" in order to avoid SPDIF and the related jitter. As far as I am aware, nobody actually makes a USB DAC that operates in asynchronous mode (which is the only mode where the DAC's clock is in control and would indeed eliminate jitter across the USB interface). Cheap USB DACs work in synchronous mode, which typically has very high levels of jitter, certainly far higher than most SPDIF implementations. The better USB DACs use adaptive mode, which can result in low jitter levels, but if you are genuinely bothered by jitter, you should be aware that it still exists in USB DACs.

Perhaps you can post data that show these conversions would be audible?



IIRC there are conflicting published (JAES, web) reports on threshold audibility levels of jitter (which seem to be frquency- and correlation-dependent). See discussion here:

http://www.hydrogenaudio.org/forums/lofive...php/t51322.html

and here

http://groups.google.com/group/rec.audio.o...e0dea3671c45094

Well in the past week since my last post, I've got it mostly figured out. Currently have Window sounds / Winamp via ASIO going out the analog ports on the E-MU (it has a better sounding DAC than my Denon receiver). Have to use the Soundmax optical out for watching movies, since the E-MU apparently doesn't Dolby/DTS passthrough. Furthermore, had to play around with the various default sound devices in Windows, since PowerDVD 7 doesn't let me select the sound device for output.

I've got the line-in (from my tv card) going into the E-MU, which works fine except volume control in WinFast TV doesn't work; must rely on the receiver for that. Semi-annoying. For Skype, yes I have specified for it to use the Soundmax for the playback and mic, which works great.

I tried taking the optical output from the Soundmax and routing it into the optical in on the E-MU. While it worked fine for general audio, it didn't properly pass the Dolby/DTS signals to the receiver. The soundmax doesn't seem to always output audio on the optical port either, when its not the 'default device' in the Sound control panel; annoying. While I was able to get digital sound, the receiver didn't properly detect the type of digital signal and do surround sound.

You would need some seriously crappy components and broken fiber to make the conversion audible. The transmitter is a LED driven directly from the binary stream, and the switching times are in the range of microseconds. The receiver is (most likely) a phototransistor with similar timing plus some filters to eliminate transition glitches. Nothing in there introduces any jitter that the bitstream didn't already have.

It's true that the bandwidth of TOSlink is lower than coax and that the fiber loves to break. Most polymer fiber also has a really high attenuation, so you need to keep it as short as possible. With coax, you can get to relatively long distances even with a simple mono RCA cable if you can't find your coax.

Well, for ground loops, TOSlink is the ultimate cure... In the case of coax, you'd need some kind of ground breaker, since those carry ground. But since ground loops are only an issue with analog equipment, it doesn't really matter, except perhaps for obscure setups.