[leeperry]

let's say that I'm bored to death w/ crappy soundcard drivers...that are unstable(C-Media) and constantly resample everything...

if I were to use these 100% bit-perfect GPL drivers on a compatible card: cmediadrivers - Project Hosting on Google Code

..so USB suffers from jitter apparently, Onkyo has made it clear on their website that they found ways to overcome this problem, though..

the 0404USB apparently works in a diferent USB mode to overcome this issue as well...but there's jitter associated w/ S/PDIF, right?

and you can't do 24/96 w/ the windows stock USB audio drivers I think?

Ideally, I'd like to use something like the RSA Predator over USB...because Reclock cannot work in S/PDIF w/o reencoding to AC3.

I think I'm looking for something that simply does not exist

I want a device that's not on S/PDIF, that has very low jitter(the STX>ST improvement is quite drastic soundstage-wise, 1ppm FTW

), a quality DAC/HP amp....and 100% bit-matched/bit-perfect KS/WASAPI drivers up to 24/96.
I'll send my letter to santa claus now

 

[Bredin]

Just ordered a new sound card and something came to my mind. Will the RCA connector of the COAX cable fit the S/PSIF out from the card XD. Chord Indigo got some thick acrylic plastic on the connector, furutechs are just thick by nature and so on XD.

 

[linuxworks]

let's say that I'm bored to death w/ crappy soundcard drivers...that are unstable(C-Media) and constantly resample everything...

if I were to use these 100% bit-perfect GPL drivers on a compatible card: cmediadrivers - Project Hosting on Google Code

cmedia chips do not resample. I've been using the famous cmi8738 series for over 10 years now (well over that). they don't resample and ARE bit perfect.

that google code location is key; that driver is great and I highly recommend it. if you are forced into windows.

if you use linux, the driver (again, from 10+ yrs ago) has always been stable and perfect.

the hardware is just fine. user contrib'd software is just fine. the company can't write drivers to save their lives, though

Quote:

..so USB suffers from jitter apparently, Onkyo has made it clear on their website that they found ways to overcome this problem, though..

onkyo? really? overcome it?

Quote:

the 0404USB apparently works in a diferent USB mode to overcome this issue as well...but there's jitter associated w/ S/PDIF, right?

that's emu not onkyo.

Quote:

and you can't do 24/96 w/ the windows stock USB audio drivers I think?

on usb 2.0 you can (in theory). its usb1.1 that had a speed limit. not the stack, the hardware itself.

 

[leeperry]

yeah, these drivers look like the pinnacle of PC audio! too bad that's only useful in S/PDIF...as these cards usually have crappy DAC specs? and I can't use S/PDIF in Reclock, as it requires to reencode in AC3 and I'd rather avoid that

so I'm stuck either w/ internal soundcards(such as the Asus Xonar) and their lousy resampling/buggy/unstable drivers....or USB? but I need some proper USB implementation w/ very low jitter.

Onkyo yes: Google Translate

Quote:

Internal crystal clock synchronization to reduce jitter values Conventional USB audio noise caused by the gap in time for signal transmission to synchronize the computer clock, jitter (fluctuation of the waveform) was 10 times the normal number of audio equipment value. SE-U33GXV greatly reduces the jitter value in the internal body clock with one crystal to provide a sound comparable to high-quality audio equipment.

I even asked pricejapan to add it to their website : PriceJapan.com

it's a baby Wavio USB, that doesn't require an external PSU and w/ an improved jitter apparently.

ah well, I'll keep whining to Asus to release bit-perfect drivers...they promised me to do so, maybe before next year?

and I want bit-matched drivers that forbid SRC, I loved the Echo Digital cards for this...unlocked hardware samplerate and no SRC whatsoever :



but well, even M-Audio forbids SRC in ASIO...so what I'm asking for is not undoable, it's just that the market leaders/joesixpacks don't give a damn

 

[leeperry]

the Onkyo soundcards use the stock VIA drivers, anyone knows if they are bit-matched in ASIO/KS/WASAPI?

this USB soundcard + a serious HP amp might be my golden ticket

PS: http://www.google.com/search?hl=en&s...&aq=f&oq=&aqi=

apparently the Envy24 generic drivers are indeed bit-matched...US $200 shipped sounds like a sweet deal

 

[slytown]

Quote:

Originally Posted by gabrielo /img/forum/go_quote.gif If you got enough free USB ports (not HUB) get USB only (assuming it is cheaper) There is no real-world benefit to having Firewire HD vs USB when you deal with music. In theory Firewire is a faster interface, but in practice the bottleneck is the HD speed. And anyway music bitrate is so low any interface is an overkill

Thanks man. I ended up getting the Lacie USB Little Desk. Another Dj recommended it to me. Thankfully the new MBPs have two USB ports, so I'm good. I'll probably use one port for the Lacie or my iPod and the other to connect my hub.

Would it be ok connecting my Compass to the USB hub, or would I need to do a direct connect to my MBP?

 

[Patu]

Any comments on ESI Juli@ vs. cheap sound card + cmediadrivers? Do you think there can be some improvement on noticeably more expensive Juli@s optical output vs. some $15 sound card + cmediadrivers combo? Can there be better jitter reduction with Juli@ or something else?

 

[joe_cool]

Sadly, the optical output from the excellent Juli@ is at a dis-advantage for audiophile use because of the required double conversion from electrical to optical at the PC and then back to electrical at the DAC. It is quite good enough for AC3 and DTS, but is typically worse in terms of jitter than a coax connection.

As far as cmedia cards you may need to use the open-source drivers to get audiophile grade stability and accuracy. Jitter will again be a concern as not all cards use quality clocks and SPDIF transmitters. Also look for the RCA or BNC coax output.

 

[tosehee]

Quote:

Originally Posted by joe_cool /img/forum/go_quote.gif Sadly, the optical output from the excellent Juli@ is at a dis-advantage for audiophile use because of the required double conversion from electrical to optical at the PC and then back to electrical at the DAC. It is quite good enough for AC3 and DTS, but is typically worse in terms of jitter than a coax connection. As far as cmedia cards you may need to use the open-source drivers to get audiophile grade stability and accuracy. Jitter will again be a concern as not all cards use quality clocks and SPDIF transmitters. Also look for the RCA or BNC coax output.

How much is this concern? I'd liketo Linux_Works opinion on this if possible.

 

[leeperry]

I love how nothing is ever designed the way it should

what we need is a 1ppm card w/ coax S/PDIF and a CMI chipset compatible w/ these GPL drivers...so that does not exist?

 

[Patu]

Quote:

Originally Posted by joe_cool /img/forum/go_quote.gif Sadly, the optical output from the excellent Juli@ is at a dis-advantage for audiophile use because of the required double conversion from electrical to optical at the PC and then back to electrical at the DAC. It is quite good enough for AC3 and DTS, but is typically worse in terms of jitter than a coax connection. As far as cmedia cards you may need to use the open-source drivers to get audiophile grade stability and accuracy. Jitter will again be a concern as not all cards use quality clocks and SPDIF transmitters. Also look for the RCA or BNC coax output.

Thanks for the input. I do have Trust SC5200 card (cmedia compatible card) also and I plan to test it now when my Bel Canto DAC3 arrived. It should have some kind of isolation in the coax input so I shouldn't have to worry too much about the electric connection.

I'm also interested about linuxworks comments.

 

[linuxworks]

Quote:

Originally Posted by leeperry /img/forum/go_quote.gif what we need is a 1ppm card w/ coax S/PDIF and a CMI chipset compatible w/ these GPL drivers...so that does not exist?

coax spdif is never native; what is native is 5v ttl.

you can easily convert that to coax, if you really dislike toslink. its not a chipset issue.

I have an older cmi8738 (just tried it a few days ago) that has coax-out on the back tab (pci card) and a dongle for opto in/out on a 2nd metal tab.

I recently just tried a VIA onboard spdif port (on my mobo) and that seems every bit as bit-perfect as the 8738 is. its a VIA EPIA LN board (mini-itx).

 

[linuxworks]

Quote:

Originally Posted by joe_cool /img/forum/go_quote.gif Sadly, the optical output from the excellent Juli@ is at a dis-advantage for audiophile use because of the required double conversion from electrical to optical at the PC and then back to electrical at the DAC. It is quite good enough for AC3 and DTS, but is typically worse in terms of jitter than a coax connection.

electrical to optical and back is not a problem, in my experience

waveshape suffers but dacs don't NEED square waves in order to pull out clock and data (my experience, anyway).

the diff is academic and not anything that a self respecting dac will care about.

Quote:

As far as cmedia cards you may need to use the open-source drivers to get audiophile grade stability and accuracy. Jitter will again be a concern as not all cards use quality clocks and SPDIF transmitters. Also look for the RCA or BNC coax output.

once you send data to the card, I'm not sure the host/cpu is really all that active. given that, I'm not sure 'drivers' matter for jitter or so-called accuracy. either you are sending (at the card level) what the host tells you or you are not.

layers ABOVE the driver might be bit-scaling (vol control) or resampling (usually windows, since it has a built-in 'pref' for 48k) but the driver should not. the 'google code' driver does not and the linux driver never did.

drivers are not 'audiophile'. at the driver level, things work or they don't. its not a 'higher precision clock' via driver or anything like that. things have stopped being software-based timers for a LONG time, now

hosts don't monkey around when a hardware device has its clock. it sends dma buffers to it, the pci bridge does the transfer and data squirts out. aint no 'audiophile' about it; it just plain works.

 

[leeperry]

Quote:

Originally Posted by linuxworks /img/forum/go_quote.gif I recently just tried a VIA onboard spdif port (on my mobo) and that seems every bit as bit-perfect as the 8738 is. its a VIA EPIA LN board (mini-itx).

well there's a very simple test to make :
-play a 16/44.1 file in DS
-then simultaneously play a 24/96 file in ASIO in foobar

M-Audio/Echo Digital cards will give you an error msg in their ASIO driver stating that they couldn't set the right hardware samplerate.

the other way around works, ASIO sets 24/96, then KMixer resamples 16/44.1 to 24/96...the way it's meant to be!

I've tried this on several consumer soundcards and they all failed

another test is w/ a S/PDIF receiver, each time you play a file w/ a different samplerate it should reflect in the receiver...meaning no SRC takes place. HDCD/DTS-CD is also unforgiving to check bit-perfectness.

I've been told that Realtek drivers are 100% bit-matched/bit-perfect in WASAPI exclusive mode on Vista...so it's not that hard! I don't understand why all the audio drivers don't forbid SRC...KMixer is what should be resampling, nothing else! and especially not in ASIO/KS/WASAPI

but I guess the Realtek's must be a jitter party over S/PDIF

 

[Solitary1]

Quote:

Originally Posted by FallenAngel /img/forum/go_quote.gif Reproduce recorded media as closely as possible to original "what"? Original as in how it would sound live, or how it sounds in the booth? What about mixed recordings where there are barely even original instruments being used? So then we need to know whether the purpose is to reproduce what was "recorded" vs how it actually "sounds". Personally, I try listen to music and configure my system to sound as pleasant to me as possible, not necessarily how it was recorded or processed. Studio equipment and audiophile equipment is quite different, as I'm sure you know. Studio equipment is generally designed to accent everything and analyse sound, not enjoy music as audiophile playback equipment is. Honestly, this is getting a bit off topic, we can start a different thread and discuss this there if you'd like.

You're right, and I was just tired of talking about it. We are just going to have to agree to disagree. No hard feelings at all on this end. We'll engage again. Take care.