The "Sound" should not be one iota different versus a PCIe x1 or even an "ancient" PCI.  Now - Low Latency scaling might vary - but sound quality in itself won't...

As far as which is better for Latency and CPU Scaling - you'd probably have to try as the disk will need to access that I/O just like the CPU and RAM will.  Probably boils down to which would have the least latency as we know the actual bandwidth requirements are TINY compared to PCIe 2.0's 250MB/s max bandwidth - it's all about the latency)

HI Atom,
The "sound" difference is coming from either RME's Steady Clock technology improving your audio clock. Your converters, cabling or Lynx card where introducing clock jitter that the RME is removing or decreasing. The only way the PCI or PCI-e slots would be effecting thing would be general compatibility issues between your Lynx card and you motherboard. But this would normally only seems as clicks and pops or crashes.

Chris

Not sure what you are hearing here, but before everybody will start shifting all-digital audio cards in search for better sound, the official position is that there is no way the PCIe slot can affect sound quality. THere is also no way there can be a difference between RME and Lynx in terms of "noise", provided the configuration is correct.

Regards
Daniel Fuchs
RME

Do you also subscribe to the theory (cough cough  fryingpan ) that different HD's sound different, and AMD and Intel also sound different?  It is all chunks of data at this level - clocking as in reference to a digital audio stream is not even involved at this level AFAIK (that is handled on the card itself with a crysta, SteadySync, and/or PLL's - certainly not being fed from a MoBo PCIe clock)...  The Audio Clock and the PCIe Bus clock have no correlation (as long as the PCIe clock is getting that data to the card in a timely manner - otherwise dropouts would occur). 

What happens when you raise the PCIe clock to 105MHz?  Does the audio change or speed up?  Nope

If RME takes a similar stance (as they loosely stated above), then I don't think your theory holds any water - if you don't mind me saying so HeadScratch

Now - the fact that the PCIe x16 slots in i5 and i7 platforms talk directly to the CPU might have some performance and low-latency scaling implications - but no effect on actual SQ IMNSHO...

There is a market for Shakti Stones as well

http://www.shakti-innovations.com/audiovideo.htm

Green Markers are also a "must have" tweak

So why doesn't the sound change if you change the PCIe bus frequency?  If the PCIe bus clock and the Audio Wordclock are intrinsically linked - you SHOULD hear some type of difference in the sound - no?  What about if you overclock your FSB/QPI and CPU Clock Speed?  Why don't AMD's sound different from Intels since the "signal path" ( ) is WAY different betwen them?   This part of the data stream is not "linear" - it is chunks of data sent across asynchronus busses.  Once the sound card recieved these chunks - it buffers them slightly - and spits them out to the card's OWN crystal or PLL - the incoming chunks of data will not affect what the card spits out as the 2 clocks don't interact.

Speakers are ANALOG TRANSDUCERS - AD/DA's are also ANALOG to an extent and actually QUANTIZE the ANALOG audio into DIGITAL bits.  Do you think the original wordclock timing is somehow transmitted along with the digital packets through the CPU and RAM?  Nope - the original digital wordclock signal has NO AFFILIATION with the digital packets or "chunks" that our CPU's process.  Once the CPU processes these packets - it sends them back through an asynchronus bus to the PCI(e) audio card.  THEN the audio card re-clocks these packets using ITS' OWN CLOCK OR PLL (not the CPU or PCIe bus clock) to form a Linear PCM type stream.  Now that the stream is actually Linear PCM - the clock will absolutely matter - but ONLY at the DAC (where the actual conversion happens from PCM to Analog).  The only way I know of a PCIe bus affecting Audio SQ is if the PCI bus can't keep up with the bandwidth (dropouts) or has unacceptable interrupt requests that also cause dropout.

I think I'll trust the people who design and manufacture these fine products over someone's "Golden Ears" any day.  PCIe vs PCIe x16 vs PCI will have no difference to the PCM digital signal - and even if it DID (which it doesn't) - the DAC's clock and ability to deal with jitter is what really matters (or more importantly - the ADC's clock is most crucial in the capturing process).  The original thread topic was about PCIe x1 vs PCIe x16 performance and reliability, not SQ

Don't waste your keyboard! I already tell you that page of theory don't interest me. If I found that a piano sound better with my RME on pci express than a linx on a pci, you can write 100000 word to tell me is faulse, I will kept my RME DeadHorse. Have you made the experiments I've made? No I'm sure so don't be so assumptive...
There is a lot of peolple and journalist who hear differences beetween digital cable, Can you write me why? It's a nonsense: It's digital: 0 and 1...
rotfl too

And this is straight from RME's mouth   I agree - this is a "Fact" based forum - take the fairy dust and unicorn tears to Gearslutz or to any Audiophile forum in the world - they would love to tell you that you're right - and give you a thousand reasons why (all of which are not rooted in anything resembling an accurate reality) .  They will also tell you why Shakti Stones have a substantial impact on your sound rotfl