Oversampling rate in J River

I have a W4S Dac 2 that I am using with my laptop. I am having a ball downloading hi res music and ripping CDs. Sounds great, lots of fun and boy do I love the convenience. Also have solid state Mac and a pair of B&W 803Ss.

I have been fooling around with different DSP settings and have found virtually very little change or change that I care for...Until this weekend when I changed the oversampling from none to 192,000. Wow. Quantum leap. More presence, detail and what many in these forums would call involvement.

What happened? is this normal? PC audio is fairly new to me. Looking for advice/input from those who know.
This is upsampling. Often the D/A chip behaves better with higher sample-rates, particularly when the digital filtering is not adjustable. Digital filtering is one of the things that causes digital audio to be fatigueing, along with jitter. Unfortunately, upsamplers are usually not as good as the real hi-res files, particularly hardware upsamplers like this one.

BTW, There are other ways to make your W4S sound a LOT better without upsampling.

Steve N.
Empirical Audio

To better analyze your results, could you tell me what output method you're using in Media Center? As a rule of thumb, the DSPs in Media center should only be used if you've got a less-than-ideal situation, such as a home theater. A lot of the options are there to accommodate SW cut-offs, or legacy receivers that aren't capable of decoding higher resolutions. You're W4S DAC is capable of playing native 24/192 files, so you shouldn't need any of the DSPs. Also ensure that your Windows sound device settings are correct. Make sure Media Center can take exclusive control of the output device, etc.

To more directly address the question, you theoretically shouldn't be hearing a difference when using up-sampling or oversampling. For example: If you're playing a Redbook CD rip (16/44,1) and up-sampling it to 24/192, you're only padding the output with additional 0's. No information is added, but the sample size is just much larger. Depending on the output method, you may be experiencing some other interactions with the Windows kmixer that are doing more harm than good.

"There are other ways to make your W4S sound a LOT better without upsampling."

Pray, tell . . .


I used the settings published by W4S for the DAC for J River. I am using Media center 16 and had it set on no oversampling. I am using Kernel Streaming, as recommended by W4S. I had no other DSP settings in use.

of course, I have the W4S driver selected in the setup which I thought bypassed all the sound drivers in Windows Vista. Have I missed something?


If there are ways with my current setup to improve the sound, I am all ears and would be very appreciative. If it requires me buying an Off Ramp,then not so much.
Luckily Steve is always forthcoming about being in the business, and of course he'll tell you that his product will help; how else would he stay in busniess?

I read through W4S's J River setup guide and they hit all the important spots. Your setup is solid, and I'm sure you'll get plenty of enjoyment from the DAC as-is. More to Steve's point: sometimes the "all-in-one" solution isn't the best. For PC audio this becomes especially true, given everything that happens from the moment you press Play on the file, to the moment it reaches the analog outputs of your DAC. I cuurently use a modified M2Tech Hiface that runs "off-grid" on battery power. You would not believe the improvements this made in my system. To each their own, and I'm never one to push people to buy more gear; I'm just telling you my findings.

Just remember: You can NEVER recreate the music that has been destroyed during the recording, downconversion, or complression process. You may hear a "difference" when upsampling, but you're not really improving the source material.

Since your hapiness is really all that should be important, I say upsample away, you're not hurting anything.
Thanks. No knock on Steve's product. I'm sure it has added value.

Thanks to everyone. Will continue the journey.
One thing to understand is that the USB interface, and in particular, the master clocks, is the MOST IMPORTANT thing in a USB digital computer audio system. More important than the DAC itself. More important than the sample-rate that you are playing back.

To prove this to you here are some anecdotes:


Latest issue of TAS, Steven Stones review of the iDAC

Steve N.
Empirical Audio

I wonder if you've seen anything about the SOtM tX-USB and SOtM In-Line SATA Power Noise Filters out of South Korea. Chris of Computer Audiophile recently built a new audio server with these two components and they seem to address a lot of what's plaguing consumer-grade PC audio systems. The USB card in particular uses it's own molex power, uses a capacitor bank for filtering and then implements their own 3.3V and 5V rectification. It also includes a switch to turn the power off and only send data (I know some DACs require power); this is pretty revolutionary stuff for us PC audio freaks.

The best thing about it, the is an entirely seperate USB bus (because it's acutally a PCI bus) and won't be shared by other devices.

This coupled with a HiFace or Off-Ramp would take a system to the next level...

Sorry to get OT, Steve just mentioned something that sparked my memory.
" One thing to understand is that the USB interface, and in particular, the master clocks, is the MOST IMPORTANT thing in a USB digital computer audio system"

I believe that proper word instead of "master clocks" would be the "real life jitter". The master clock's jitter shows the potential, however, when you build your server or DAC (or buy elsewhere) number of factors will degrade its actual performance and it depends solely on the builder to isolate and "annihilate" them. Its not "off-shelf" stuff.

Also, I must mention output stage... for generations, most modifiers here made their living on upgarding output stages and every one with ears know the improvments of "better" output stage.
Of course what we are talking about here is jitter and all of the contributors to jitter. I have written several white-papers on this:




The point I'm making is that if the master clock is not REALLY GOOD, all other bets are off. It does not matter how careful you are downstream. The damage is already done. And this damage makes all of the fancy technology that you put into the DAC worthless.

Steve N.
Empirical Audio