At the moment I am running a Turtle Beach Santa Cruz sound card into a Cambridge Audio DAC3 digital to analogue converter, and into my hifi system. However, I am very tempted to get a Terratec EWX24/96 soundcard after reading reviews praising its excellent sound quality which is apparently on par with mid to high end CD players. My mate has one, and it does indeed sound very good.

The question is, with this being a 24-bit, 96K soundcard, does this mean a program like foobar will be able to upsample to a higher resolution, in a similar fashion to upsampling DACs? And if it does, does it sound any better than 16-bit/44K?

Cheers

Yes, you can upsample. No, it will not sound better.

You can't create something out of nothing.

Edit: But it is a damn fine card. I own one myself.

We may try...

Frequency response of the EWX for example is more linear in the highs when testing 24/96 mode. So, theoretically, using high quality settings, software upsampling could do some good.

However, even if its measurable, I think playback in 16/44 is already so good that you can't tell the difference. Also, upsampling is going to cost you cpu-power.

But testing it would still be nice...

Here's a quick test. Note that I did it in loopback mode, so any differences can be caused by either the playback being better in 24/96, or the recording being better in 24/96 - thus, the test doesn't tell you anything, really

RMAA5.0 sav-files are here: edit: oops, fixed link
There are four:
a loopback test in standard 16/44 mode on my DMX 6fire
a reference test 16/44 (written to hdd then read)
a reference test 24/96 (written to hdd as 16/44, then upsampled in CEP2.00)
a loopback using the upsampled wav (playback in CEP, recording in RMAA)

The upsampling in CEP2.00 used a quality factor of 400 and enabled pre/postfiltering.

Patsoe, forgive my obvious "newbieness", but what do I do with those files and what are they?

Hopefully this weekend, I am going to have a loan of a Terratec EWX24/96 soundcard and am going to put it directly head to head with a Pioneer DV-350 DVD transport along with the Cambridge DAC3, and see how it fairs. The DAC3 is (IMO) pretty good for the money, and so if this soundcard sounds better, then in my eyes it'll truly be an amazing piece of kit, especially given the "downers" most audiophiles have on using a soundcard (and PC) as a hifi source.

If anyones interested I'll certainly post my findings.

You can see graphs from these files using RMAA, to be found here: http://audio.rightmark.org - click on 'Load (saved) Results'.
I could have produced some graphs, but since 96kHz results can't be put in one graph with 44kHz results, it wouldn't be so informative. So that's why I put the sav files up.

If you are going to do a comparison of the DAC3 and the EWX, it might be interesting to take a quick measurement using RMAA (this software doesn't tell everything, but it's easy to use, and will give an idea of the type of differences you may hear). You can try both EWX output to EWX input and DAC3 to EWX input (where the DAC3 can be driven from EWX digital out). If you like to, you can even put the test samples on CDR and play from the DVD player.

Weird... for me upsampled 96kHz graph (IMD) looks worse and I can't find that peak of -70 dB(!) noise in not-upsampled case...
Probably these results are just lucky. Not to mention large difference between RMS levels, while nothing like that is visible. :/
Additionally, RMAA seems to not detect IMD properly in upsampled case - it looks higher. Probably it's confused by THD. :/

I didn't look at the results, but how do you compare distortion in a 44.1 kHz file with distortion in a 96 kHz file ? It seems that if there is some distortion above 22.05 kHz, it will be missing in the 44.1 kHz file, leading to better results.

If you compare results below 20 kHz and get worse ones for 96 kHz it can mean that there is some intermodulation or aliasing with the ultrasonic noise, or also that the EWX upsampling algorithms (oversampling 44.1 kHz playback) are better than your software upsampling...

Oh, BTW, the results of the Pioneer vs Terratec test would be interesting if the test is done in an ABX way, trying to tell the difference between the two.

I am open to suggestions on how to do this in the ABX way, but I feel it might be difficult, since I will know which one is playing. Actually, I could get my friend to help here - maybe i can wear a blindfold or close my eyes, and get him to switch sources without him telling me which one is which.

What I was going to do however, was just play the same song at the same time, and use the amp to switch over and see if I can tell the difference that way...

But yes, however it works out, I will report back on my findings...

You can always play through the DAC and then record it, or use friend-changes-setup method.

CEPs resampling is inferior to SSRC, even highest quality - I recall some graphs...

So I should use foobar2k upsampling + diskwriter then?

You bet it.

"Resampling to higher sample rate doesn't increase perceived "sound quality",
only resampling to 48000hz avoids issues with certain types of sound hardware."

We're checking whether with this card it avoids some issues.
I think the difference would be quite hard to ABX...

I have now completed my comparison of the Pioneer DV-350 + Cambridge Audio DAC3 v Terratec EWX24/96, and have here a copy of my review which I have wrote for some other forae also....

The PC is currently linked to a Cambridge Audio DAC3, which was a highly regarded, but now somewhat ageing Digital to Analogue Converter. The soundcard being used at present is a Videologic Sonicfury (its actually made by Turtle Beach, called the Santa Cruz, but is renamed for the UK market, and sold under Videologic, now known as Pure Digital). The soundcard resamples internally to 48KHz, and thus is not really an optimum solution, with or without a DAC. Given that however, it only sounds the tiniest bit worse than using a Pioneer DV-350 as transport, it suggests to me that the process in this case does not result in significant degradation of the sound. So, to make comparisons easier, I simply used the DVD player, which was playing the original discs. I used the Terratec to play MPC files that I ripped from said originals. Not fair on the Terratec, some might say, but I cant discern any difference from Q6 MPC files to the original discs. I wanted to ABX test, but as it was impossible to match the volumes up the same, and no attenuators were on hand, this was always going to be mission impossible. So I just auditioned in the normal way, in the comfort of my bedroom.

The accompanying equipment was my Rotel RA-01 stereo integrated amplifier, driving my Mission M51 speakers, and driving Beyer DT880 headphones. The music used was largely electronic, trance, techno, ambient, but there was also some Nine Inch Nails thrown in to cater for rock music.

First, we listened to some of the trance music on the DAC3, just so our memory could be refreshed on to how it sounded. Then in its place, we hooked up the PC with the Terratec.

The first song I played (and a good audition song since I know it like the back of my hand) is Oxygen Feat. Andrea Britton – Am I On Your Mind (12” Original). My first impressions of the Terratec was its outstanding bass performance, something not quite expected given the DAC has a dedicated power supply. If anything, I was expecting the bass to sound a bit lightweight, but not so. This soundcard produces sound with real authority, with very solid, firm, tuneful bass that extended very deep. Theres some nice deep bass and this, and it was well defined and deep, with authority. However, at the same time, it is clean, and does not pollute the rest of the frequencies, as some bass heavy things can do. Maybe its not bass heavy as such, but the bass is definelty there, and with real quality.

The next thing I noticed was the extra detail over the DAC3, this thing is noticeably more detailed, and it adds ambience to the sound. Not only is this thing more detailed, but it is not bright, it is just right, and the sound is much smoother, but this extra refinement does not seem to cost excitement to the sound either. The extra smoothness is particularly beneficial to headphone listening, where it sounds really nice.

Finally, the sound is noticeably more extended at the treble side of things, the net result is it sounds less closed in than the DAC, inviting you into the music a lot more. It isn’t brighter though, it just sounds more open.

So we tried a number of different trance songs, and the conclusion was always the same. The soundcard simply has better clarity, musicality, bass, treble, is more refined, and more smooth.

To make it better still, we then replaced the Van Den Hul The Bay C5 interconnect cable with a solid silver DIY shark interconnect cable, terminated with Eichmann Bullet plugs. This made a further improvement to the sound, further increasing smoothness and detail levels, and the bass was a bit better controlled. The sound was very, very smooth and nice with these two in the system, very musical and bouncy still, but now with the refinement the system has always lacked. My friend described the sound as heavenly and euphoric. Pretty impressive then…

So, to summarise, the DAC3 was good in its day, and is still not bad now, but it is outclassed by the Terratec, which is a potent little card for the cash (£130), and offers a 24-bit DAC and sound quality as good as some of the best midrange CD players in my opinion.

I know where £130 of my cash will be going on payday

Thank you for sharing, but we're not much interested in subjective opinions about sound. It sounds like a review in a hifi magazine...
We try to avoid any general comment on the sound (exept when backuped with blind tests) and focus only on either blind test results or measurments. This is explained here : http://www.hydrogenaudio.org/forums/index....showtopic=11442

OK, fair do's but as I explained, it was not really possible in this instance, but I can see your points....

Cheers

Just one thing about resampling: the resampling at CEP at quality around 400 and over with pre/post filter checked is very high quality, and I believe is transparent by far. At max. quality I don't think it has anything to envy to SSRC resampling, only processing speed.

Does anybody want to setup a blind test of resampling algorithms?
I vote for SBLive as an anchor!

I'd be interested in doing that. Not this week though. But if you keep remembering me, I will

I can give you the old one again.

I just thought I'd mention that Terratec have released updated drivers for the EWX 24/96 on Aug. 29th.