Is 48Khz Upconverting Ruining The Sound?
Mar 7, 2007 at 5:18 PM Thread Starter Post #1 of 16

classicalguy

500+ Head-Fier
Joined
May 9, 2004
Posts
536
Likes
12
I have two computer sound systems in my house. The one downstairs has an M-audio transit (proprietary ASIO) ---> T-Amp ---> bookshelf speakers. The one upstairs has an Xitel Pro (ASIO4All) ---> T-Amp ---> bookshelf speakers. Both USB sound cards.

I one downstairs (Transit) sounded great. The one upstairs (Xitel) sounded bad. I thought it was the speakers, so I swapped them. The downstairs still sounds good, while the upstairs still sounds bad. So it was not the speakers. The only differences between the two are the USB sound cards and the room (which I don't think would make this kind of a difference since in both cases I'm sitting close to the speakers, in similar locations on the desk).

I noticed that the Xitel converts to 48Khz from the native 44.1. I suspect the Transit doesn't, and that is the source of the sound problems. It sounds as though the life of the music has been sucked out. It dulls the notes - they are all there, but lack the clean impact that makes music sound dynamic. It's like you put a blanket over the speakers. I suspect that the sound problem is caused by the conversion to 48Khz. The difference is not subtle to my ears.

When I get some time, I'm going to hook a good CD into the T-Amp in the bad sounding system to make sure I don't have a bad cable and to isolate the problem to the source.

Do you think I'm right about the 44Khz upconverting causing the problem? Is the only cure a new sound card? How can I find out whether available usb sound cards upconvert?
 
Mar 7, 2007 at 7:04 PM Post #2 of 16

mriguy2

New Head-Fier
Joined
Nov 16, 2006
Posts
22
Likes
0
Yes... here's my story. I use an Apogee Mini-Dac, connected to a Macintosh runing iTunes. It works fantastically well, with most songs encoded from CD using AppleLossless encoding.

One day I noticed that the Min-Dac was indicating a 48khz sampling rate. I went to the MIDI Contol panel and set things back to 44khz sampling. The Mini-Dac changed rates accordingly and sounded considerably better. I went back and forth several times and confirmed that the sampling rate change is not a good thing... so, I agree with your observation.

PS - If you want to control iTunes from your listening chair, check out a program called "Clicker" from Salling Software at http://www.salling.com/
It functions perfectly and is a must have for all of us who are serving thier music from a Mac or PC, but the computer is located away from the listening seat(s). The guy who writes/owns this software is very helpfull, responsive and pleasant.
 
Mar 8, 2007 at 1:08 AM Post #3 of 16

Sycraft

100+ Head-Fier
Joined
Jul 6, 2004
Posts
440
Likes
11
It can. If done well resampling introduces no audible distortion and thus doesn't hurt sound quality. However it done poorly, and lots of things do it poorly, it does introduce distortion that's easy to hear and easy to verify on a scope. When possible no resampling is better as there's less places for something to go wrong.
 
Mar 8, 2007 at 9:13 AM Post #5 of 16

Calroth

100+ Head-Fier
Joined
Apr 16, 2006
Posts
179
Likes
0
It's true that sound cards locked to 48KHz often sound bad, but I wouldn't say that 48KHz resampling causes the problem, since it can be done decently. As said above, try a high-quality software resampler and see if the sound improves.

I think sound cards locked to 48KHz and sound cards that sound bad are two symptoms of the same root cause: the manufacturer is cutting corners to save them (and you) money. Of course, I don't know whether the Xitel Pro is such a cheap sound card - I've never heard of it (which is telling in itself, I guess).
 
Mar 8, 2007 at 8:47 PM Post #6 of 16

classicalguy

500+ Head-Fier
Joined
May 9, 2004
Posts
536
Likes
12
I'd prefer to call it "inexpensive." It's approximately the same price as the M-Audio transit, which sounds much better to me. I think there is some improvement using the resampler in foobar, but it still does not sound as good as the transit by quite a bit.
 
Mar 8, 2007 at 10:27 PM Post #7 of 16

EnOYiN

Headphoneus Supremus
Joined
Jan 13, 2007
Posts
2,269
Likes
18
Software resampling can actually make a signal better than it originally was when using interpolation. However this goes to waste when you are using a DAC which isn't that good.

The most important thing is the DAC. A good DAC will use interpolation instead of just giving you the signal like it is.

Maybe I can draw a small picture which will explain this better than I can with words. I haven't got the time right now, so I'll do it tomorrow or the day after.
 
Mar 8, 2007 at 11:48 PM Post #8 of 16

CSMR

1000+ Head-Fier
Joined
Mar 24, 2004
Posts
1,162
Likes
11
Quote:

Originally Posted by classicalguy /img/forum/go_quote.gif
Do you think I'm right about the 44Khz upconverting causing the problem? Is the only cure a new sound card? How can I find out whether available usb sound cards upconvert?


You can check by upconverting in foobar and seeing if that improves things. But you may as well use get decent sound card in any case.
 
Mar 9, 2007 at 2:13 PM Post #9 of 16

classicalguy

500+ Head-Fier
Joined
May 9, 2004
Posts
536
Likes
12
Foobar helps a bit, but there is no bass. I'm thinking seriously about dropping $100 on a better sound card. E-MU 0202, M-Audio Revolution, silverstone, m-audio transit? Which would be best?
 
Mar 9, 2007 at 9:50 PM Post #11 of 16

lowmagnet

100+ Head-Fier
Joined
Mar 20, 2006
Posts
342
Likes
13
Quote:

Originally Posted by mriguy2 /img/forum/go_quote.gif
Yes... here's my story. I use an Apogee Mini-Dac, connected to a Macintosh runing iTunes. It works fantastically well, with most songs encoded from CD using AppleLossless encoding.


The comment in the Benchmark DAC1 manual is basically this:

a) run at full volume because OS X's mixer causes rounding errors if you don't
b) run at your source's rate because OS X's rate conversion causes rounding errors if you don't.

I've got my Midi panel locked at 24 (no choice there) and 44.1KHz rate. 48 does sound a little off, if you're listening to something critically.

I also miss sound check. But, you guessed it, it causes rounding errors.
 
Mar 15, 2007 at 1:11 AM Post #12 of 16

classicalguy

500+ Head-Fier
Joined
May 9, 2004
Posts
536
Likes
12
I don't know if anyone is interested, but I figured out today what was ruining the sound in my desktop setup. It was the rat shack cable (stereo RCA --> minijack) from the usb card to the t-amp. I changed to some old audioquest cables that I had around into a minijack converter, and it sounds much better.

I'm not a member of the $50,000 power cable club, but bad interconnect cables can hurt the sound - badly.
 
Mar 15, 2007 at 12:58 PM Post #13 of 16

lowmagnet

100+ Head-Fier
Joined
Mar 20, 2006
Posts
342
Likes
13
Quote:

Originally Posted by classicalguy /img/forum/go_quote.gif
It was the rat shack cable (stereo RCA --> minijack) from the usb card to the t-amp. I changed to some old audioquest cables that I had around into a minijack converter, and it sounds much better.


Rat shack cables have no shielding. Look how thin most of them are.

Quote:

Originally Posted by classicalguy /img/forum/go_quote.gif
I'm not a member of the $50,000 power cable club, but bad interconnect cables can hurt the sound - badly.


Yeah, bad cables hurt, good cables work, and expensive cables hurt. $50,000 power cables break the budget and don't change a thing.
 
Mar 16, 2007 at 3:15 AM Post #14 of 16

classicalguy

500+ Head-Fier
Joined
May 9, 2004
Posts
536
Likes
12
I believe the audioquests have no shielding at all, yet sound pretty good. I think there is more to it than just shielding.

Better wire of appropriate thickness is my guess. Rat shack cable is pretty thin and flexible. There can't be too much wire in there. Still, for a short run, I didn't expect it to make that much difference.
 

Users who are viewing this thread

Top