Can harmonic distortion in sound card capture be improved?

QuadraphonicQuad

Help Support QuadraphonicQuad:

This site may earn a commission from merchant affiliate links, including eBay, Amazon, and others.

jdmack

1K Club - QQ Shooting Star
QQ Supporter
Joined
Feb 27, 2005
Messages
1,256
Location
Maryland
Tonight, I was re-calibrating my turntable using a record with sine wave test tones. I usually don't record this, but tonight I decided to. And what I found is that recordings of sine waves made with my Sound Blaster X-FI Titanium card have a whole bunch of harmonics added. A 1K tone sounds fine when I'm monitoring it, but once recorded, it now has harmonics that I can hear at 2K 4K 8K etc, decreasing in strength as it goes up the harmonic sequence (the spectral frequency display is useful here). For comparrison purposes, I tried recording the sine waves through the RealTek audio card built into my motherboard and got the same result.

Is there something that can be done for this? I'd consider buying another sound card, but the THD spec for the Sound Blaster is 0.004%. Apparently, this spec isn't of much value so I don't know how to be sure I'm buying a sound card that won't have the same problem.

J. D.
 
If the distortion is audible, it sounds like your recorded sine wave is getting / has gotten clipped somewhere. Have you tried a slightly lower record level?

You could try loading the recorded sound file into an audio editor like Audacity, which can also do a spectrum analysis. If the distortion is present in the recorded file, then too-high record level is probably the culprit. If the file is clean, then there's something clipping in your playback path.

-- Jim
 
First, check out who's really to blame.
1) using Audition or something else, create a set of sine wave tones at the frequencies you want to test - 500hz, 1khz, 2khz... whatever. Sample rate and bit depth should be set for the desired target, es 44100/16/2 for audio cd, or higher for dvd a/v.
2) burn these files as audio track on a cd or dvd
3) record the output of the cd player with the titanium and see if the harmonics are still present, or how they varies.
4) Let us know the result.
 
I'd rather "blame" the stylus...

I did the same with a COmmand 1960 LP with sine waves and got ALL kinds of harmonics from the tones...

Maybe it's just why vinyl sounds so sweet, even LPs sourced from 44/16 sources have harmonics that go higher than 20 K (on my system)
 
I'd rather "blame" the stylus...

I did the same with a COmmand 1960 LP with sine waves and got ALL kinds of harmonics from the tones...

Maybe it's just why vinyl sounds so sweet, even LPs sourced from 44/16 sources have harmonics that go higher than 20 K (on my system)

So vinyl sounds good because the stylus introduces distortion that you can't hear anyway? Interesting.
 
(shaking my head...)
NO nononononono
it is NOT distortion , they are HARMONICS that you can't hear but add a richness to the signal...
why do you think stuff is now recorded in 96/24????

(still shaking my head)
 
OK, here's what I found out from the test suggested by Winopener. A sine wave created in Adobe Audition was transfered to my MP3 player (what's a CD player?). I unplugged the cable from my phono preamp output and plugged it into the output of the MP3 player. I recorded a few seconds of the sine wave in Adobe Audition, and then looked at the spectral frequency display. No harmonics at all! What puzzles me is that when I am recording the sine wave from the record (the same Command Record that kap'n crunch has), I am monitoring the record through my sound card, and it sounds like a pure sine wave. It's only on playback of what I recorded that I instantly hear the whistling caused by all the added harmonics.

As a second test, I tried it with a different turntable, and got the same result. Turntable #1 has an AT 440MLA stylus. Turntable #2 has a cheap conical stylus (I use this turntable for playing old styrene 45s). The phono preamps are built into my Tascam M-30 mixer. If the levels on the mixer are set to 0 DB as shown on the meters, then the level going into Audition is about -10 db, so I don't think I'm overloading anything.

Any way, it's not the sound card, as some of you suggested, so I'm happy for that.
 
Thank you for posting the results. That's why i started my post hinting "who's really to blame".
Kap is correct, there are harmonic distortion that are inherent to the vinyl playback and the real offender is the stylus: it's just mechanical resonance and no matter what damping you have you can't eliminate them at all.
The fact that on recording you saw only the pure sine and not the harmonic can ba traced back at the speed of your graphic card / drivers etc; doing a realtime display is very different to a freezed-out one when you load the entire wave and then magnify whatever you want.

If you're looking for more technical fun, try this:
1) take a late 60s/early 70s album, with a decent hissing background (a original pressing of Santana Abraxas is perfect for that)
2) record a snippet of the album
3) using a equalizer, do a +30db (yes + thirty) from 10KHz to the end; if you're dealing with standard cd parameters, 22.050.
4) DON'T play it, you're already 100% sure it will sound like crap and it does, but now turn on the Spectral Analysis and you can see clearly where the original master tape hiss ends. That's the real frequency response of that record as musically intended by the master + record cutter, and for late60-early70 you're in the 15KHz range. All the stuff above the hiss are just mechanical resonances of your system.

So much for hi-rez vinyl and digital bashing typical of today trendy kids.... :violin
 
Back
Top