Unmasking Lossy Compression Artifacts When Upmixing

QuadraphonicQuad

Help Support QuadraphonicQuad:

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

Sonik Wiz

đź‘‚ 500 MPH EARS đź‘‚
QQ Supporter
Joined
May 30, 2005
Messages
5,303
Location
Kansas City
Much has been said elsewhere about the need for quality sources when doing upmixing. Sure about anything from low rate MP3 to high bit/sample rate can go through a a PC upmix program or Surround Master, etc & give some of surround sound field. I've re-discovered a dts white paper on my HDD that explains quite well why this is, or at least another aspect. Like a lot of white papers it is ther to show dts Neural Upmix superiority it has some good insights as well.

Me thinks the paper is long gone from the dts site so you can grab it here on Dropbox.
 
Nice find SW.
I usually add the DTS Neural:X upmix to standard Dolby D 5.1 TV broadcasts and am continually amazed at the discrete elements that come from the rear surrounds. So clean, discrete, and natural, just sonic magic to my ears.
 
Thanks for the link, Mr. Wiz.

Jamoke, I too am surprised at how great these things can sound sometimes. Some of my favorite upmixes are because they just sound gorgeous as much as any multichannel trickery on display.

I think perhaps the simple act of spreading a given mix out over more speakers by some means or another, is enough to significantly reduce intermodulation effects, but that's just me talking out my bu- <ahem> button-down neck-hole.
 
Last edited:
I think perhaps the simple act of spreading a given mix out over more speakers by some means or another, is enough to significantly reduce intermodulation effects, but that's just me talking out my bu- <ahem> button-down neck-hole.

What a charmin reply!
I think the take away from that paper was that whether you're talking IM distortion or other artifacts is that they are revealed rather than masked when spreading over more speakers. I did not know that of lossy codecs gave less bit depth to the L-R component. That ex[plains a lot.

Sometimes this unmasking results in a benefit. I have done a lot of upmixing with video. I demux the audio / video from a video source & work the audio around in AA 3. I export as high bit rate dts. Then using a DVD authoring app I mux them back again with menu & chapters. On a Chisato Moritaka live performance on one song I kept hearing a staccato clicking bongo like noise I had never noticed in the stereo mix. Once upon a time I built an SQ decoder with such a pitiful power supply it bottomed out every bass beat. That's what this noise remained me of.I proceeded & eventually made the completed project. When I watched that song this time in surround there it was! That noise! And now I could see it. Conga drums! It was buried in the stereo mix & unmasked in surround & very happily coming out of L front. And it seems so much more appropriate than the stereo mix cuz if you see an instrument on stage you expect to hear it as well. IMO.
 
Back
Top