An experiment in digital audio

Yesterday I did an experiment that some of you might find interesting. I had been noticing bright, shrill sound from downloaded audio files and, in order to eliminate the incompetence of the person doing the original encoding, I dug out some DVD players from storage that are capable of making copies of SACD discs.

First attempt was with my Oppo BDP-103D and I used Sting's Sacred Love SACD as the original material. I copied an ISO of the whole disc via wireless LAN to my PC using SACDExtract, then converted to DFF stereo with ISO2DSD. Playback chain is PC -> Wi-fi -> Oppo UDP-103 -> Victor JM-S7 VFET -> Jordan J6T linear arrays.

Result: 5/10 Dry, bright, flat soundstage , no air.


Second attempt employed the Oppo BDP-105D using the same procedure.

Result: 7/10 Dry, less bright, more detail, no air.


Third attempt was using the BDP-105D ISO but converted to multi-channel.

Result: 8.5/10 More detail again, starting to sound normal.


Final test was to put the original disc in my UDP-2013 and play it.

Result: Huge soundstage, sweet, open, details for miles.


Conclusion: If you're serious about listening to music, don't download stuff off the internet. I just performed a BEST CASE assessment using the highest quality player capable of ripping SACD and lost 15% in the process. The average person uses equipment at BDP-103 level or lower quality, which just sucks the life out of the music completely.

And if you don't believe that digital information is changed in the process of copying it, I challenge you to make a copy of an HDCD disc and see if it lights the HDCD indicator on your player. I tried three discs downloaded off torrents, and none of them were HDCD any longer.
 
Thanks for the info. Ripping SACDs looks tricky.

I wonder if you've done a similar trial for CDs eg ripped one of your CDs in a PC with EAC and compared streaming vs USB vs disc playback with the Oppo?

I've now completed that comparison as well. Didn't use HDCD since it turns out the HDCD discs I downloaded do contain that information after all; my BDP-105D seems to be a bit flakey on detecting them, but with a power off reset they all worked.

So instead I used Checkfield's "A View From The Edge" which is an excellent recording. I found the results of ripping that to HDD and replaying across the network to be very similar, maybe 97-98% the same as the original. There was a little more space around individual instruments with the CD but it was a hard call to say they were different, and it would take a very well-trained ear to pick it in a double-blind test.

I guess that tells us that the extra data available with hi-res formats needs to be handled with extra care, but not so much the standard Redbook stuff. I may do a second test using a couple of the Japanese SHM pressings if I can find any in my collection. They're reputedly better quality than a standard CD, despite being 16/44.1 format.
 
Last edited: