High resolution (Hi-Res) audio has become an increasingly important topic in audiophile circles over the past few years. Some audiophiles have eschewed digital mediums because of the perception that they were sonically harsh or inferior. Hi-res audio promises to change all that, of course.
Many audiophiles have blindly jumped onto the high-res bandwagon and Dr. Mark Waldrep of AIX Records has been advocating on the concept of digital provenance — the origin or lineage — of a digital file from recording to mastering to hi-res. Mark correctly points out that many hi-res files marketed as hi-res are nothing more than CD-quality files at a higher sampling rate. In other words, just because something says it’s hi-res doesn’t mean that the fidelity of the track is high definition. David Pogue’s scathing review of the Pono Music Player and hi-res music just added fuel to the fire and the ire.
Living with the Benchmark Media Systems exceptional DAC2 HGC digital to analog converter and preamp and the AHB2 power amplifier over the past several months made me wonder about hi-res audio quite a bit. Indeed, one of the big selling points of the DAC2/AHB2 combination is it’s ability to play back the fidelity of hi-res audio. Once my review period was over, I decided to embark on a few additional tests and see if I myself could truly tell the difference with hi-res audio.
I had been aware but never taken the experiment conducted at Scott Wilkinson at AVS Forum and Dr. Mark Waldrep. The purpose of the experiment was “to explore whether or not true high-resolution audio files—with information beyond what a Redbook CD can accommodate—can be reliably distinguished from the same files that have been downconverted to CD specs.” To emphasize, this was merely an experiment and not a scientifically controlled environment. You can read about the parameters of the experiment here.
To summarize, Mark took three hi-res recordings that were recorded, edited, and mastered at 24/96. Mark did not alter the tracks dynamically. He then took each clip and downsampled them to 16 bits/44.1 kHz (Redbook CD audio), using the “gentlest” triangular PDF algorithm in Sonic Studio’s Sonic Process. Then, the downsampled version was upsampled back to 24/96 without adding audio data that wasn’t in the lower-res file. By doing this, someone would not be able to immediately which version was which based on file size or sampling rate indicator on a pre-pro or computer. The two versions of the files were then made available for download in uncompressed WAV format.
|The AVS/AIX hi-res audio chain identified the above as the ideal signal chain setup. With the Benchmark/Revel combo, I had such a setup in-hand to take the experiment myself. Image courtesy of AVS Forum.|
With the Benchmark electronics driving a pair of Revel Ultima2 Salons, I felt as though the setup clearly met and exceeded the parameters set by Scott Wilkinson’s experiment. I then loaded the files on my Mac and played them with VLC and made sure that my MIDI audio settings on the Mac were accordingly adjusted to output the hi-res file sampling rate.
Even though I was aware of the experiment and aware of the statistical results of who was able to tell the difference between the hi-res and non-hi-res versions. Moreover, I had made it a point to deprive myself of finding out which file was hi-res and which was the 16/44.1 version. I also did not open the files in any audio editing software in an attempt to look at the waveform. In other words, if I was going to take the experiment, then I was going to do it openly and honestly—even putting myself out on a limb.
I will confess that I was a bit nervous taking the experiment. Was hi-res audio really a farce? Throwing caution into the wind, I spun up the first track, Just_My_Imagination. I didn’t wait to listen through the entire song. I flipped to the alternate version and then back and forth and I could perceive sonic differences. The saxophone and percussion were areas where I could clearly hear the difference. I went to the next track, Mosaic, and sure enough, the difference was there too. With mosaic in particular, the triangle and upper frequencies shimmered with what I felt was the hi-res file. The final track, On_the_Streets_Where_You_Live, didn’t disappoint either. As soon as the vocals kicked in, I could not only tell a difference between the two files but the file I perceived to be the hi-res version had instant dimensionality to it compared to the Redbook CD version which sounded more muted and didn’t seem to have as airy a presentation.
I felt that the differences were so apparent and so clear that I didn’t need to listen through an entire song. I quickly dropped Mark Waldrep and John Siau of Benchmark a note documenting my results and I sent a separate email to Scott Wilkinson copying Mark and John.
There is no question in my mind that the Revel-Benchmark combination lived up to its billing.
I then brought my wife and my teenager in front of the setup. “Surely,” I thought, “They too would be able to hear the difference!” It was late and we had just come back from seeing a movie but they accommodated my request. My teen sat in the sweetspot and my wife was lying down at almost 90 degrees to the left Salon2. Sure enough, both my teen and my wife heard differences between the files but who preferred which version of the file varied. My teen was more closely aligned with my results and my wife seemed to prefer what I thought were the Redbook CD versions. Depending on the results, I guess I’ll either be vindicated or I’ll need to hand off the audiophile reins of this site to my wife!
I hope to hear back soon on whether or not my perceptions and answers were right or wrong. I also plan on taking the same experiment again using the Benchmark DAC2’s headphone amplification section and plugging in the Sony MDR-1A hi-res headphones and Sennheiser HD-600 headphones to see if I can still perceive a difference in the files when wearing a pair of headphones.