I think I could contribute a few things to this discussion. I was educated as theoretical physicist in quantum theory and gravitation and also worked in Radio Astronomy aperture synthesis with VLBI and developed programs for image reconstruction in 2D, phase reconstruction, things like that. I am active in computer science and IT for quite some time now, so my previous knowledge can be rusty.
My equipment is relatively simple: an Oppo BDP-95 source connected via XLR directly to the Ayre V-5xe Evolution amplifier (which is Stereophile recommended at any price in 2012). This drives two Monitor Audio PL300 speakers connected via top class bi-wired cables. The room is not acoustically treated but is a good size room with generous ceilings, good distance from the walls etc. Acoustic treatment might help to further improve things but as things are, I am very happy with the results, although I am quite possibly losing some resolution due to going to the amp directly with no pre-amp in place. I run the test below at the same volume setting set directly in Oppo. The listening position is approximately 12 foot from the speakers, which are toed in just a little but are not facing me directly.
What I am looking for in music is definition and realism. I am extremely happy with the recordings from HDTracks I have so far acquired. To me, the difference is extremely obvious, I would use even a stronger word, perhaps, overwhelming. I enjoy Classical Rock, classical orchestral music and progressive rock but I do listen to many other genres of music--and watch Blu-rays, but that is a separate story.
In this post, I won't go theoretical but will simply reproduce four snapshots from the Spectrum Analyzer app running on iPad 3. To clarify, the iPad is recording using the built-in mic what it hears at my sitting position, so we have a full path from the source to the speakers and via air to the iPad and to myself. The test subject are the Yes, Close to the Edge album's title track with two episodes: the first guitar attack and the second guitar attack in the first 1 1/2 minutes or so of the track. I have a CD and an HDTracks 192/24 recording and I am simply taking the full sweep of Fourier with audio sample rate of 44100 Hz with 2048 sample Fourier transform.
HD 192/24 version
HD 192/24 version
I will make relatively few comments as, I am sure, you will be able to pick out many differences in these versions.
I can hear in a double blind test with my partner a marked difference between these passages, with the HD version sounding much sharper, with all the transients and attacks sounding whip-like, life-like, and also very pleasing on the ear and really beautiful.
The CD and HD snapshots have hotspots very similar in color under 2KHz. I can analyse the snapshots further but it is clear the volumes are the same and the highest levels are the same. However, the snapshots are remarkably different.
- Quite obviously the transients go much higher in the HD versions.
- In the first passage, the HD version has sharp ticks between 9KHz and 11KHz prior to the attack at around 5.7 seconds offset (the snapshots need to be aligned by hand between CD and HD). The CD version has those washed out.
- The transients are much sharper in the HD version.
- The horizontal stripe around 7K from the offset of about 15 seconds stands out very clearly on the HD version and is washed out on the CD version.
- In passage 2, you can observe the lower left-hand corner. Note how the signal goes to blue quite clearly between the horizontal stripes in the HD version. The CD version is all washed out.
Bottom line. These are just random observations. You can find as many differences as you like but it is obvious that the CD version looks like a blurred out version of the HD snapshot. All troughs are filled up and the peaks are washed out horizontally--creating smudged attacks--and vertically, creating a halo around each note.Edited by warrior-kid - 6/3/13 at 6:40am