Re: The things at work in a listening test
Jeff,
Four things:
1. Virtually every speaker on earth listened to at 95dBA at 1 meter away, if that is indeed where you are listening, will sound harsh(ish). That is screaming loud so close to a box. Even a few extra feet of air absorption will dramatically change your impression of the top end.
2. Songs/CDs are mastered at a huge variety of levels, and a lot of different eq balances as a result. Many cds are mastered very crispy on the top end so that they have "air" at low volume levels. You have to be very VERY careful about your source material.
3. Certain brickwall mastering limiters also do REALLY BAD things to the very high frequencies. The waves L2 when pushed hard was/is particularly egregious. Anything that sounds like "frying bacon" on the decay side of transient events is almost always IMD from some component of the limiting chain.
4. Some playback devices have very poor analog headroom in front of their reconstruction filtering. Tracks mastered right against 0dBFS in the digital realm can have intersample peaks more than 2dB higher after the reconstruction filter. The intersample peaks can be very much in the "not pretty" realm of the analog electronics at the end of the playback chain. TC Electronics presented an AES paper on this a few years back using consumer CD and DVD players. I also saw a test (which I cannot find) that shows certain Apple iXYZ products as poor performers (especially the original iPhone 3G).
In addition to my live audio work, I also do some mastering, and it never ceases to amaze me the diversity of top end balance that is requested. Its also amazing to see how much a half dB shelving filter above 12kHz can change one's perception of the track brightness. Not only does the overall track air change, one's perception of the balance between 2-5kHz also shifts! The latter effect is especially tough with low end earbuds that have nothing above 10kHz. A track that is balanced on a quality playback system suddenly becomes spitty on the earbuds that are missing the last octave.
So, in summary, mastering is a compromise, and poorly chosen source material will badly skew a demo. I've seen too many demos of good products ruined with poor source material, level choices, and listening positions. Unless I get to pick the music, and the level, and the listening spot, I don't make loudspeaker conclusions from listening tests.
Meyer has historically been one of the good guys, and the UPJ class products are what put them on the map, so don't immediately write them off as a result of this listening experience.
