Klystron said:
but that by definition sampling techniques do not capture all the information present in a musical performance.
Yes, but no recording technique captures all information present. There's no such thing as a recording technique with infinite bandwidth and dynamic range, so you will always lose something. However, with digital, we can carefully control and exactly understand what we lose, and we can therefore select our technique so that we only lose the inaudible part. With analog, the distortion, noise, and lost information is much harder to quantify, which is a large part of why it's worse. You will always lose some information, so why not select your recording methodology so you can know and control exactly what is being lost?
In addition, amusingly, if you look at the actual capability of analog recording techniques, they come up measurably worse than digital in basically every way. A really good brand new vinyl record can achieve about 70dB of dynamic range, in perfect conditions. This is about the equivalent of what you can achieve with 12 bit digital audio. Studio reel-to-reel tape can approach 80dB (though it would typically be lower), or a bit over 13 bits. CDs have 16 bits though, so they handily outperform analog for dynamic range, and digital recording and mastering is usually done in 32 bit float (though in practice, the actual dynamic range can basically never exceed 20 bits or so, but there are good reasons why 32 bit is more convenient for recording and mixing before downconverting to 16 or 24 bit in the final result).
In frequency response, on the high end, vinyl does have a bit of an advantage over a normal CD. In perfect conditions, vinyl can theoretically reproduce up to 100kHz or so, and it's pretty easily demonstrable that it can do 45kHz, since the CD4 quadraphonic record format relied on a 45kHz bandwidth in order to work. However, this relies on nonstandard stylus heads - most record players will not be able to achieve this. In addition, it would rely on everything throughout the mastering and recording process also handling those frequencies, and most studio analog tape is not really able to do this without running at pretty high speeds, but then you sacrifice low end to do it. You pretty much can't get tape to both record well down to 20Hz and also to get usable high frequencies above 18-20kHz or so (I have found a few claims of being able to do 25-30,000Hz on tape though, so you could maybe achieve that in perfect conditions?).
Most records are also cut on machines that have an 18kHz low pass applied to the signal as well, so even if they had preserved high frequency content, that gets removed at the record manufacturing stage (because high frequency requires more power to cut which burns out cutting heads, and it's largely just noise anyways in most cases). Even if you do happen to get a record that was cut with high frequency content, and the entire mastering chain was able to capture and preserve that content, and the playback system is able to reproduce that content, there still isn't any evidence that it's audible in any way.
And, of course, if you really care about high frequencies that much, high sample rate digital is much better at capturing them than records anyways.
Also, on the low end, CDs are vastly superior to vinyl - vinyl typically sums bass to mono below 100Hz, and a lot of cartridges have noticeable deviations from a flat response as you approach 20Hz. Vinyl also struggles with phase misalignments in the bass. CDs on the other hand will happily and perfectly reproduce stereo bass all the way down to 20Hz and below, and this difference can be clearly audible.
.
