Sound Quality of Audio Systems

Are vinyl LP’s better in certain ways than music digitized and recorded on CD or into a computer file? This is an interesting area of inquiry that I’ve thought about a lot over the years since digital recordings became a practical reality. There are legitimate arguments for both sides of the question, but both types of recording come with their own special issues. I was recently intrigued by an article that mentioned “a youthful LP convert” saying, “There’s less compression, more, much more information. More air. More depth.”

This statement is interesting because CDs, despite their arguably inadequate sample rate (or bit depth), require much less compression than vinyl, because the full dynamic range of vinyl is much lower than for CDs (65 dB vs 96 dB). Technically, the dynamic range of vinyl is limited by the physical velocity capability of the cutting head and the subsequent playback stylus, so at high frequencies, where stylus excursion is minimal, vinyl dynamic range can approach that of the CD (about 90 dB). But at low frequencies, the dynamic range may be little better than 25-30 dB. (Some dispute this on mathematical grounds, but I strongly support it on empirical grounds.)

As often happens in discussion of audio perfection, the debate often gets quite silly, but even in 2014 vinyl has to be mastered with compression, while many CDs that I buy, or create myself, have no compression at all. I still remember the thrill of hearing, for the first time, Emil Gilels performing a Beethoven sonata, mastered for CD—the loud notes were as loud as a real piano, and the soft notes were as soft. I had never heard that on a recording before, and it was stunning. That kind of dynamics can’t be reproduced on vinyl without the use of non-standard (and risky) compression and decompression processing (such as dbx). The standard vinyl mastering technology of the last several decades uses a simulation of active compression for recording (the RIAA EQ curve), which must then be reversed for playback.

To be fair, my uncompressed CDs also are quite capable of destroying conventional speakers due to high dynamic range of the signals they provide, which was only very rarely a concern with record players. But that’s a risk I’m happy to live with in exchange for accurate dynamics.

The LP vs CD discussion often dismisses what I consider to be an abominable problem that severely interferes with my enjoyment of the playback: tape hiss and surface noise. Furthermore, playback is intrinsically an ablative process, gradually destroying the medium as it is used. (True, a very high quality tone arm and cartridge and stylus can minimize damage and prolong the life of an LP to many hundreds of plays, but realistically only a few people possess such gear.) The accumulating ticks and pops are (to me) completely intrusive and destroy the ambiance of a good recording.

The “Magic” of LP Listening

But there’s a really fascinating psychoacoustic element in all this. Let’s say, for the sake of argument, that vinyl and CDs can reproduce audio signals of exactly the same quality, with the proviso that vinyl’s unavoidable surface noise and accumulation of microscopic contaminants are still present, along with the electro-mechanical issues (without these provisos there would be nothing left to compare).

When we listen to a record, and I do remember this ever so deliciously, there is a moment before the playback begins, when the audio system is absolutely silent, and absolutely nothing is heard (barring amplifier hiss and hum, I hope). At this point, “we’re just waiting.”

Then, in a moment of anticipation, we hear the quiet but distinctive tap of the stylus gently contacting the polished vinyl between grooves. “It’s about to start…”

After a short but unpredictable interval of silence, the stylus slides into the groove with a soft thump. This is an exclamation point for our listening experience, and has nothing to do with the sound quality. It announces the performance, and we’ve conditioned ourselves to respond to this through countless playback experiences. “It’s beginning!”

Then the surface noise begins, lending a subtle ambiance to the entire recording that never stops until the final lift of the stylus and a return to absolute silence 20 minutes later. The surface noise is often rather faint, but the speakers are thrumming with it, and the air in the room is being pumped, ever so gently, with an almost imperceptible energy. More anticipation.

Right after the brain settles into the thrum, we hear the beginning of the (hopefully faint) tape hiss as the recording begins (assuming an analog recording that’s not direct to disc); that hiss is also minimal, perhaps almost subliminal, but it’s unmistakable. Now the music is about to start, immediately, and we know it. Our brains are already reactivating internal subjective filters to ignore the thrum and hiss and savor the music.

Finally, the differential signal appears at the speakers, real music, and both relief and delight fill the heart—along with whatever other reactions the recording may evoke.

I submit that listening to music in this way, always bracketed or wrapped in the sound of the mechanics of physical playback, enhances our experience in subtle ways. It also primes our neural systems to deal with the music in a special context that is always apart from the unmanaged auditory experiences of normal life. This context is subtle, creative, symbolic, and probably also Pavlovian in the presentation of recorded music via LPs.

Pros and Cons

There are definitely some advantages of vinyl over CDs in certain areas, and some advantages of CDs over vinyl in others, and they’re worthy of note.

With vinyl, excellent reproduction of excellent pressings definitely affords a much higher upper bound to the frequency response, probably up to twice what we can hear directly (50kHz, by some estimates). Excellent + excellent also affords almost no surface noise, potentially zero tracking error, perhaps no dust ticks, and so on. But to achieve this, we must hope the vinyl was exceptionally high quality, that the spindle hole was punched in exactly the center of the record, that the vinyl is absolutely flat, that no dust has fallen on it, and that it hasn’t been played too much (or even once on a badly adjusted system). And we must not mind the inevitable harmonic and intermodulation distortion of both the reproduction electro-mechanics, and the RIAA equalization applied before mastering and then undone during playback, and the extra stage of amplification required to get the moving coil cartridge’s signal up to standard preamplifier levels.

CDs, of course, lack many of the picky and problematic issues just mentioned, but they have a few of their own. On the plus side, there is no surface noise, no dust, no off-center rotation, no warp, no rumble of turntable mechanics, no susceptibility to a heavy step on the floor, no destruction of the recording from playing it, no tracking error, no stylus compliance issues, no mistracking (shatter), no RIAA pre-processing, no playback RIAA un-processing, and usually at least one less analog amplification stage.

But 44.1 kHz isn’t sufficient for a Nyquist frequency that can accurately exceed human hearing and be guaranteed to reproduce all the subtlest phase relationships in the high frequency domain (where the brain excels in an ongoing spatial interpretation of every incoming vibration). And 16-bit amplitude resolution is insufficient for a really smooth transition from 0 to 96 dB of loudness. And the typical CD playback device is usually a minimalist electro-optical-mechanical gadget that sells for $30 or less, while the typical “serious” LP playback device (system) is usually at least 5-10 times more expensive, often with turntable, tone-arm, cartridge, and stylus all purchased separately for much more than $30 each. Therefore  many of the comparisons we read about are made between superior LP gear and inferior CD gear. (Superior CD gear would, perhaps more than anything, make use of highly accurate DACs running on a clock with absolutely no drift or jitter.)

And I won’t get into the fact that the left channel (the outer side of the record groove) is traveling a little faster than the right channel, and should therefore have slightly different (better) dynamic range and frequency response.

Other Factors

The thing is, of course, that we don’t listen to theoretical vinyl or CDs anyway. We listen to those particular LPs wedged onto that shelf, or strewn around that table, and to those CDs that get stuffed into this player and that, from car to livingroom. And some of the comparisons cited in the article are likely to be between MP3 and vinyl, which is like comparing a postcard to an 8×10 darkroom print.

Then, of course, some LPs are mastered from the same digital master recording as the CDs. In these cases they have to undergo RIAA compression and equalization, and this remastering for vinyl may include other compensatory acts by the engineer. Some source material intended for vinyl is mastered specifically for more audio-sensitive listeners, and so might not be fairly compared to the CD’s different source. Compression to MP3 further mangles the subtleties of the sound (the “youthful LP convert” spoke of “air” and “depth” and “more information”).

Once we get into the realm of air and depth, however, we have to address the recording process itself, because modern albums tend to be recorded with almost no regard for phase at all. Multiple mics are the norm, and they are processed in dozens of ways en route to the final mix. Mics at various distances record the same instrument, directly mangling phase, and “room mics” sometimes record natural ambiance in the studio, to be mixed in with the close-mic tracks of each instrument. Ambiance synthesis, dynamic EQ, and a whole rack-full of physical or digital signal processors are usually applied to each track, to the submixes, and to the final mix. The final is then usually “mastered” by passing it through yet another chain of processors so it can be brought into some kind of subjective conformance or compatibility with the other tracks on the album. For most modern pop recordings, a huge amount of compression is used throughout this process, instrument-specific and then again submix-specific, so that maximum loudness can be assured on the CD. Recordings such as these don’t fare very well on vinyl, and (to my ears) sound equally horrible on CD. (But much of the pop-music vinyl is mastered this way!)

“Bucking my parents’ blind acceptance of CDs,” (quoting my own hypothetical LP convert) “is one of the reasons I prefer vinyl. Getting something better than a crappy MP3, something big and graphically splendid and tangible, is another. Getting into the process of listening is another. But my friends who just rip their LPs and play them as MP3s are nuts.”

All arguments about theoretical quality are moot—almost nobody, realistically, is attempting to extract maximum quality anyway.

As often happens, over time my opinions have evolved. With the latest DACs, and with better than average players, the deficiencies of CDs are, while still quite real, now quite acceptable, while the deficiencies of vinyl remain (to me) almost intolerable. The advantages of CDs are numerous, while the advantages of vinyl are few and rarely realized.

Recently we have seen vendors selling 24-bit 96 kHz remasters of the classics. Some are better than others (some are just the same low-resolution masters floating in a bigger digital box full of zeros), but some are stunningly better (again, only on gear that preserves these subtleties, and in a listening situation where the full dynamic range can be heard). Now that I can master my own compositions at 24/96 or 192 and conveniently play them back from computer direct to my power amps, my primary concern is that we don’t yet have a standard consumer medium for 24/96 material so my audience can reliably hear what I heard when I created the music.

Coda

Since psychoacoustics is as interesting to me as audio accuracy (to the horror of my high-end-audio friends), I am not averse to playing with signal processing, from tone controls to (for example) an early Phase Linear processor called a “Peak Unlimiter and Downward Expander (Autocorrelator).”

Some of the perceived “clarity” of vinyl probably comes (in my estimation) from inertial overshoot effects in the mechanical action of the groove-stylus-moving coil (or magnet) assembly. This could be generating a very subtle peak emphasis that would lend an air of crispness and detail to playback.

I’ve experimented with this kind of effect, using the latest digital processors, and the results (totally subjective, mind you) are as expected—certain kinds of distortion can make the music sound better. Further, these very brief expanded peaks are right in the region where the 44.1 sample rate is likely to cause a little smoothing or dampening, compounding the perceived loss of “air” going from LP to CD. In fact, at the 96 kHz sample rate, it is just that same “air and detail” that constitutes the main improvement. (Along with much greater definition in the low bass.)

If we are willing to play with the psychoacoustics, then the relevance of most objective measurements gets skewed this way and that, and we could postulate a post-processor, inserted between the preamp and the power amp, that could recreate the sensations of the best vinyl from almost any source. We could even add that magical tap, thump, thrum, and hiss from the days of yore, to get the lizard brain back into the mix.

 

This entry was posted in Uncategorized and tagged , , , , . Bookmark the permalink.

Leave a Reply

Your email address will not be published. Required fields are marked *

This site uses Akismet to reduce spam. Learn how your comment data is processed.