Why CDs May Acutally Sound Better Than Vinyl

[ernstinen]

I know this is taking a tangent, but vinyl records and first generation reel-to-reel masters are a completely different animal. I recently had my 15 i.p.s. reel-to-reel mastering deck refurbished, and A/B'd the original tape mix of a song of mine to the CD version. It was no contest --- The reel-to-reel original mix SMOKED the digital copy! The detail was astonishing. There were parts that were mixed quietly in the digital version that I couldn't even hear, but listening directly to the analog master tape was a true revelation --- I could hear EVERYTHING that was going on in the studio! Of course, I rented the best microphones and analog gear that was available, and had the best musicians. That counts for a lot, too.

I know, I know, it all depends on the digital converters used. I'm aware of that. But to my ears, the best-sounding recordings ever (like Tom Petty's "Wildflowers," Steely Dan's "Gaucho," and Nirvana's "Nevermind") were all recorded on analog gear... Steely Dan used digital for one album a few years ago, and switched back to analog because they thought it sounded better. Donald Fagen said that working with digital "Loosened the fillings in my teeth."

And their engineer, Elliot Scheiner, said: "I grew up and learned analogue and I'm an analogue geek. It's not that I'm kicking digital, but analogue has a much better sound. When you are able to A/B analogue and digital, which we could do in this case, there's simply no comparison. The top end is so sweet and beautiful. I've never heard anyone say about digital, even at 24-bit/96kHz or 192kHz: 'Isn't the top end as sweet and beautiful as you've ever heard?' You don't because digital just doesn't sound that way." And none other than Bob Moog wrote a paper explaining why analog tape is a superior recording medium.

So while I agree that vinyl records are mastered to roll off everything below 60 Hz (because too much low end can cause the stylus to jump out of the groove) and are noisy with clicks and pops, a quality reel-to-reel mastering deck can sound better than all but the most expensive digital gear. Vinyl and tape are apples and oranges; vinyl is fun because you have a lot more art to look at!... And of course, 24 bit and high sampling rates are vastly superior to the archaic 16/44.1 of CDs. I own some CDs, but everything I buy commercially now is 24-bit/ 96kHz. I never owned an iPod because mp3s sound SO bad that I waited until smartphones came out that would support high-resolution audio. And digital is so easy and fast to work with (and portable). Personally, I mix digital and analog gear to make my recordings sound as close to the warmth of analog as possible.

My 2 cents,

Ern

[Len911]

Bob Clearmountain, "To me, analog is unpredictable; it does that funny thing to the bottom end. You work really hard on the bottom to get it exactly right, and then you play it back on your analog tape, and it's like, 'Oh, what happened there?' The storage medium is making decisions about what the bottom end should sound like."

Below are frequency response curves I've made myself by testing various analog recorders I've worked with over the years. What I wanted to know was, after a standard alignment procedure was performed (or, seemingly, not performed) by the studio techs (or sometimes by myself, to be sure), what was coming out of the machine versus what went in? Analog machines all sounded different to me, and I wanted to know why.

http://www.endino.com/graphs/

It's not only the medium, it's the human and the gear they use between the tape machine out going to the digital in. I can't imagine that "low-end head bump" being particularly kind to the input of the a/d converters. A fairer comparison might be recording tracks with the signal split with one signal to a tape machine to be recorded and the other to be recorded on a hard disk. It's possible that "the sweetest top end you've ever heard" is the lack of a top end?? -or the low end head bump?? possibly it's the lack of s/n ratio smoothing things over a bit??

I'd conclude from the comments is that if anything, analog is not either uniform or consistent. Hot and cold also tends to affect the outcomes, warming up or overheating. But what I'm not concluding is whether it sounds better or not, sometimes maybe yes, sometimes, maybe no.

[ernstinen]

http://www.endino.com/graphs/

Ironically, I looked at these exact same graphs the other night! And it DOES show individual analog tape decks can vary WIDELY. The particular Studer that stood out as the flattest I found in eBay for $9000. Past my budget!

And analog decks have to be aligned and serviced. When I went to engineering school, aligning 24-track 2" decks was beyond me. Not my thing. I'm a musician, not a techie!

That said, when everything is working correctly and you have the best engineer working with analog, I still think it sounds better. The question is WHY? I think Bob Moog explained that. Mr. Moog was obviously an analog genius, and knew just about everything, so his answer is very complicated, but what I took out of it is that a professional, perfectly aligned and maintained tape deck records HARMONICS that go way above the perceived range of human hearing (about 23kHz for the best ears). Human ears can't HEAR above that, but they can SENSE when the harmonics are not there. A CD's sample rate is 22.05kHz (44.1 divided by 2). So what Moog stated from his experimentation is that the aforementioned tape deck using the best quality tape (and tape technology has INCREASED even after digital took off), the frequency response goes up to somewhere like 30-35 kHz (!) or higher. Hence, on a CD you are no way hearing (or "sensing") all the harmonic information that quality analog tape can record.

So I'm with ol' Bob Moog. It would be interesting to A/B a great sounding original analog recording, 'A' being the commercially available CD, and 'B' being the original master analog tape. I'd bet my first born son that the tape recording sounds a LOT better than the CD. It's just simple mathematics. A CD cuts off the top end at 22.05kHz, whereas the analog tape goes WAY above that. Even on that analog tape recorder graph, some of the decks went up to at least 27kHz before the high end dropped off dramatically... And by personal experience, when I used to master analog recordings, I'd rent the #1 EQ unit available, a Massenburg. That unit went up to 26kHz. And I'm positive that cranking up that frequency made the recording sound better, adding more "air" (or as Moog would call it, harmonics).

Peace,

Ern

[Len911]

When you said, "engineering school", that reminded me of a Navy A-school I attended. We had a class called, "Logical Troubleshooting" where the instructors would insert a problem in these R390 tube radios, and we would then find the problem. The problem was that these radios were so old and out of tune, they were almost impossible to troubleshoot, so you had to tell them how you arrived at the problem, so they could tell you, no, that wasn't the problem they inserted and you could continue on troubleshooting. Believe me, that doesn't teach logical troubleshooting, but it does make you question the competence of the school,lol! https://www.google.com/search?q=r390+ra ... 43&bih=604

I totally agree with you about Massenburg, Father of parametric eq. He and Doug Sax have some of the finest mixed and mastered
cd's imo.

I found this about some of the characteristics and limitations of tape on wiki

The storage of an analogue signal on tape works well, but is not perfect. In particular, the granular nature of the magnetic material adds high-frequency noise to the signal, generally referred to as tape hiss. Also, the magnetic characteristics of tape are not linear. They exhibit a characteristic hysteresis curve, which causes unwanted distortion of the signal. Some of this distortion is overcome by using an inaudible high-frequency AC bias signal when recording, though the amount of bias needs careful adjustment for best results. Different tape material requires differing amounts of bias, which is why most recorders have a switch to select this ... Variations in tape speed cause flutter, which can be reduced by using dual capstans. Higher speeds used in professional recorders are prone to cause "head bumps," which are fluctuations in low-frequency response.

https://en.wikipedia.org/wiki/Tape_recorder

AC bias is the addition of an inaudible high-frequency signal (generally from 40 to 150 kHz) to the audio signal.

The missing link in hearing is how the brain processes what the nerves from the ear send or don't send it.

[ernstinen]

BTW, recording engineering school was a joke. The "owner" of the school couldn't even do a simple punch-in or punch-out. He knew nothing about music at all. I never learned a thing there; everything I know about recording was completely by trial and error. I took a couple classes, and dropped out. The main reason was because I wanted to be on the "other side" of the glass, i.e. being the musician, not the engineer. I learned how to engineer because I realized it would in the long run be much cheaper to do it myself, rather than pay for studio time.

As far as tape noise, after I left 4-track I NEVER had any tape noise on my recordings. That's because I recorded analog tape way into the red (and most of the material was rock). That's another great thing about analog tape: Natural tape compression. My old analog recordings "sound" nearly twice as loud on digital meters as compared to pure digital recordings. The tape was in the red, the inputs were in the red, the board output was in the red with zero distortion. Try THAT with digital! Analog recording is an art, and takes many, many years to learn how to do it right. As does digital, but it's a completely different way of recording audio.

I took a digital audio class at UCLA a few years back. The only thing that made sense to me was "fill up all the bits." That's like in analog: Saturate the tape as loud as possible without distortion. I still use limiter/compressors recording digital to "fill up the bits," that is make the signal as hot as possible without distortion. And in digital, when there's distortion you hear it immediately! It's a horrible sound... As Ray Charles once said "If I don't HEAR any distortion, there ISN'T ANY." So much for being conservative with the meters in analog recording LOL! Trust your ears.

Ern

[Len911]

I agree with you Ern about "schools". I have discovered the same thing in my experience. First I would say that there really are a few really smart people who do know their stuff, then there's the rest. I was disappointed in the Navy schools I attended. I realized rather quickly that it was taught by folks who were on shore duty, and even though they had been in the field, they knew little about electronic theory, they swapped boards in the field, and they learned their course material in about a week from the instructors they were replacing. It was utter chaos and confusion. But then again, I've always been the kind of person who questions things until they start making sense. I'm the "but why kid".

Analog is different than digital as a storage medium, primarily because analog effects the content, whereas in digital it's still the analog that effects the content. Analog is part of the sound, in digital, it should never be a part of the sound. Digital is more like the printed word, not a part of the painting, if it's smeared and illegible, it's useless. There is no digital without analog, and it's the integrity and accuracy of the analog that determine the quality of the digital. Voltage and clock accuracy in the circuit.

The fascinating thing about digital to me is that it only does one thing, it marks the amplitude. It's only into how many pieces it divides the usually 5v into, and how often it marks the amplitude, sampling frequency, that determines the quality. The phase is a byproduct. It's like measuring a speaker moving positive and negative, or in and out. Sine waves. Because it is so simple it almost seems unfathomable. Theoretically, you could write your own wav or aiff files, you wouldn't need any audio equipment, I just don't know how practical it would be to mark 44,100 points a second, or how you'd know where. I did actually wonder once if it would be possible to just edit a wave file instead of messing with a daw and tweaking plugins.