Saturday, October 31, 2015

The Ghosts of Tom's Diner

Here's an article that gets its geek on by exploring in extensive depth precisely what MP3 compression leaves behind. I poured over this like a scientist who'd been handed a time-travel manual from a disembodied arm extending from a glowing time portal. (Here's another version of it. Don't know why there are two.)

As the article mentions, that particular version of "Tom's Diner" (which is an artist-approved, overly-hot remix and not the original) is the primary song used to develop the MP3 file format because of its dynamic range, the abrupt transitions from sound to silence, and Suzanne Vega's complex voice - a classification I wouldn't have thought of had I not listened to her from that perspective. Singers whose vocal delivery is closer to speaking than singing are more complex by their very nature since they are further away from pure musical notes. So this is the perfect song to see what's left behind. Also, here's an article on the history of Tom's Diner, and how many times it's been sampled. Oh, and here's a brand-new one that popped up as I was finishing this.

The sound of what's left behind IS very ghost-like - static-laden, rumbling, with a snippet of a double-track voice looming out of the static. Yeah, it's kinda creepy. If there's a hell, this is what the piped in music sounds like.

Which got me to thinking. For years there's been grousing and debates over what MP3 compression does to the sound quality of a song. Personally, I've noticed that about 1 song out of 100 has some artifacting (burbling, echoing, ringing, phasing, etc.) or degradation of sound. One song in particular that I ripped long ago with an early generation of software - "Crash" by the Primatives off the Dumb and Dumber soundtrack (which is a different, hotter mix than the one on their albums) - resulted in the guitar intro getting jacked into ringing harmonies of itself, and I've always liked the effect. When I play the actual CD, I'm always slightly taken aback by the original sound of the recording as my aural memory expects it to sound like the MP3. I've ripped it again since and it always has come out perfect - so I've kept track of that one funky rip because I like it so much.

Other than that, the typical bad result is a little less bass and treble response; it's just not as bright or present as the source. This is only noticeable (to me) on a good system in a quiet room. In a car or in headphones (which I think are limited by the size of the mechanism that reproduces the sound - even in really good ones), I have yet to hear the difference between a CD and the MP3. Your mileage may vary.

Neil Young has been one of the more vocal grousers about the sound quality of MP3s and decided to do something about it. Thus, we have his where you can buy both high resolution (192.0kHz/24bit) FLAC music files (Free Lossless Audio Codec) and a Pono player that plays the FLAC files, which are translated through an Ayre Acoustics audio circuit that by all reports is totally bitchin'.

A music label can fuck up the (re)release of an old album so many ways. Leaving songs out from the original track list (which will never fail to be several fan's favorite). Substitute alternate versions of songs - Fleetwood Mac's Tusk was marred by this on every single subsequent release. Lose the original master tapes, as they have for so many; Joni Mitchell, Garbage, and ELO easily spring to mind. But, right up there is charging too damn much. The music industry is making the same mistakes it always has when rolling out something new: it's overpriced and you can't buy what you want. Quoth Bob Dylan: When will they ever learn?

One of the many reasons the music industry is in trouble is it charged too much for CDs for years, particularly after promising at the onset of CDs that prices would come down once the technology was "paid for." That day never came. So, people who would usually go to the record store to see what to buy this week started going online and downloading the one song they wanted. After a few years of trying Gestapo techniques to stop file sharing, Steve Jobs was the perfect asshole for the job of saving the biz from itself by offering legal digital downloads (so his new iPod could have some legal content), and insisting the prices be reasonable, so he got $1 singles and $10 albums.

Ponomusic finally opens up and most albums cost $20. The standard range is $24 for new/important albums to $13 for promotional albums. Perhaps, since it's "audiophile," the extra cost may be justified from the label's perspective (be warned however, the majority of the files are at the same kHz/bitrate as CDs). is worse, because each step up in resolution costs more; this album costs $25 at the best resolution.

Another question is do the songs sound that much better to justify the cost? I wanted to buy the song "Yeah Yeah" from Jackson Browne's latest album because it's a fantastic song, one of his best. (There's a few solid songs on the album, but "Yeah Yeah" is a classic.) Read somewhere that the high resolution (192.0kHz/24bit) version sounded heavenly - and the way it's arranged, I bet it is. So I went to buy that one song, even though I already have the album, and can't because they don't allow the purchase of singles. (Much profanity.)

Sad thing is, they could just OWN the market if the prices were competitive with the cost of an MP3 or iTunes MP4 version, and you could buy the songs you want. My daughter loved buying from the Apple store because she said it sounded better than an MP3. When she moved to a Droid and had to convert the songs to an MP3 to transfer, she was sorta bummed. (And it was a pain in the ass, because Apple requires you to convert each MP4 to and MP3 manually.) So my daughter would be a loyal client if she could download high-def FLACs of her music for the same price. Her whole generation would take to it like ducks to reasonably priced water. Guess, Pono (and HDTracks) will just be the BluRay of music.

I'll address the question of does the supposed better sound justify the cost a different way. As I was writing this, Donald Fagen's "Security Joan" popped up in my MP3 music queue, and oh my gosh, the walking bass line and the cymbal ride sound fantastic in my headphones. I bought the version that included a Dolby Digital/DTS 5.1 high-resolution DVD of the album. Of course this MP3 doesn't sound like the DVD and it would be cool if it did.

But I've noticed something about those DVDs , because I have ones from Dave Matthews Band, the Foo Fighters, and The Flaming Lips: they're only interesting and a good listen if the music has a lot of space in it, sort of jazz-like. Thunder rock and even something as dynamically limited as a Tom Petty song (I loves myself some Tom Petty, so that's not a crack on his music) often gain nearly nothing from high-definition. The Foo Fighters DVD actually sounds worse than the CD of the same. It's such a wall of sound there's not of lot of discernible distinction in the overall sound, it just seems louder, and the sonic depth makes some of the songs overwhelming rather than pleasant. To give you an idea of what's that like, have you ever gone to see a hard rock band in a room that was too small for the volume they used? All you can really hear is vague changes in tone around the roar of the sound that neither your ears or brain were meant to process.

It occurs to me that the mixing on these high-res DVDs might be the issue. Sound engineers have worked forever on how to put a big sound into a smaller space, like a vinyl album (with the problem of too much bass throwing the needle) or a CD, with the limitations of the digital sample rates. Perhaps mixing to the expanded range is difficult to do, or maybe no one has managed to figure out how to do it well. It's not as if there isn't a lot of sonic info available on the Foo Fighters DVD, it just sounds like a big buzz.

Simply put, Joni Mitchell sounds great on a high-resolution recording; KISS not so much.

Until I can actually buy a hi-res FLAC, this assessment will have to do. I hope I have the opportunity sometime soon.

Finally, here's some trivia, ephemera, and doo-dads I've tripped upon:

The sound geek magazine Sound on Sound has several "how it was made" articles of classic albums, one of my favorite wastes of time. Dug this nugget out about ELO's "Don't Bring Me Down":
"Gradually, Jeff started getting into it, and, as there was a plan for ELO to start a concert tour in Australia, the song was originally titled 'Don't Bring Me Down, Bruce'. This was meant to be a joke, referring to how many Australian guys are called Bruce, but we couldn't leave it like that, so eventually we replaced it with 'Gruss', based on the Bavarian greeting Grüß Gott — 'greet God'. Gruss, not Bruce, is what you hear in the song immediately following the title line. A bit like Freddie Mercury joking around at the end of Queen's [1985 single] 'One Vision', singing 'fried chicken'."

Don't remember where this is from, but the Beatles' American concerts had the distinct odor or urine as part of the sensory sensation, as the hysterical girls would not only scream ceaselessly, they would wet themselves, too. According to the people who ran the venues, urine would run down the aisle in rivulets.

Rolling Stone has one of their listly articles on the best songwriters. It's pretty good, with not a lot head-scratchers.

Found this interview with the boss of 70s and 80s themesongs, Mike Post, who was the only other guy besides Henri Mancini who had theme songs that regularly became radio hits. In this 3 hour interview (with some highlights busted out for your impatient pleasure) I discovered that Post formed the 60s band "The First Edition" monkees-style, and was dubious about Kenny Rogers being in the group. He also produced and arranged the hit "Classical Gas." The orchestra break in the middle was his idea and he wrote it.

Here's the best thumbnail description of classical music composers I've come across:
"If the music was joyous: Mozart; if it was angry: Beethoven; regal: Hayden. If I heard an organ: Bach; a harpsichord: Corelli; a lone violin blazing up and down the scales: Vivaldi."
(The article it's from is rather harrowing. It's about a father's alcoholism and how the now-grown son relates it to an unfortunate childhood viewing of "The Exorcist". )

'Til next time.


Whisky Prajer said...

Oof -- "harrowing" indeed. Happy Halloween.

I might have shared this with you already, but I absolutely love the aural tweaking on the Donald Fagen Nightfly Trilogy DVDs. The set comes with mp3 files, but those are of the original digital releases and, of course, sound way sub-par. Near as I can make out, the two early discs are remastered into four-channel surround-sound. I pushed it through two channels, and ripped that -- using Audacity (I think -- it's been a while) -- and the result has been eight years of delicious portable listening.

While I was at it, I pushed the Morph DVD through two channels, just to hear how it sounded. Answer: really muddy and awful. Ripping the CD: really great. If I had to guess why pushing the most recent DVD through two channels resulted in such lousy sound, I'd say they probably engineered Morph to the highest current standard -- DVD multi-channel Surround-Sound for 16-plus speakers -- then re-engineered that for the CD release.

I can't believe how thoroughly Apple is messing up what was a really sweet ride (for them). If you use iTunes, it's true -- there's a discernible difference between the m4a and mp3 files. But I've got all sorts of mp3s ripped via various Linux applications that are easily the equal of, if not superior to, Apple's default format. I've used these files to do some comparisons with Suzan Vega-like performers, and I can't pick out deficiencies.

The ears are definitely going. High-end frequencies really bother me, especially watching sporting events on television. Crowd noise irks me. Hockey, with the skating noises, bothers me so much, I've usually got the sound down to a level that invites ridicule from my daughters. I used to twit the Old Man about the same, back when he was my age. Guess it's my turn.

Yahmdallah Bjorknickerfod said...


Thanks for the tips on the Donald Fagen albums. Very interesting, as they usta say on Laugh-in.

As my post indicates, I'm filled with dismay at Apple, Ponomusic, etc. etc. Why can't major labels figure out how to sell music?

The ear thing is exactly age. I have the same thing. My eldest and her buddies who know me caught me just nodding and agreeing a couple times at a recent party. My daughter sold me out and said she can tell when I can't hear her.

Oversharing moment: I've discovered it gets worse if we old(er) folks don't pay attention to earwax buildup. You have to get an Ear Syringe, take it into the shower with a cup, and blast water through your ears a few times to clear it all out. If it's been awhile or never, the first time will result in horrors at the bottom of the shower. But you will notice less ringing and you can usually hear much better. Apologies to the kids whose bodies haven't turned on them yet.

And apologies for taking so long to respond. I don't do personal stuff on the web at the new job. I knew the folks who monitored that stuff at the old place, and knew they had my back. But at he new place, I don't, so I have to wait until I'm home to do this stuff.

Whisky Prajer said...

No need to apologize. Heck, I still recall the days of "pen pals." And YES to the Ear Syringe. Absolute necessity, that.