It’s not a lot of fun listening to “Track 1” on “Unknown Album,” but that’s what happens in your car or home music player if the titles aren’t recognized. I’m fascinated by the process in which my computer consistently recognizes the data on even the most obscure music, but the fancy systems in the cars I test fail to properly identify even the most common music
. Let me explain to you exactly why that is.
On one level, I understand the problem — cars come with onboard databases that need an Internet connection to be constantly updated to recognize newer albums. That’s not a problem on Web-connected computers. But I’ve had my car’s CD player fail to list the tracks on even older catalog material. And why isn’t the information, which can’t take up much space, simply embedded on the CD itself for any player to read? What gives?
I’ve noted that when I pop some more obscure CDs into my computer, it pops up a box asking if I want to upload the track information to Gracenote. I figure I’m helping the system learn by saying “yes.”
One way Gracenote works is through track time recognition. If the album has, in sequence, tracks of 3:43, followed by 2:19, 10:55 and 7:20, why, it’s Derek and the Dominoes. I’m sure I’m oversimplifying, but that’s essentially it. I’ll leave it up to Gracenote to explain why this is the best method out there.
Stephen White, the president of Gracenote, was patient enough to explain the weird and wonderful reality of track naming. “In the early days of CDs, when the standards were created, there were specifications for putting the track data on the discs, but the labels didn’t want to be bothered by it,” he said. “It would have meant a human being assigned to typing in all that data. There’s a place for it on CDs, but they don’t do it.”
That odd situation opened the door for early programmers to create downloadable open-source databases of track names for people to download. Out of that work grew CDDB and, ultimately, Gracenote. A major problem was that, say, Seals and Crofts could also be spelled as Seals & Crofts, or James Taylor listed as Taylor, James, and that led to a lot of confusion. It reigns to this day — I’m constantly correcting data misspellings.
Only one label, Sony (Gracenote’s current parent), had a brief go at professionally entering that data, which is why one in eight or so CDs yields data on otherwise clueless cars. Oh, and I found out that the reason some older CDs don’t register is because they pre-date Gracenote’s database.
We’re supposedly in better shape now. As White explained, many of today’s automakers either use an onboard Internet connection to cue into Gracenote’s 13 million online track names (rare, since only 5 percent of cars are wired) or embed a much smaller library of 250,000 to 500,000 CD descriptions into the computer’s memory. Most carmakers are Gracenote’s customers.
The drawback of the latter option is plain, since hundreds if not thousands of CDs are released every week and the database goes instantly out of date. White says that Ford and GM, as well as Audi, are pioneering the best connected systems today, though MyFord Touch has its issues. The state of the art is the Tesla Model S sedan, which has an always-on Internet connection and so reliably gathers track names and album art. Seriously, you can use the 17-inch touchscreen to surf the Web even when the car is in motion.
Since adding memory and paying for Gracenote can add $20 per car, some automakers leave their lower-end cars with no track naming at all, hence a lot of “Unknown Album” and “Track 1.” I don’t get that. What’s the point of using precious dashboard space for track descriptions without any way of ensuring that the information is actually available? Also dumb is the use of huge type (supposedly to reduce distracted driving) that requires five screens to get a whole track name.
White predicts that fairly soon every car will have some kind of Internet connection, which will go a long way toward solving this problem. Meanwhile, the challenge grows, because consumers are accessing music from hard drives and their cellphones, and they want the track data — and cover art — displayed instantly at the head end. Gracenote is trying to keep up with all that, according to White. “They expect a rich interface,” he said. The video below offers an online interview with White, if you'd like to know more about how Gracenote works:
One might expect Gracenote to worry about the imminent demise of the CD, considering the business is based around the shiny discs. “Don’t give up on the CD yet,” White says. “Some 50 percent of all music sales are still CD-based. “We’re bullish on the CD. Our research shows that it still dwarfs everything else.”
Scherf told Wired
that Gracenote is confident that it can survive in the post-CD environment. “CDs effectively never wear out,” he said, “and as new forms of digital audio arrive, people will be ripping and re-ripping for a good long while. I fully expect our disc-recognition service to be running for decades to come, even if not a single CD were sold after today.”
People are, however, moving from unruly piles of CDs to hard drives of MP3 files. Lots of them. Ten years ago, the average consumer had 70 songs. Five years ago it was 1,000. Today, White tells, me, it’s 12,000. Wow. Of course, I have more than 80,000, which brings me to one more issue I have: Today’s cars can’t effectively read a hard drive with that many songs: The indexing function just spirals endlessly.
Infotainment types regard me with amusement when I tell them how many songs I have, but I’m hardly alone — I know dozens of music freaks with libraries like mine. So get with it, guys. “That’s a simple programming issue,” White said.
One more interesting aspect of Gracenote regards public vs. private content
. Gracenote was originally CDDB, and its core database created from many uncredited open-source contributors. The company filed for a patent on that kind of database in 1999, and newly named Gracenote was granted it in 2005. What the pioneers think of their groundbreaking and altruistic work becoming private is unclear.
White says not to worry about the public/private thing. “The reality is that you see this all over,” he said. “Facebook built a multibillion-dollar business based on people entering their data.” True, that. Rock on!