By Roger Nichols
I thought seriously about changing professions last week. How about tourist helicopter rides in Maui, or maybe just a hot dog cart on the beach in Miami. What about a house pet massage therapist in Montana, or a storm chaser in the Midwest? I could sell Pope-sicles in front of the Vatican in Rome. Yeah, that’s it!
Two weeks ago, I received a test pressing from the CD plant that is pressing the new Steely Dan CD. The record company asked me to approve it before they pressed 72,000,000,000 of them. I was staying in a hotel at the time, and decided to play it back on my portable CD player and listen on my Sony MDR-7506 headphones. Before the vocal came in on the first song, I thought that it didn’t quite have the clarity and sheen I remember from the mixing last summer. Maybe it was my CD player and I would have to listen the next day at the studio. In my suitcase, I found a CD reference disc from the mastering session. I put it in the portable CD player and listened to the first song up through the entrance of the vocal. It was perfect. Exactly what I remembered and ten times better than the test pressing. The next day, I listened at Sony Studios in New York with Elliot Scheiner and Dave Smith. The difference between the two CDs was amazing. I remember some friends talking about a project that they were working on in Japan. Whenever there was a problem, the boss would say, “Who is fired?” I was wondering that same question.
At first I thought that the pressing was so bad that error correction was causing the difference in sound. Dave Smith at Sony ran the test pressing through an error checker, and it came back perfect.
Maybe I didn’t have the latest ref before the masters were sent to the record company. Maybe there was a screw-up in the transfer to the master tape used for the CD plant. I assumed that the test pressing was an accurate representation of the master they used. I was going to be wrong. I verified that the CD ref that I had was made from the exact master tape that went to the CD plant. I had a new ref made from the same data file on the Sonic Solutions to verify that my CD ref was accurate. It sounded exactly the same as my CD ref.
Don’t tell anyone this, but during the mastering we turned up the overall level of the whole album 6 dB. It was exactly a shift up of 1 bit. No math would be performed on the data to raise the noise level, but, because of it, there were a few overs on loud passages during snare drum hits. We listened carefully and decided that, since the overs were not audible, we would leave them alone. We tried using limiting, but you could hear a slight change in the sound. We sent an over list along with the master to the CD plant.
The Overcompensation Process: If a CD plant sees overs, they will usually contact the source of the master and request an approval before pressing CDs. Lots of artists want their CDs to be so loud that you can listen to the CD without a CD player. Cranking the level up with compressors and limiters is just not enough for them, so they goose the level up even further until the over lights come on at every boom or smack of a backbeat. To keep the CD plant from going ballistic, the mastering house makes a copy of the audio and turns the overall level down by .1 dB. This assures that the chopped-off wave forms that result from digital clipping do not show up as overs when checked by the CD plant. Copy the audio from a Brandy CD into your waveform editor and look at it once. The waveforms look like square waves because they have been chopped off so drastically.
Sometimes the CD plant (they usually ask first, but who knows) will run the master through a processor that drops the level down .01 dB, thus eliminating the over problem. I asked the CD plant if they did this to our master, and they said no. I suspected a falsehood, and decided to check for myself. I read all of the audio data from my CD ref (the good sounding CD) into my computer. I also read the audio in from the test pressing. I ran a program that compares every sample of one audio file with every sample of another audio file, and produces a third audio file containing the difference between the two files. The computer said, “The file you have created contains only zeros.” This told me that the audio data was exactly the same on both CDs. Then why did they sound so different? We are talking a big difference. Everyone could hear the difference. My MiniDisc copy sounded better than the test pressing.
It Was Jitter! The plant used the correct master, cut the glass master at 1X, and, as far as I could tell, did everything right. But the CDs sounded different…unless you played them on a $10,000 CD player. Then they sounded the same. I copied the audio from the bad pressing into my computer and burned a CD-R from the audio file. The CD-R sounded just like the ref. We thought that maybe the CD ref was wrong. Maybe the test pressing was right and something about the jitter in the CD-R burners made the CD sound better than what it was supposed to sound like. I had the mastering facility make a DAT tape from the same Sonic Solutions file. I played the DAT tape and the CD ref through the same Apogee PSX-100 converter. They sounded exactly the same. That
meant that something had to be wrong with the LBR (Laser Beam Recorder) that was used to cut the glass master, or something in the plating or pressing process had to be causing the sound to change. HELP!
We sent new masters to the CD plant — a 1630 tape and a PMCD master. The plant pressed test CDs from the 1630 and the PMCD feeding three different LBR glass master machines. We had test pressings coming out of our ears. They all sounded different, but none of them sounded as good as the original CD ref.
I called Glen Meadows at Masterfonics (now Emerald) in Nashville. He told me that many of the producers in Nashville were having similar problems with CD pressings, so they did some tests. They sent master tapes to the JVC plant in Alabama on Exabyte, 1630, and PMCD. They cut glass on two different systems, one of which was called the K2 system. The pressings done on the regular LBR sounded different than the original masters. The K2 glass masters (special low-jitter system) sounded the same as the original masters. Some record labels in Nashville who own their own CD pressing plants have their CDs pressed at JVC using the K2 system.
We called the record company and the proverbial sh*t hit the fan. No way — not even for Steely Dan — were they going to allow the pressing to be done at JVC, period. Well, I guess that answers that!
The only possibility was to find a place that did custom high-quality glass mastering that was compatible with the plant the record company was using. We were in a crunch because the CD was supposed to hit the stores February 29th. This meant that, on Friday February 11th, the CD plant had to press up at least 100,000 to pre-ship to distributors for the small record stores so that everyone would have the CDs for sale on the same day. After some searching, I found out that the Denon CD plant in Georgia does custom glass for many customers. BMG Classical uses Denon to press their CDs. Denon is a high-end audio company, and I would expect nothing less from them. I contacted Denon and they agreed to rush us a test pressing to evaluate. I sent them the Steely Dan masters, and two days later (last Friday, February 11th) I got the test pressing. We listened to it, and it was fine. It sounded like the ref and we were good to go. The metal parts were sent to the record company’s CD plant and pressing started.
The CD plant is supposed to send me a pressing from the Denon metal parts on Monday, February 14th so I can see if it sounds as good as the test pressing right from Denon.
Hang on — here is FedEx with my test. I’ll be right back with the results… Yeah! Whoopie! It was perfect! Somehow the problem was in either the glass masters or the plating process afterwards that makes the stampers that press the CDs.
Well, at least I don’t have to change jobs just yet. I will be doing further investigation into the problem and let you know. Just remember, all of the expensive test gear said the pressings were perfect. The quality control people at the record company said we were crazy, the test equipment said the pressings were perfect. When asked if they listened to the CD pressing, they said, “No, do we need to?” Still, the best piece of test gear is your ears. Use them and protect them. Cool.