News

I've had a lifelong love of and involvement with audio, and have often been frustrated by the lack of quality. The following are some real-world examples of the struggle for quality sound from my personal work.

Marie-Lynn Hammond, the Toronto-based singer-songwriter I'm working with, had released five albums by the time I met her. I've re-released albums 1 and 2 on one CD, made from the original master tapes, and although the original masters were 22 and 17 years old, the CD has received positive "sound quality" reviews. These two albums were recorded in high-quality studios and were originally released on LP.

Albums 3 and 5 were released only on cassette and were lower-budget recordings (but the songs are still great). I've found something approaching the masters and will be re-releasing those on CD, but the sound quality is not as good as albums 1, 2, and 4, although they were made later. Album 4 was recorded at a high-quality studio and released on CD. I obtained the original 2-track master tapes and was amazed at the difference in sound between the 1990 CD and the 2000 CD-R that I made from the original master tapes. The difference was immediately obvious on a $250 shelf system, as well as on my home systems.

So, what were the problems with early CDs?

First, there have been improvements in players. A colleague of mine thinks that there is still a way to go, but I'm pleased with what I hear.

Second, digital mastering tools and techniques have improved greatly. The example of Marie-Lynn's album 4 is indicative of the changes in just 10 years—and my A/D converters aren't the latest, slickest, and most hyped, but rather a set that has a good, solid reputation for "sounding good."

Third, many CDs were rushed to market and were made from protection masters or worse. No thought was given to how to make them sound good for CD.

Examples of this from my collection abound. Laura Nyro's Eli and the Thirteenth Confession CD sounds as bad or worse than the LP, with the typical-for-the-time "we make our LPs sound good on $30 phonographs" equalization (thanks, Columbia). It sounds like no remastering from the LP version. Judy Collins's first compilation album, Colors of the Day, has been recently re-released in a variety of formats, including the original horrid-sounding CD, revised HDCD version (don't credit HDCD with the improvement; it sounds better on a normal player), DCC Gold Audiophile CD (note the remastering credit . . . hmm again, more care), and even a DCC remastering on LP! I think the DCC sounds a bit better than the stock HDCD Elektra, but it's the remastering, not the medium. Same thing with Joan Baez's Diamonds and Rust on gold disc—it was more carefully made. There are other horrid examples in the Joan Baez discography where the tapes from which the CDs were made had deteriorated from the time the LPs were made. David's Album is perhaps the worst.

A compilation of Lynn Anderson's early recordings on a compilation CD also sounds absolutely horrible compared to the LP. Remember "I Never Promised You A Rose Garden"? Well, that has an exciting bass line on the LP that's just plain gone from the CD. I suspect the LP mastering engineers were topnotch and made something much better in the mastering process, and the same wasn't done for the CD. I'm sure there are other explanations as well.

If Marie-Lynn and I had been doing the re-release of her albums 1 and 2 on a budget and schedule, we would not have spent months looking for the original master tapes of album 2 and would have done the re-release from the safety masters. But that would have been a big mistake! The improvement in sound quality was remarkable. Also, I tried three different dbx decoders before I was happy with the sound of album 1.

My point is that there are many steps in the process. The 16-bit/44.1kHz technology is not the best that is available now, but it is entrenched and was a stretch at the time it was developed. I also contend that it is not bad, and is perhaps even pretty good. There are very few scenarios in which the 96dB dynamic range provided by 16/44 is appropriate. I certainly can't easily reproduce that range at home, and would get complaints from my family if I was reproducing peaks at 111dB SPL (assuming my background noise level is 15dBA, which I suspect it is with air-conditioner and appliances off and no dogs barking or car alarms going off).

16/44.1 can sound a heck of a lot better than most early CDs, not because there were flaws in the basic system, but because the system wasn't used properly and the equipment was not fully up to what it is now capable of. Oh yes, forget a –20dBfs reference—no one uses it. Many of the early CDs tried, and they are horribly low in signal level, throwing away 8–12dB of dynamic range. We now normalize all CDs for the loudest peak to be 0dBFS, which was hard to do pre-computer, but now it's as easy as highlighting the project and hitting the "N" key in Samplitude.

Remember, sound quality (or pixels) is not what sells music (or movies). Marketing and the music itself sell music. Remember that, although theoretically LPs sounded much better than cassettes for those of us who cared, at some point in the mid to late '70s cassettes outsold LPs. Then the CD came around and people could have quality that we only dreamed of in the LP days—and could have it for the long term with frequent use. But now the public is listening more and more to MP3s, which are a step down in quality from the CD but a step up in convenience.