Why do old audio CDs sound duller?

I have quite a few hundred CDs collected over the last 20 years or so. Recently I put everything I had onto my iPod via iTunes. Now when I listen back, the older CDs are definitely quieter and seemingly more lifeless. So what gives? Surely my computer is reading digital information so shouldn’t the information remain unchanged? Are bits being lost as the CD ages?

Are they of music that was also released on vinyl? Quite possible it was originally recorded on analog tape in the studio, then digitized. The quality of conversions was pretty variable when CDs first came out - I’ve heard two different CDs of the same album that sounded very different. So they probably sounded lifeless when you first bought them too.

There’s always research (can’t find a cite right now) that shows all else being equal, listeners often think louder music sounds better. There’s a trend that gradually increases perceived volume, either maximum volume or normalizing an entire track upwards so that the average volume is louder, such that maybe the music you’re hearing today actually is louder and more defined in the sense. Kinda the Hollywoodization of music.

I was considering this. Most of my early CDs were from contemporary artists, but it makes sense if they always sounded that lifeless and only now have I started noticing. But many of my old CDs are considerably quieter on my iPod compared to newer releases.

Reply: I’ve noticed this with classical music as well, but it could be a number of factors such as dynamic range and digitising technology. I have read contradicting things about the loudness war, and I’m not sure it entirely accounts for the differences I’m noticing. Less dynamic range should not mean the entire album would sound quieter (even its peaks). I was wondering if the CDs themselves had an effect somehow.

I’m not sure what you are hearing, but this I can say with certainty: the bits are not being lost (or even changed) over time.

Error correction software in CD players will not let a block of sound pass if any bits which cannot be read are not recovered. If they are recovered, they are 100% perfect from the original recording. That’s the charm of sophisticated error-correction schemes, and why about 1/3 of the data is reserved exclusively for error-correction data recovery, not actual sound.

Caveat, but just for nitpickers: While a data CD requires 100% reconstruction if data is read imperfectly, audio CD circuitry has a little more leeway. If a fraction of a second of music cannot be reconstructed from an error reading, a tiny slice of the previous sound may be substituted. This is so little, and hopefully so infrequent, that the human ear cannot detect the repetition of the slice. But it does mean that the resulting output might not exactly match the original.

Nevertheless, even if slices of sound are frequently substituted, this has negligible effect on things like sound levels and frequency response. About all that might result is a slight distortion.

My conclusion is what you are observing is changing styles in recording concepts, frequency response, mixing factors, equalization styles, reverberation factors, microphone characteristics, instrument placement, etc.

Don’t overlook the possibility that what has changed is you. My speakers don’t sound the same as they did 30 years ago, but that’s me, not the speakers. Get an audiometer test to be sure!

Interesting, thanks Misicat. The only thing I’ll say is that new CDs of old recordings do sound different but they don’t sound quieter. But perhaps they’ve been remastered with modern techniques.

Remastering is probably what you are hearing. You can but alot more audio on a CD than tape but mastering techniques took a long time to catch up. I have a copy of “Exile on Main Street” remastered by Bob Ludwig that you can hear things just don’t exist on the original, background vox, guitar parts etc.

Compression has quite a bit to do with this, it makes the mix punchier, louder and overall less dynamic.


To further add to Musicat’s point about error correction. Some stats were done a while ago on the frequency of non-correctable errors on CDs. Most CDs play with none at all, and those with errors, have a one only every few minutes at worst. Basically there is essentially no chance at all that there is any sonic change attributable to data loss.

The loudness wars really do account for a change in apparent level. It isn’t just that the levels are compressed, but they are all pulled up to maximum. So much so that a quite real result is actually commercially released recording that are provably clipped at digital maximum. Needless to say this doesn’t help the sound, but it seems there is less and less care in many areas.

The first few years of CDs saw CDs cut directly from the masters used for LP disk cutting. This wasn’t smart as by then the recording had been mixed with the intent that it go to vinyl. So it had, by design, dynamics that avoided the problems inherent in vinyl.

CD production has brought about the invention of a new person in the chain, the mastering engineer, whose only job is to get the recording to sound right on the delivered media. Sadly this guy is often asked to compress the hell out of a recording, but for high quality music releases his job will be to tweak things so they sound as good as possible. This is a more recent thing, and early CDs were cut directly from the master produced by the mixing engineer.

(The word “master” is now so totally overused to be almost useless. In principle there is only one “master” recording - but now everyone seems to want to be the one creating a master. So “remastering” can mean a range of things.)

If that’s what we’re talking about, there are many factors involved.

Just to illustrate a simple one…stereo separation of L/R tracks on LPs wasn’t the best. When CDs came out, the separation was equal to the original, which was much better than the LP versions I was familiar with.

Listening to some CDs of Mamas and Papas tunes revealed some things that made me go, “Hmmm…” since I was able to detect some stuff in one channel that I never heard before. In particular, a flute track that seemed “out of place” in the mix. I could hear it quite clearly if I turned off one channel and turned up the other.

So I went back to the original LP, and extracted the same channel from the same song. The same bogus flute performance was there, it just wasn’t as prominent, and I never noticed it before!

My conclusion was that the relatively poor stereo separation in LPs covered up a multitude of sins that CDs revealed for the simple reason that CDs didn’t have any crosstalk, and LPs had not only mild crosstalk, but poor separation.

I was able to confirm this in another Mamas & Papas track, but that’s another story. You’ll have to ask to get it. :slight_smile:

So to say it in one, the old CD’s were mastered with a lesser dynamic range so that final result was ready to go to CD, vinyl record, tape . .and also to be played on the AM radio.

These days mastering is better, it keeps the full dynamic range as its allowed by the CD, and reproduced by hifi 's, and the radio stations will compress (to reduce dynamic range) before broadcast (lest they only broadcast half the track… the noisy bits. You know, like when you tune to a distant station and you only get the drum beats…)

As a CD is played a highly focused laser tracks the disc’s “pits” and which encodes the music onto the CD. Over time, the playback laser, since it is so highly focused, wears down the edges of those soft polycarbonate pits and rounds them over. Eventually, this causes the music to sound dull and lifeless.

Our Patented DiscRevitalizer[sup]TM[/sup] process resharpens the edges of those pits and restores the brilliant and sparkling sound that you used to enjoy.

Simply send your disc to us along with your credit card order (checks and money orders also accepted) for $29.95 and we’ll perform our Patented DiscRevitalizer[sup]TM[/sup] process to your disc.

Pit edges can be resharpened indefinitely, so after the first sharpening and when the edges wear down again, we’ll resharpen the pits for only $19.95 per disc!

Avoid future disappointment and restore those old discs today.

Shut up and take my money!

It would be funnier if you weren’t already 25 years late. There were guys that would cryogenicly treat your CDs to do just this. Along with your super special cables, tubes, connectors, pretty much anything they could convince you to freeze. It was, and is, altogether a bid sad.

But Patented DiscRevitalizer[sup]TM[/sup] is all new! Different from common cryogenics!

Check your ripper log. Programs like DBPoweramp and ExactCopy (EAC) logs let you know about read errors on a disc. A badly scratched or dirty disc can get enough errors to be noticeable.

When CD players were first released, they were plugged into home stereo systems. You youngsters don’t remember the joy and science of shopping at a stereo store to put together the best sounding system you could afford. Now they’re tweaking stuff so it’ll sound good on ear buds, not large Advents, JBL L-26s, or JBL 4310s.

Interesting stuff - thanks all.

Are you aware that, unlike analog media, the dynamic range in a digital medium has a maximum that cannot be exceeded (you can’t store 16 bits in a 15 bit word)?

Are you familiar with normalization? Ideally, all CDs were mastered post-normalization. To NOT do so would waste available dynamic range and sound quality. You would have to extract the digital data from each one to prove that it hadn’t been.

Or are you talking about compression? Styles in compression do change, and the result of one mix can sound louder than another, but it still cannot exceed the maximum allowed for the digital word size in use.

I’ve heard that they do deteriorate over time and that error correction methods sometimes can’t compensate, but that is addressed above in another post that opines that it would not cause the effect you speak of.

Older CD’s usually had a rating, something like:
Where A meant analog and D meant digital and was what step in the recording the music was digitized. AAD meant that the 1st 2 steps were analog, something like the master recording + the copies - the CD’s were made from those copies, ADD meant something like only the master was analog, the copies were digital and that used for the CD’s, and DDD meaning all 3 were digital. Perhaps that is part of it, perhaps not.