loudness wars: digital recording to improve?

interesting article here: http://mixonline.com/mixline/reierson_loudness_war_0802/

let's hope the thesis is correct!
This exact scenario played out twice in my living room this month: I introduced a non-audiophile friend to my sytem and played some stuff that I like to listen to so that he could see what it was all about. Then he said, "Oh, man, I want to hear the Black Keys on this stereo." I said okay and let him play his favorite Black Keys CD.

After a minute, the guy looks over sheepishly and says, "Why doesn't it sound good?"

...thus creating a teachable moment about Dynamic Compression and How the Kids These Days are Ruining Music.

Two different friends, same exact story.
Both encouraging and sad. The good news; we get better sound. The bad news; software programmers have to add more technology into the process, because software programmers have better taste than way too many recording engineers, who should know better and be ashamed of themselves. Couldn't the recording engineers just do it right in the first place, without all the added doodads? Go figure?!
Over and over, they say better sound recording is coming.. Well, it turns into worse sound instead.
The guys who master and process the recordings have to get with it, or they will continue to get worse. Though it is hard to imagine how much more harm they could do..
(though I am afraid i will be wrong on that count)
I'm with Elizabeth on this one. To expand on it - just wait till high-res becomes the standard (don't hold your breath though!). More bits and space to brick wall a recording! Imagine how 'loud' Metallica will try to be.

I just wish they'd declare Metallica's last 2 albums the undisputed winner of the loudness war so we can all move on.
How the Kids These Days are Ruining Music.

I think it is more the record companies that are ruining it. They are the ones who actually decide how an album is going to sound, not "the kids". You can probably blame some of the artists as well. On the flipside, there are some really great sounding pop/rock albums released since the turn of the century, which only goes to show that if you have people behind an album that actually care how it will sound on a decent stereo system, you will end up with a good sounding album. If the kids want louder volume there is always the volume knob on all of their mp3 players. Go nuts with it!

And here's a good question - the record companies hate it when people download music for free (i.e. steal). I do too, because I agree it is stealing. I have personally refrained from doing it, but then again why should I have to pay good money for a lousy product? And don't they realize that many of those who still buy CDs do so because they want a high quality recording? There is nothing worse than putting a CD into your player, hitting play, and listening to one big mess.
If the recording is botched and dynamic peaks compressed or clipped in the recording, no algorithm will fix it properly.

Algorithms can change what is there but not put back information that is lost in the first place, a least not with complete accuracy.


All these algorithms do is allow the average listener to listen to a mix of different recordings with different overall loudness levels on a server at more of a consistent volume without having to adjust volume as much. Not sure it will do much more than that for audiophiles.
All I can say is that I really like the new Black Keys CD and would like it even more if it sounded real, instead of recorded for MP3 users. There is so much potential in that CD. Its really a shame what the labels do.
I agree with virtually all that's been said. The current state of affairs saddens me more than I can say. I'd be very happy if compression as we know it now ends. I am very skeptical that that will happen, however. It seems to me Mapman makes a great point. Maybe the guy who wrote the piece will listen to "us"? Who knows...
Another good point is that the recording industry has allowed its legal distractions to take their eye off the target in regards to quality product in many cases.

I suppose that their argument is that there is not much money to be made in making high quality recordings available to the public these days because most people care about cost first and quality second with these things.

"Sad, but True" I suspect. That seems to be how things in general work these days.
Mapman, you might want to reconsider your thoughts on what algorithms can and can't do, especially when given more brushes and more delicate brushes on a bigger canvas. Getting it right the first time is certainly preferable, but being able to offer some level of repair might be better than leaving it broken. I suspect that the potentiality of these kind of offerings, might more often than not, be squashed by business politics rather than inherent failure.

Most likely the results of applying corrective algorithms should be better, but not likely as effective as fixing this data quality issue at the source.

Audiophiles tend to care about these kinds of details more than most which is why I am skeptical of the benefit to that cranky bunch.

Heck, a lot of audiophiles will never accept digital even if done right, much less if botched and corrected somewhat after the fact.
Remember, CD's have been exposed to error correction since it's inception.
"exposed to error correction"?

I don't understand that statement. In any case, compression is an entirely different - and *relatively* new - thing (at least to the degree that it is used these days).
Again the role of error correction is misunderstood.

Error correction is not a band aid to fix undesired screwups. It is a method of handling an expected rate of errors resulting from operating the hardware at a much higher bandwidth than it could handle without errors. As long as the error rate does not exceed that for which the algorithm is designed the output data will be exactly the same whether errors occur or not, and, incidently, will be identical to the original data before it was encoded for error correction. It is not simple interpolation of adjacent data, although that may be done as a fall-back if the error rate is excessive. Not the greatest, but better than aborting the playback.
I didn't misunderstand. I was merely suggesting that things have been manipulated to sound better (or at all, for that matter) for some time. It was not meant to be an exact comparison.
In a sense everything recorded is "manipulated" in some way.

What matters I suppose is whether or not the customers like the results.
Mapman... True.

Manipulation in the form of equalization, compression, peak limiting, echo, etc are intended to change the way it sounds, and are therefore fair game for opinions. Error correction, on the other hand, changes nothing between the original data and the data played back, and so is inappropriate for discussion in context of sonic quality.
"Error correction, on the other hand, changes nothing between the original data and the data played back, and so is inappropriate for discussion in context of sonic quality. "

Eldartford & Mapman:

Agreed as well
I suppose you like the sound of: click,click,click,click rather than what the artist recorded? If the original data is compromised, it can be fixed.
This war will never end and I'll tell you why in one simple word:


Tracks that are compressed will always sound louder than those not compressed.

Compare an orchestral piece like Mahlers 3rd or a live recording to a typical Pop studio recording. If you look at them with DAW software you will see that all three have peaks that meet 0dB. However, the studio recodring with compression will have significantly more energy in the audio band than the other two non-compressed recordings.

This is not likely to change. Here is why: Studio Pop recordings are intentionally compressed to sound good on boomboxes and car stereos that have compromized sonics. The orchestral recording is made primarily for audiophiles, so it usually have no compression and maybe even no EQ applied.

Steve N.
Empirical Audio
More likely redo without compression from any masters available. Assumes that the band was doing something other than just making noise.
If you want examples of the same track with compression and without, go to bluecoastrecords.com and download the same track in 96 and then 44.1.

Steve N.
Empirical Audio