Great find, Rab. The video is a great visual illustration of the issue, and the article does a great job of detailing the problem. I know I'm guilty of the "make it loud" approach. Question for you: Most DAWs have a "normalize" feature, which attempts to find the loudest sample in a given track, and raise that one sample to the loudest possible signal (without clipping), and increase the other samples a relative amount. Is that an issue? I would think not since the range is maintained. Yes?

yes - informative and brief. I always suspected - a lot of my stereo recordings i gave up treating and they sounded more airy and punchy.
is there any technology that can process the transients quickly enough that you can get punch and loudness? and how about a happy medium where it has that 'elastic' quality but is also simmering and sensitive? - obviously we're not talking about mp3's :)

I'd say it's better to get the recording level right to begin with and make sure the loudest transients in your program material just peak under 0db. This will make sure you fully exploit the bit depth available [particularly when working above 16 bit]. Normalising only pulls your audio up in level so that the maximum peak on a given track reaches 0db. It's not the best idea to record a weak-levelled track then rely on normalizing to get it louder. Better to get it recorded properly in the first place. It often takes a bit of experimentation and some "dry runs" to get better levels (and it can be done). This will pay you back better in fidelity rather than normalizing a track. Personally, I'd rather work with a slightly quieter (non-normalized) native track than a normalised track that started off as a track that was recorded too quietly. Normalizing is useful when used wisely - but it's not the miracle tool that solves all level problems. The ability to normalize can actually make you lazy when it would be better to refine your recording technique and take "proper" advantage of the available bit depth rather than letting a particular tool alter the original fidelity. The "wimpy loud sound" can be achieved not only on a mix, but also on a given track. Do this to all your tracks and the dynamic range in your mix falls flat on its face.

Hi, Ben. You posted just before me.

Yes... we're not talking about MP3's.

You asked: "is there any technology that can process the transients quickly enough that you can get punch and loudness"?

One way to do it would be to use a multi-band compressor / limiter that has a look ahead feature which acts to "predict" the peaks early enough to "react" to them. You basically tune it to the offending high-energy peaks so that they can be either compressed or limited as required. Sometimes these work really well - and other times they make the source material sound a bit unnatural. Let your ears be the judge when using this tool.

On your question: "and how about a happy medium where it has that 'elastic' quality but is also simmering and sensitive"?

There is no real "trick" here. It's just a matter of learning to record properly. With some very basic (and not necessarily expensive) gear, it's painlessly easy to jump to the next level and get that simmeringly sensitive elasticity we all crave :)

I followed the link and found another with an article I found very satisfying. "Singing legend Bob Dylan, 66, said: 'Modern records are atrocious. There?s no definition of anything ? just static.'"

That quote makes me feel much better about what I was hearing. I thought it was just me.

Having said that, I should point out that I think QOTSA makes this work. They somehow pull it off. But every album from the Foo Fighters since Colour and Shape has left me wondering "what the hell"? Which is all very curious considering Dave Grohl produced QOTSA best album.

There is a lot I could say about this. . . and even more I need to learn. Thanks Rab.

Thanks Rab.

Something else struck me on this subject on my commute this morning....TV Commercials. They have to be some of the most guilty parties in squashing dynamics and making everything LOUD.

You're absolutely correct, Paul. You'll notice it particularly well on softly-spoken female advertisment voice-overs. Works well.. far sexier too - which is probably why it's done. :)

I'm skeptical of the fidelity of Bob Dylan's ears at this point.

Ha! Matt, you're right about that!