DR, Brickwalling, Fidelity, and Perceived Loudness

[Bill B]

Actually, it is to get louder play on the radio. They think that sells more CDs.

I've heard it is also to sound good on cheap earbuds.

 

[par4ken]

Actually, it is to get louder play on the radio. They think that sells more CDs.

That is totally unnecessary as radio stations use their own audio compression.

 

[JediJoker]

That is totally unnecessary as radio stations use their own audio compression.

And, therefore, the less-compressed version going into the broadcast processor will sound much better coming out of it than the more-compressed one.

 

[MidiMagic]

That is totally unnecessary as radio stations use their own audio compression.

They do now. They didn't in the early days of CDs.

 

[par4ken]

They do now. They didn't in the early days of CDs.

They always used compression, it was more important in the analogue days. Without any compression there would be worries about over modulating/deviating the carrier causing not only distortion but interference to other services. Also the desire to sound louder than the competition. Stations that cared about sound quality would use the minimum amount of compression necessary.

 

[JediJoker]

They always used compression, it was more important in the analogue days.

Yep:

1975 - OPTIMOD 8000 audio processor introduced for the new FM format. Bob Orban and partner, John Delantoni, set up Orban Associates as privately held company.

 

[Plan9]

It would take an awful lot of filtering to alter the DR value. Sorry but I don't understand the point that you are making. DR value is a measure of the average difference in level, low level to high. Brickwalling flattens the level, so the less variation the lower the DR value. Extra A/D and D/A conversions should not change the DR value.

The vinyl that I have digitised myself usually show DR values of 9 to 16, with 12 being by far the most common. LP's always sound good. I'm a believer in DR readings as every CD that I have ever thought didn't sound right proved to have a very low DR value.

I would open the offending CD in an audio editing program and immediately see the brickwalling. That was before I took stock of DR readings at all, but since have found that the DR value and sound quality are closely linked. The brickwalled CDs all had a low DR value. Audiophile CD's usually run about a 12, the same as the vinyl.

What angers me most about this is that with the introduction of the CD we were promised greater dynamic range than what we were getting with vinyl. Instead we get less in the name of loudness, or to appease the smartphone user!

The problem with the DR meter is that it is extra sensitive to infra bass and limiting. Take the same CD track after a D/A conversion (no flat-top from the limiter anymore) and a high-pass filter @25Hz (both of which shouldn't affect the dynamic range too much, right?) and you gain 1-2 DR points (or more). It is like it analyzes the waveform visually. That why you usually get 1 DR point more when comparing the same master lossless vs lossy compressed (lossy encoders create stray peaks above the limiter's limit).
It is also extremely variable for vinyl readings. One stray click and you will gain 3 DR points on a particular track. Same if you have a high-pass filter on your rig to avoid rumble. It is good when comparing vinyl rips from the exact same system with perfect pressings but less useful when comparing different rips.

 

[barfle]

They always used compression, it was more important in the analogue days. Without any compression there would be worries about over modulating/deviating the carrier causing not only distortion but interference to other services. Also the desire to sound louder than the competition. Stations that cared about sound quality would use the minimum amount of compression necessary.

I worked on a compressor/limiter at Altec Lansing in 1973. And it replaced an earlier model.

 

[Stephen W Tayler]

They always used compression, it was more important in the analogue days. Without any compression there would be worries about over modulating/deviating the carrier causing not only distortion but interference to other services. Also the desire to sound louder than the competition. Stations that cared about sound quality would use the minimum amount of compression necessary.

Funnily enough in 1979 I was living in Chicago when Pink Floyd's 'The Wall' was released - a local radio station played it continuously for 24 hours - it sounded amazing!! When I bought the LPs I was really disappointed it didn't sound anything like as good - luck of the draw with radio compression!!

 

[Dynamic Editor]

Funnily enough in 1979 I was living in Chicago when Pink Floyd's 'The Wall' was released - a local radio station played it continuously for 24 hours - it sounded amazing!! When I bought the LPs I was really disappointed it didn't sound anything like as good - luck of the draw with radio compression!!

The thing about The Wall is that it's so dynamic that it sounds abysmal in loud environments, with the quieter, more subtle aspects of the mix dashed by the sound around you. It's the same with Roxy Music, to an extent (at least, in my opinion).

 

[par4ken]

It is also extremely variable for vinyl readings. One stray click and you will gain 3 DR points on a particular track. Same if you have a high-pass filter on your rig to avoid rumble. It is good when comparing vinyl rips from the exact same system with perfect pressings but less useful when comparing different rips.

That is why you should start with "clean vinyl". Large clicks are very visible and easily removed. My vinyl rips are very consistent, most DR12.
Even if the vinyl is 1 or 2 points higher than CD due to anomalies, the results are close enough to be meaningful. Bass is often limited on vinyl releases which should result in a lower DR value.

The bad digital releases (of which there are many) have consistently low DR values. To me they are unlistenable!

 

[JulesRules]

DR value is a measure of the average difference in level, low level to high. Brickwalling flattens the level, so the less variation the lower the DR value. Extra A/D and D/A conversions should not change the DR value.

Even converting a file from WAV/CD format to mp3 often changes the DR by one point. Once clipping is involved, all bets are off. There's all kind of evidence that the cutting, pressing and playback of vinyl introduces enough steps that the DR of the underlying files isn't represented anymore. We've talked about it in this thread:
https://forums.stevehoffman.tv/thre...te-working-again.1105885/page-2#post-27915691https://forums.stevehoffman.tv/thre...te-working-again.1105885/page-5#post-29626601I won't find it now but there's a posting somewhere that shows how the LP of Rush's Vapor Trails, when you rip it, has "diagonal cuts" in the spots where the CD was clipped hard. So, the DR meter won't detect it, but the damage is still there.

 

[ar surround]

The thing about The Wall is that it's so dynamic that it sounds abysmal in loud environments, with the quieter, more subtle aspects of the mix dashed by the sound around you. It's the same with Roxy Music, to an extent (at least, in my opinion).

In the last few months, I've learned that too much dynamic range can be almost as much of a problem as too much compression. I was making my own 5.1 of a rock song from the 60s where I moved the center (almost) mono drums to the right channel to maintain a good balanced 5.1 mix. However, after doing so, the drums were buried in the mix; and I simply was unable to make them loud enough without employing a good bit of compression.

 

[barfle]

In the last few months, I've learned that too much dynamic range can be almost as much of a problem as too much compression. I was making my own 5.1 of a rock song from the 60s where I moved the center (almost) mono drums to the right channel to maintain a good balanced 5.1 mix. However, after doing so, the drums were buried in the mix; and I simply was unable to make them loud enough without employing a good bit of compression.

One thing to remember about dynamic range is that most people have limits in their hearing that is less than the electronic limits we have these days. I’ve attended classical concerts where my tinnitus competes with a soloist, and even though my room is pretty quiet, jets and motorcycles sneak in from time to time.

It’s just as possible to be too soft as it is to be too loud. Artistic intent is worthy of respect, but don’t be a slave to it. Practical considerations happen, too.

 

[MidiMagic]

In the last few months, I've learned that too much dynamic range can be almost as much of a problem as too much compression. I was making my own 5.1 of a rock song from the 60s where I moved the center (almost) mono drums to the right channel to maintain a good balanced 5.1 mix. However, after doing so, the drums were buried in the mix; and I simply was unable to make them loud enough without employing a good bit of compression.

I always place the drumkit in the center so that both speakers aid each other in reproducing the lowest frequencies. The bass guitar is often placed there for the same reason.

This also carries over to QS and SQ mixes for the same reason. The front channels of most matrix systems (not BMX, H, and UHJ) are kept in phase to each other in the encoding process, so the strong low bass is maintained. Any other panning will weaken the bass.

 

[ar surround]

I always place the drumkit in the center so that both speakers aid each other in reproducing the lowest frequencies. The bass guitar is often placed there for the same reason.

This also carries over to QS and SQ mixes for the same reason. The front channels of most matrix systems (not BMX, H, and UHJ) are kept in phase to each other in the encoding process, so the strong low bass is maintained. Any other panning will weaken the bass.

Had I kept the drums in the center on that particular mix, I would have ended up with a 'dead' RF channel. Thus I moved the drums to the right and juiced them up with compression.

 

[Dynamic Editor]

Had I kept the drums in the center on that particular mix, I would have ended up with a 'dead' RF channel. Thus I moved the drums to the right and juiced them up with compression.

Shh... You're not supposed to use the word "compression" in an audiophile forum. It scares the wildlife.