DR Dynamic Range

[ar surround]

And to confuse things even further, here are DR readings for Nights On Broadway:

1) The source stereo file shows a miserable DR of 5
2) However, the Quad that I made from stems taken from that source stereo file shows a DR of 11

And my Quad sounds that much better too. Can anyone please explain this phenomenon?

 

[par4ken]

And to confuse things even further, here are DR readings for Nights On Broadway:

1) The source stereo file shows a miserable DR of 5
2) However, the Quad that I made from stems taken from that source stereo file shows a DR of 11

And my Quad sounds that much better too. Can anyone please explain this phenomenon?

View attachment 84304

I suppose that it all depends on just how the various components add and subtract with each other. Interesting result!

 

[par4ken]

Maybe your more sensitive to DR issues?
In any case your so locked into this position there's no room for debate.
But I will stand my ground when I say there's more to SQ than a simple DR number reveals which shows only the margin between soft and loud. It tells you nothing about inner detail resolution, FR, imaging ablity, etc etc etc.

Exactly, there is no inner detail when everything is the same level! Flat line brickwalling produces a sound character that is downright fatiguing to listen to!

More importantly why do this at all? My vinyl rips are pretty consistent most running with a DR value of 12 (some 11 some 13). The consistent DR value of vinyl shows that compression was used on them as well. The compression used for vinyl was due to the limitations of the medium. The compression used on vinyl was not excessive. The good sounding CDs have similar DR values and show no sign of brickwalling. CDs brought up the possibility/promise of even greater dynamic range. That promise was stolen from us! Absolutely no excuse especially with the even higher resolution digital formats.

I recall an early article in an Audio magazine about recording to CD. In it was recommended to keep levels about -10dB. While that sounds low by analogue standards it makes sense for digital, giving a full 10 dB extra space, plenty of room for the odd peak. Many early CDs have maximum output levels below zero, they sound fine except when compared to their overly loud counterparts.

 

[Sal1950]

That promise was stolen from us! Absolutely no excuse especially with the even higher resolution digital formats.

I absolutely agree with you!
Please don't take my comments to in any way condone the heavy compression being used by some today. It's a travesty!
We need to stop arguing between ourselfs on the listen-ability of these releases and find ways to redirect that energy at the offending labels.
Falling back on 1960s vinyl tech isn't the answer, I haven't been able to enjoy listening to it since I got my first CD player. With it's surface noise, wow & flutter, mono'd bass, inner groove distortions, pops & clicks, other speed irregularities, and all the rest. Vinyl is best left to the collector of archived material and it's supporting gear.

 

[ar surround]

In my vast collection of 'stuff' that I often don't remember that I have, I found a 1996 remaster of The Moody Blues Blue World. This 1996 version has a DR of 10. Although it has the same clarity as the 2008 remaster (DR of 6 - 7), it sounds a bit better especially where the song gets louder.

(Lower two waveforms are the 1996 remaster)

 

[HomerJAU]

Ruin?

Maybe that’s an overstatement, but whenever I listen to Yoshi the low DR is as obvious as dog’s balls, and that’s not a pleasant experience.

 

[Plan9]

The DR meter is especially sensitive to digital limiting and low bass. Add these to any track and watch the DR value go down.

You may have a comparatively a greater amount of analog compression on another mastering of the same track and the DR value will be much higher. Same if you shave the bass below, say, 40Hz.

 

[Plan9]

And to confuse things even further, here are DR readings for Nights On Broadway:

1) The source stereo file shows a miserable DR of 5
2) However, the Quad that I made from stems taken from that source stereo file shows a DR of 11

And my Quad sounds that much better too. Can anyone please explain this phenomenon?

View attachment 84304

Without going into too much technical details... not sure if that will make sense.
The DR meter "sees" the brickwalling on the source = low DR value.
Extract something from the brickwalled source, and by definition it won't have the final limiting that the stereo track has on it.
The DR meter will then be thinking it is more dynamic.
But no dynamic range has been created. You just extracted a sound from inside a dynamically limited source.
Reorganize these extracted sounds and you will have a new mix. By definition a different mix = a different DR value.

 

[ar surround]

Without going into too much technical details... not sure if that will make sense.
The DR meter "sees" the brickwalling on the source = low DR value.
Extract something from the brickwalled source, and by definition it won't have the final limiting that the stereo track has on it.
The DR meter will then be thinking it is more dynamic.
But no dynamic range has been created. You just extracted a sound from inside a dynamically limited source.
Reorganize these extracted sounds and you will have a new mix. By definition a different mix = a different DR value.

Makes sense. I also believe that remixing into multichannel instead of stereo helps: Easier to stuff 5 lbs of potatoes into a 5 lb vs 2 lb bag.

 

[Dynamic Editor]

Brickwalling looks much worse than that. Only parts of that waveform look 'bricked'

Compare to, say, Oasis's What's the Story from 1995. *THAT* is brickwalling.

Oh yeah. That one has distortion all over it. I tried to fix that up here with a declipping algorithm I use. The video below switches between the original release and my edits, starting with my edits.