Skip to main content ↑ ↓

Notice

Please note that most of the software linked on this forum is likely to be safe to use. If you are unsure, feel free to ask in the relevant topics, or send a private message to an administrator or moderator. To help curb the problems of false positives, or in the event that you do find actual malware, you can contribute through the article linked here.
Topic: From the mouth of Billy Corgan... (Read 23483 times) previous topic - next topic
0 Members and 1 Guest are viewing this topic.

From the mouth of Billy Corgan...

Reply #50 –
Quote
When we sent the new GUITAR WOLF record to the mastering lab for inclusion in our recent in-store play sampler, the mastering engineer called back, mystified by the volume level on the CD-R. The levels exceeded the theoretical maximum possible on compact disk audio. In other words, JET GENERATION is the loudest CD in history.

This story proves that if you rock hard enough, you can break the laws of physics. And Tokyo's premier punk trio definitely rock hard. In their most brain-pummelling album yet, GUITAR WOLF attempt to reproduce the legacy of Johnny Thunders, Joan Jett and the Ramones and make them much much louder and more distorted. There is no doubt that they succeed.
Hehe... I have this record, and it's really sick, Zwan comes nowhere near.

Here's the proof:
[span style='font-size:14pt;line-height:100%']07-Roaring Blood.mp3, Track gain -18.1[/span]

Am I the winner so far? I'm gonna check out some Jap-noise records later on...
EDIT: I just saw clima mentioned this band a few posts up.
EDIT by rc55 (removed swearing)

From the mouth of Billy Corgan...

Reply #51 –
Quote
When we sent the new GUITAR WOLF record to the mastering lab for inclusion in our recent in-store play sampler, the mastering engineer called back, mystified by the volume level on the CD-R. The levels exceeded the theoretical maximum possible on compact disk audio.


You've gotta love marketing BS!

The theoretical maximum signal is digital full scale, and it's not a theory, it's a fact.


How much perceived loudness you can generate depends on how to go about hitting this maximum.



If you're judging the loudness using Replay Gain, then a simple bit of reverse engineering should allow you to get the "theoretical" loudest signal ever. With mpc and mp3 there's no limit, but for a .wav there must be one - I haven't calculated it though.

Looking at the equal loudness filter in replay gain, lots of noise around 3-4kHz will give the greatest calculated perceived loudness. Maybe just a 3.5kHz square wave will do the job - I've just crashed CEP trying, so maybe it really is the limit

Cheers,
David.

From the mouth of Billy Corgan...

Reply #52 –
Maybe unrelated comments :

Music seems to follow the path initiated by the cinema industry,
louder, louder, louder...
If i had to classify my CDs by release date,
I'm sure I can have a good guess only by watching the volume level I use to listen each CD.
La forme tend à dominer le fond...

I like the work from Billy Corgan, but noise is not music,
I doubt something that can destroy my speakers will please my ears.
But, hé, music is often subjective, it's only my opinion
and the artist's freedom should not be forced.

From the mouth of Billy Corgan...

Reply #53 –
Quote
You've gotta love marketing BS!

Yup yup, the
Quote
The levels exceeded the theoretical maximum possible on compact disk audio.
statement is pathetic. But the cd still sounds quite loud.

Edit by rc55 (removed swearing)

From the mouth of Billy Corgan...

Reply #54 –
Clima - your gv.zip link is broken


2B - heh..  I had thought that someone could reverse-engineer ReplayGain and generate the loudest sound ever.  I don't know squat, so would look to someone clever to do it


About the Guitar Wolf hype..  how can you all be sure they didn't send in a demo cd full of 17 bit PCM files?  They would easily go way beyond what would ever be expected from the theoretical maximum of a cd
< w o g o n e . c o m / l o l >

From the mouth of Billy Corgan...

Reply #55 –
Quote
Looking at the equal loudness filter in replay gain, lots of noise around 3-4kHz will give the greatest calculated perceived loudness. Maybe just a 3.5kHz square wave will do the job - I've just crashed CEP trying, so maybe it really is the limit

That reminds me of the story of the killion: a number so large that, if you see it, you die instantly. We need to warn these mastering engineers that hundreds of thousands of listeners could be killed by their deeds!

From the mouth of Billy Corgan...

Reply #57 –
Quote
Looking at the equal loudness filter in replay gain, lots of noise around 3-4kHz will give the greatest calculated perceived loudness. Maybe just a 3.5kHz square wave will do the job - I've just crashed CEP trying, so maybe it really is the limit

I tried a lot of different combinations on that range and the highest attenuation was: -24.06 dB on a square sweep from 3000Hz to 4000Hz.

When I compressed it with different encoders I got these results:

MP3 --APE:
RG: -23.87 dB
Peak: 49569

MPC with --xlevel --quality 5:
RG: -24.09 dB
Peak: 53320.08

Ogg Vorbis --Quality 10:
RG: -16.80 dB
Peak: 35575

So I dare to say the biggest perceived gain based on RG calculations would be ~113.09 dB.

On the other hand the absolute minimum for RG calculations on a 16bit wave would be a wave with -1,+1 bit peaks, (-90.3 dB) in that case the calculation is +64.82 dB, which means the wave percieved loudness is 24.18 dB

So, to sum up, RG goes from -24.09 to + 64.82 dB.
"You have the right to remain silent. Anything you say will be misquoted, then used against you."

From the mouth of Billy Corgan...

Reply #58 –
Quote
Music seems to follow the path initiated by the cinema industry,
louder, louder, louder...

I'm not sure if I agree with this.  Unlike CDs, film soundtracks have a standardized volume that is consistant with almost every movie.  There are exceptions, however, and trailers are generally recorded at a higher volume.  A fader setting of "7" on a film sound processor corresponds to 85 dB for the quieter dialogue portions of nearly every film.  Loud explosions should be at about 105 dB.  This is the recommended setting, but it is probably too loud for most theatres due to poor equipment and acoustics.  Actually it would be kind of nice if the music industry followed specifications such as this.

From the mouth of Billy Corgan...

Reply #59 –
ephaestous, that's a cool post!

It reminds me how simplistic replay gain is. A 3.5kHz full scale square wave (which, generated digitally in CEP, actually has a perceived pitch of 700Hz due to aliasing), doesn't sound as loud to my ears as a really distorted track - however, RG always thinks that the 3.5kHz square wave is louder:

square wave: 113dB SPL (track gain -24.1dB)
distorted track: 106.9dB SPL (track gain -18.1dB)


There are other replay gain algorithms out there that are more intelligent (Frank's worked on some, and someone else emailed me about one once), but I don't think they're used anywhere yet. I really need to find the time to look at replay gain again!

The 20dB "silence" on a 16-bit recording is about what you'd expect when following the SMPTE RP200 standard that RG is based on. 20dB SPL is very quiet (sound "proof" room kind of level). It's often said that you need 20 bits to store the full dynamic range, and this is correct if you want to allow signal peaks well over 100dB, and also acheive really silent silence! This doesn't seem relevant when listening to most music!


To my ears, movies are getting louder. Or maybe cinemas are installing louder sound systems, or just ignoring the calibration and turning up the gain. Also, there is meta data (similar to replay gain) built into the AC-3 format which may be used to make things louder without actually compressing the audio signal - maybe studios are playing with this data, or even setting it incorrectly?


Cheers,
David.

From the mouth of Billy Corgan...

Reply #60 –
Quote
Actually it would be kind of nice if the music industry followed specifications such as this.

Yep. Imagine the party that would happen here if someone actually released a CD labelled "Mastered to ReplayGain Loudness Standard" 

clima: I listened to your file, it sounds pathetic! But for the most part, that type of distortion is intended by the artist and obviously not a result of mastering (part of it must be though, at this volume level).

Here's a clip with loads of clipping which is entirely a result of bad mastering.

From the mouth of Billy Corgan...

Reply #61 –
Quote
Yep. Imagine the party that would happen here if someone actually released a CD labelled "Mastered to ReplayGain Loudness Standard" 

Hehe, that's actually what I'm aiming to do with the next one I'm working on..  I wanna fill the inside cover with nerdesque audio stuff    The first two tracks have a  replaygain -3dB, and need speakers with at least 30Hz-20kHz to hear some of the details
< w o g o n e . c o m / l o l >

From the mouth of Billy Corgan...

Reply #62 –


no comment... this is the worst example I have (Metallica - (1998) Garage Inc. CD 2)
The whole CD looks like one giant, filled rectangle.

From the mouth of Billy Corgan...

Reply #63 –
I've been toying around lately with dynamic compression and some other stuff to get really loud music and get a high replaygain score.  The highest I managed was -17.6 db with foobar's RG. But then the audio sounded absolutely crappy. What surprises me, is that it is possible for the mastering engineers to get to similar levels and still have the audio sounds half way decent. Yes there is clipping and lots of distortion, but I've never heard anything that sounded as bad as what I came up with to reach the same loudness. What is their secret? L2 Ultramaximizer maybe, I don't have access to that one. Hm.. maybe I should convert to 32 bit, apply compression and dither back down to 16bit.

From the mouth of Billy Corgan...

Reply #64 –
A 0 to 1 peak to peak signal should be quieter than a -1 to 1 one. Also, a signal with 1-level samples largely spaced (low frequency) may be interpreted as a much quieter signal (like if it had been dithered).

From the mouth of Billy Corgan...

Reply #65 –
Quote
well, ya see, the problem is that playing back music that is clipped like that can and will blow speakers if you turn it up too loud, tweeters are the most susceptible due to the size of their voice coils etc being too small to be able to handle all the heat.  Sound quality definitely is a concern here with the bad mastering going around, but so too is equipment damage.

What about with radio broadcast, I made some FM radio rips, then I normalized the waves files and the wavGain values was between -16 to 19 dB, I think that this is insane, I don´t know if this only happens here in country or happens in most of the FM radio stations of the world, the more quiet WavGain value for a radio station here in my city was -9 db, the level of most of modern rock/pop albums. I think that too high levels can blow the speakers and damage it too. The trouble with radio stations is the same with the music labels, every broadcaster wants be the loudest, the same happens with the music labels.-
MPC: --quality 10 --xlevel (v. 1.15s) (archive/transcoding)
MP3:  LAME 3.96.1 --preset standard (daily listening/portable)

From the mouth of Billy Corgan...

Reply #66 –
Someone already said Merzbow, but really I think the entire power electronics/nose scene shouldn't count in this? I mean, I'm pretty sure my Aube, Merzbow, Whitehouse, etc. stuff would likely generate some rather nasty scores. One looping sample alone I was doing for a friend alone just measured in at -17.  MPPENC said there were 120 or so clipping errors encountered for a five second clip. Not exactly giving those sound engineers a fair chance to compete against 'boom'.

I am certainly glad I have the non-remastered version of the Stooges' Raw Power, though. If I want noise and clipping, I seek out those who specialize in it. I'll see if I can't track it down and post the original's replay scores.

EDIT: Curious. Anyone have the new Dead Kennedys 'remasters'? I bet they're compressed similarly since I heard they sounded terrible. How they could manage to make the DKs sound quality *worse*, given their limited resources to begin with is beyond me.

From the mouth of Billy Corgan...

Reply #68 –
Another goodie... 01-Chart Topping Radio Hit.mpc from the "Down With The Scene" album by Kid606.

Code: [Select]
   Level Adjustment   |   Peak Level   (Adjst)|  Filename
----------------------+-----------------------+-------------------------------
 +0.00 dB =>-17.63 dB |     0 => 49595 ( 6515)| 01-Chart Topping Radio Hit.mpc

From the mouth of Billy Corgan...

Reply #69 –
Quote
Haha, 0wned:
Venetian Snares - Songs About My Cats
Code: [Select]
    Gain   |  Peak  | Scale | New Peak | Track
 ---------------------------------------------
 -15.21 dB |  32767 |  0.17 |     5688 | 01-Chinaski.wav
  -6.49 dB |  32767 |  0.47 |    15522 | 02-Katzesorge part 1.wav
 -14.36 dB |  32767 |  0.19 |     6272 | 03-Nepetalactone.wav
 -13.83 dB |  32767 |  0.20 |     6667 | 04-Poor Kakarookee.wav
 -11.59 dB |  32767 |  0.26 |     8629 | 05-For Bertha Rand.wav
 -14.58 dB |  32767 |  0.19 |     6116 | 06-Breakfast Time for Baboons.wav
 -15.96 dB |  32767 |  0.16 |     5217 | 07-Fluff Master.wav
 -13.68 dB |  32767 |  0.21 |     6783 | 08-Bobo.wav
  -5.72 dB |  32767 |  0.52 |    16960 | 09-Katzesorge part 2.wav
 -12.92 dB |  32767 |  0.23 |     7403 | 10-Pouncelciot.wav
 -15.40 dB |  32767 |  0.17 |     5565 | 11-Kakenrooken Stivlobits.wav
  -5.18 dB |  32767 |  0.55 |    18048 | 12-Lioness.wav
 -13.82 dB |  32767 |  0.20 |     6675 | 13-Cleaning Each Other.wav
  -8.98 dB |  30185 |  0.36 |    10735 | 14-Look.wav

 Recommended Album Gain: -14.80 dB

The fact that a track could SERIOUSLY need limiting to +/-5.2k makes me laugh    Despite the 3 'quiet' numbers, could this take Californications crown?






Thanks n68   

Quote
Musepack, Liking Audio
I always liked how that font looks messed up 

Yes, but considering Venetian Snares music is the clipping really so bad??  Its meant to be pure pain anyway.
gentoo ~amd64 + layman | ncmpcpp/mpd | wavpack + vorbis + lame

From the mouth of Billy Corgan...

Reply #70 –
Quote
Yeah, but do their amps go to "11"?  Spinal Tap's do.  And Tap is England's loudest band.

Why not just make 10 louder?
gentoo ~amd64 + layman | ncmpcpp/mpd | wavpack + vorbis + lame

From the mouth of Billy Corgan...

Reply #71 –
When someone can't create music, they create crap. Loud crap is still crap. Am I glad, that I do not have to listen to it! Unfortunately that means I have not spent a penny on music lately. So I have found the perfect sound. No distortion at all. It is called silence.

From the mouth of Billy Corgan...

Reply #72 –
nah, there's still sound
 

From the mouth of Billy Corgan...

Reply #73 –
O.K. There is still noise and distortion, but no clipping!

From the mouth of Billy Corgan...

Reply #74 –
@ Sony 666 : That image defies all belief.  It really does.  Especially if you sit and think about it.