Losing (beating) the Loudness War!

[Rubycon]

Interesting comparison here between two releases:

http://www.youtube.com/watch?v=DRyIACDCc1I&

This practice must be stopped! They are ruining music. (what's left of it)

I must admit I listen to quite the variety of genre BUT if it's compressed to the wall I tend to hit the skip and forget button real fast.
[FONT=&quot]
[/FONT]

 

[coxmaster]

Is that music?

 

[Cogman]

Ruby, you failed and tried to link to a file on your local computer

I agree though, the loudness war is terrible, why anyone thought it was a good idea is beyond me (As well, loud commercials are horrible)

 

[Tiamat]

Interesting comparison here between two releases:

http://www.youtube.com/watch?v=DRyIACDCc1I&

This practice must be stopped! They are ruining music. (what's left of it)

I must admit I listen to quite the variety of genre BUT if it's compressed to the wall I tend to hit the skip and forget button real fast.
[FONT=&quot]
[/FONT]

Agreed. Except, I hit the delete key real fast. Music compressed to kill dynamics and the timbre of the instruments is just a waste.

 

[Rubycon]

Ruby, you failed and tried to link to a file on your local computer

I agree though, the loudness war is terrible, why anyone thought it was a good idea is beyond me (As well, loud commercials are horrible)

I don't know what that was in the post before it was removed. The only thing that was in the clipboard was a youtube link. Who knows?

As for commercials they typically run the program material 6dB lower than the max and commercials are RMS normalized to the max! Supposedly this will be changing.

Speech compression is a good idea but using it in music to make things louder is horrible. There is no life in drums. A CD is capable of 96dB dynamic range yet most modern recordings have less than 10dB dynamic range!

 

[Cogman]

I don't know what that was in the post before it was removed. The only thing that was in the clipboard was a youtube link. Who knows?

As for commercials they typically run the program material 6dB lower than the max and commercials are RMS normalized to the max! Supposedly this will be changing.

Speech compression is a good idea but using it in music to make things louder is horrible. There is no life in drums. A CD is capable of 96dB dynamic range yet most modern recordings have less than 10dB dynamic range!

Well, with voice (telecommunications) yeah, raising the volume doesn't distort much, but that isn't really too common of a practice (that I know of). Rather, the practice is to limit the sampling frequency. Human voices operate in a much more restricted frequency then music can reach, hence most telecommunications have a pretty limited frequency (saving bandwidth and allowing for cheaper materials to make a phone, at least, that was the thinking around the 70's)

I guess if volume is too low, then you could lose some detail that way as well, However, if the sound is lost from being too quite, it is lost regardless, you can't create something from nothing.

 

[Raduque]

Mashed up, loud, distorted garbage with no dynamic range is NOT "music". It's just noise. Maybe that's what metallica was going for, but it's still not listenable.

Dynamic compression is a very big reason while I don't listen to most modern music. The crap is even applied to country. Country!

I'm so glad this practice only started fairly recently. Could you imagine what Dire Straits "Private Investigations" would sound like today? It would probably make me cry.

 

[Rubycon]

Well, with voice (telecommunications) yeah, raising the volume doesn't distort much, but that isn't really too common of a practice (that I know of). Rather, the practice is to limit the sampling frequency. Human voices operate in a much more restricted frequency then music can reach, hence most telecommunications have a pretty limited frequency (saving bandwidth and allowing for cheaper materials to make a phone, at least, that was the thinking around the 70's)

I guess if volume is too low, then you could lose some detail that way as well, However, if the sound is lost from being too quite, it is lost regardless, you can't create something from nothing.

I HATE the voice "quality?" of modern cell phones. It's barely acceptable with a good signal. When the signal is low (-100dBM) it's so garbled and messed up and dropping out, etc. At least on old analog cell systems you heard the noise floor creeping up but voices were still clear and understandable.

Compression is often used to keep the voice loud(er) but at the expense of picking up lots of background noise. Sophisticated multihomed NR algorithms take care of this and the decline in quality is not noticeable under standard digital wireless telephony.

Telephone bandwidth is about 3kHz so one can use 6kHz sampling rate at 8 or 12 bits and have adequate quality. Of course with very lossy VBR it goes downhill from there.

 

[EGGO]

You're about a decade late, I'd say.

I think our only saving grace when it comes to cds with music we'll like is live recordings, aside from genres like Jazz, Classical and whatnot that almost never does it.

 

[Rubycon]

You're about a decade late, I'd say.

I think our only saving grace when it comes to cds with music we'll like is live recordings, aside from genres like Jazz, Classical and whatnot that almost never does it.

Actually live performances are seeing the effect of compression too!

It's much easier to make a performance louder with the same kit just by compressing the signal.

The soundtrack to Pirates Of The Caribbean was VERY compressed. Well now that we no longer have AOL CDs we can skip these! How far can YOU skip a CD across a pond?

 

[The Boston Dangler]

rubycon, tiamat, et al:

do you have an opinion on replaygain? i'm using it to normalize levels, but too often music takes on the sound of the metallica cd in the op. with replaygain turned off in foobar, it immediately sounds 100% better.

 

[Cogman]

I HATE the voice "quality?" of modern cell phones. It's barely acceptable with a good signal. When the signal is low (-100dBM) it's so garbled and messed up and dropping out, etc. At least on old analog cell systems you heard the noise floor creeping up but voices were still clear and understandable.

Compression is often used to keep the voice loud(er) but at the expense of picking up lots of background noise. Sophisticated multihomed NR algorithms take care of this and the decline in quality is not noticeable under standard digital wireless telephony.

Telephone bandwidth is about 3kHz so one can use 6kHz sampling rate at 8 or 12 bits and have adequate quality. Of course with very lossy VBR it goes downhill from there.

One thing I've always wondered is why they don't use better compression algorithms. I swear they are using something around MP2 quality, where as the AAC standard has been around for several years now. It is somewhat strange (on a somewhate unrelated note) that AAC has been around for so long and yet there are so few GOOD software solutions for using it. It has been relegated to the realm of "slightly better then MP3" when it could be much better IMO.

 

[Rubycon]

rubycon, tiamat, et al:

do you have an opinion on replaygain? i'm using it to normalize levels, but too often music takes on the sound of the metallica cd in the op. with replaygain turned off in foobar, it immediately sounds 100% better.

No definitely leave it off!

If you need a background music application and your program varies wildly you should use a processor to peak normalize and deal with differences in volume from mastering. RPG tends to over process and takes away the "life" of the music.

One thing I've always wondered is why they don't use better compression algorithms. I swear they are using something around MP2 quality, where as the AAC standard has been around for several years now. It is somewhat strange (on a somewhate unrelated note) that AAC has been around for so long and yet there are so few GOOD software solutions for using it. It has been relegated to the realm of "slightly better then MP3" when it could be much better IMO.

Perhaps someone that works for the telecoms in this specific area can chime in. I'd be willing to bet it's left as is because it's cost effective.

 

[BurnItDwn]

Lack of dynamic range is indeed a really sore thing.... Thankfully, there are a lot of bands out there who prefer for their music to sound more musical.

 

[GodlessAstronomer]

I prefer the CD version, distortion has been a part of rock music since Geezer Butler first cranked his bass amp to 11 to experiment with clipped audio. I like the "mushy" distorted sound, as long as the vocal track is intact.

 

[CKent]

I had the GH3 version within a week of release. It's actually not a bad CD, when you can listen to it without ear fatigue after 1-2 songs.

 

[Tiamat]

rubycon, tiamat, et al:

do you have an opinion on replaygain? i'm using it to normalize levels, but too often music takes on the sound of the metallica cd in the op. with replaygain turned off in foobar, it immediately sounds 100% better.

Leave it off. Every software-based filters I've tried made the quality worse. I don't know of an inexpensive solution (i.e. free-of-charge) to normalize levels without degrading the quality.

 

[vi edit]

What's sad is that I clicked the link already knowing that it would be about Death Magnetic. That's like the gold standard of how *NOT* to master an album. Even on my crappy car speakers the drum compression is insufferable. I don't even want to know the level of misery it suffers on a high end system that can really point out the horrors of dynamic compression.

 

[jersiq]

Perhaps someone that works for the telecoms in this specific area can chime in. I'd be willing to bet it's left as is because it's cost effective.

I have to totally ignore TDMA, as I never learnt it

But in CDMA you have to remember that the goal is to have the lowest possible transmit power at all times. In fact, your handset is actually at wrong power levels all the time and the Base station has to constantly correct it, but that's to prevent it from running away.

So CDMA has had these Codecs chronologically:
8k, 13k, EVRC some later variant of EVRC that no one used and finally EVRC-B

8k and 13k did not do very well with high Bit Error Rates, which when given the constraint of lowest possible power can happen frequently. And of course, there is the bandwidth argument.

Most people don't recognize a difference between the two (13k and EVRC) and half of your conversation is listening anyways.

So, not knowing how well other Codecs do at high BER, I can't comment about that aspect. However, there has been a strong move to EVRC-B as it is SIP-T compliant. Also, there are a lot of scenarios now where the Codec in the MTSO is being bypassed if you are making a call to another CDMA phone on the same network. Previously, your voice was compressed, sent to the MTSO, decompressed, compressed again and sent to the called party. In some (not all) call scenarios, this middle man is going away.

 

[CrimsonWolf]

Rick Rubin can eat shit for how Death Magnetic got mixed.

It sounds fucking terrible whenever I play it loud on half-decent speakers. The tragedy is that the songs are otherwise great.

 

[mmntech]

What loudness didn't kill, Autotune did.

I like classic rock and this is one of the reasons why I don't like "remastered" content.

 

[40sTheme]

Compressors are very useful; especially for guitarists. Except there's no way I would have the attack that early and the sensitivity that high. It would screw it up; although I don't use a comp anymore.

 

[Rubycon]

I prefer the CD version, distortion has been a part of rock music since Geezer Butler first cranked his bass amp to 11 to experiment with clipped audio. I like the "mushy" distorted sound, as long as the vocal track is intact.

No, fuzz is different and only used as distortion aka pedal. Taking the entire gig and compressing it to the wall is crucifying to say the least.

What loudness didn't kill, Autotune did.

I like classic rock and this is one of the reasons why I don't like "remastered" content.

Auto tune needs to die.

 

[SludgeFactory]

Brickwall limiting is the scourge of modern music. I honestly believe it's done *a lot* to give music a disposable, this-really-isn't-worth-paying-for quality, something that seems so prevalent with many people over the last 10-12 years. Most of them can't tell you what's different, and probably don't automatically cringe within seconds when a compressed-to-hell recording is played, but piss poor mastering is something that subconsciously hits everybody over the head with a 2 x 4.

You simply cannot listen to this crap for more than a set amount of time in one sitting. For me it's about 20-30 minutes. It's best suited to poor listening environments, i.e. your car, your shit-quality Ipod + earbuds while you're doing four other things at once, the backing track to a 30 second commercial, etc. Is it any wonder that hi-fi is dead?

Well mastered rock and roll can be cranked to 10, to higher volume than the "loud" overcompressed garbage, and be enjoyed for far longer without your brain waving the white flag and surrendering to fatigue.

The trend in mastering has been screwed up for a long time, and the rise of the Pro Tools "studio" in the wrong hands (i.e. just about every artist/producer who thinks he can self-engineer the album) is going to result in a lot of recordings that are fubared before they ever hit the mastering console. People have claimed that was the case with the Metallica and Rush albums, two of the most infamously awful sounding examples to come out of this era. Supposedly they were atrociously recorded, and there was very little a mastering engineer could have done to clean up the mess. Audio engineering is in a sad state of affairs right now.

 

[mmntech]

Auto tune needs to die.

Here's proof that it can make anyone into a pop star. "Stalker Patti" is a regular guest on the Opie & Anthony show. She is supposedly a virgin in her 50s. Don't know if that's true but she is in her late 50s and she does have a horrible long island accent. She sings Tik Tok by Kesha, edited with Auto Tune in two hours,
http://www.youtube.com/watch?v=XQbpF0_5-c8

This is how she normally talks, she's in black
http://www.youtube.com/watch?v=EAv7Tp7RP-E&feature=related