Loudness wars coming to an end

uglymusic

We've talked about the sonic compromises inherent in making recordings 'for radio'.

Here's the case some things may be changing. Or do you have to buy vinyl to get a decent dynamic range from chart material?

Alan

I would like to believe this is true. I am signed up to the DR movement, and have even measured and submitted some of my albums for their DR database, but I fear this story is simply too positive a spin.

The message is getting through, there has been some good publicity as mentioned in the article. Also bands such as Elbow are fully signed up which helps, you can see the 'turn it up' logo on the back of their record sleeves.

I think things won't change anytime soon, not in a significant way at least, but the awareness that people would like better mastering is spreading.

It's an oddity that audiophiles who spend so much on gear (ie - all of us) tend - on the whole - to often overlook the mastering of a recording as by far the greatest arbiter of sound quality. We can do something about this, but progress is slow - but our voices are being heard. Oftentimes, high quality or niche interest music is recorded very well indeed so we can content ourselves with this. However, this doesn't get the revolution rolling...

PAC

Jack White seems to have joined the vinyl revolution.  I like some of his reasoning even if it is less to do with recording quality!

http://www.bbc.co.uk/news/magazine-18545530?utm_source=twitterfeed&utm_medium=twitter

uglymusic

I think I might play Mr White's album, but digitally ;-)

PAC

I'll reserve judgement Dave until I get a chance to try out the new Hadcock GH242 with the 401....it may give my Gyro somethig to think about...or even the Droplet come to that!  (here's hoping anyway)

uglymusic

You can hear the limiting after the guitar at the start of the second track. I was in the kitchen for the rest of the album.

Docfoster

The whole studio sq issue fascinates me. I'm always surprised it doesn't generate more debate.

uglymusic

I think it's ironic that the medium that has the largest dynamic range is the one that gets messed most with on chart music.

That's not always the case, but the flipside is that niche music does more often get put out with a more realistic dynamic range.

PAC

I think part of that Dave is that it's easier to mess with digital files and allow more compression for louder playing without affecting what it's later played back on, whereas with vinyl, there's a limit to audio compression (ie dynamic) which can be achieved without the stylus jumping out of the grooves!  A good friend of mine who's been a studio engineer for many years (as well as the author of a recent biography on AC/DC bizarrely!) explained to me just how far digital recording has come. We sat in his listening room the other night listening to streamed music (lossless) and he was able to tell me from listening how each piece had been recorded and give an estimate of the compression with the reasons why it had been applied.  It wasn't all as clear cut as I had thought.  To test his golden ears, I played an LP without telling him when it was recorded and asked for his observations, and almost instantly, he got it nail on head by telling me it was recorded using an early Decca tree.  Not only did he get the date right within a one year period, he got the conductor too!  Clever people some of these studio engineers!

With respect to CD replay, many are compressed to a staggering 10dB range which is all but impossible to do with an LP depending on the music etc, which is why a similar recording on vinyl sounds better than the CD equivalent.  Its not that vinyl is superior (as it clearly isn't in technical terms) but that the CD mastering is terrible.  Get a good one though.....

Having said all that, I've invested twice on my vinyl front end now than on the Cdp as I enjoy tinkering.  May look for a London Decca Blue cart for the Hadcock as they are reputedly a very good match.  That's a story for another day though...

uglymusic

Thankfully, I have very little that has been compressed the hell out of!

This afternoon, I played @Jim my 24/88.2 AIFF of Charles Lloyd's Mirror, his album from last year. As I think he'll agree, it is absolutely stunning, putting so-called 'audiophile' recordings to shame. The Redbook will be almost as good, I'm sure, simply because it's an ECM release and they take a lot of care.

I also have a load of simple jazz recordings which haven't been messed with. I think the fall of the 'big' record industry is a great thing. Musician-owned and independent labels are putting out great digital recordings brilliant in their simplicity.

And maybe it's because so much of the music I listen to hasn't been f***** up by the record industry that I'm not using vinyl.

Docfoster

PAC - great anecdote. I love it when one gets to spend time in the presence of expertise.
Who knows, maybe I'll experience it here one day... ;-p
Also I hadn't realised that there were greater

Can I go back to basics? Why has post production compression become such a normality? I understand the reason for some compression for listening in cars, and in clubs. And with very dynamic instruments (eg drum kits) i've found compression is essential during mix down (or may be the drummers I know are crap...!)
But it's not clear to me why so much compression is used (10dB...! Blimey Charlie). In any event, radio stations and clubs probably compress everything again at their end anyway. I can only imagine that there's been a race to the bottom because it's what people have become accustomed to, and increasingly so. Folks have come to 'expect' their music to sound a particular way.

Docfoster

Do you think compression has just become the accepted norm? It's almost as if people 'expect' their music to sound that way.

When I say "people" obviously I mean tabloid-reading skytv-watching pseudo-crim' ibiza-holidaying dullards. Not us. Obviously.
Bastards.

Alan

You often get double doses too; excessive and undesirable dynamic compression on the recording, followed by still more added during broadcast by radio etc. 

The execs should realise that broadcasters can manipulate the signal any which way they want, and that their job is to provide the best quality (which means the least messing) possible. They simply dont need to compress so heavily.

PAC

I think that the reason for so much compression is twofold:  Those in the music industry that aren't technically minded saying "louder is better" and the marketeers who know from studies that kids visiting music shops to listen to the latest releases tend to prefer the "louder" ones. Add to that radio stations with limited bandwidth (way less than 192Khz) and not only do you get loud music, you get compression of data too.  End result?  Thumpathumpathumpathumpayellshoutthumpa.  Enough to make music lovers emigrate to their own island paradise!

It even permeates classical recordings which are better placed than pop to exploit CD's 90dB range, so why oh why foist a 30dB range recording on us?  It makes no sense.

Things are changing though, partly because to the backlash from the receiving end and partly due to young (and not so young) musicians insisting on less compression.  Compression is needed.  You couldn't record direct to vinyl for example (or to the master disc) without it with some music as loud passages would cause overlap in the grooves and you'd also end up with an LP which would be impossible to track, so for physical limitation reasons, vinyl is limited to between 50 and 60dB.  Thing is, most of my older classical and jazz/blues recordings seem to exploit this range very well.  Not so with a surprising percentage of my CD's.  Get a well recorded CD done properly to Red Book and you wonder why you bother with LP except for the reasons you may have 400LPs which you don't have on CD (a good enough reason).

Trouble is I think that there's a gulf in understanding between the recording and broadcast engineers and the record companies themselves.  The record companies who peddle pop (many of them anyway) are under the misunderstanding that louder on air means more air time for that track = greater sales, so push their engineers to compress the crap out of every album.  What the recording engineers fail to appreciate (and the broadcast engineers) is that when an already compressed digital file is broadcast with more compression (broadcast compression) with the compressors in a muddle where one will want to bring down the level and another bring it back up...and to a level it wasn't originally recorded at, so not only is audible compression heard, additional distortion is added into the mix.   

Chuck in MP3 streamed music with compression and other lossy formats and we're so far removed from fidelity of any sort that it's just "muzak" bordering on noise.

I keep falling back to LP at home partly because I now have a 500 or 600 strong collection (never counted them all!) but because it's rare to get this level of compression because its a physical impossibility.  That reason is one of the major factors why there has been such a renaissance in vinyl in the last decade.  Its an inditement on the music industry that give them a medium capable of superb fidelity but also capable of major flexibility in how they play with it, and the money men and ignoramuses float to the surface where greed and stupidity actually controls the end result. A bit like politicians really 

uglymusic

The s*** floats to the top, I believe.

Docfoster

Limiting is one option isn't it?
Ie simply rolling off the highest peaks to make the overall recording useable, as opposed to crunching the whole thing up into a handful of dBs as PAC describes.

PAC

Limiting at what frequencies though?  One could view compression as limiting in the opposite end by boosting low notes.  Either way, the dynamic range is compromised.  Whilst there are good arguments for some compression, one of the reasons that rolling off higher amplitude signals isn;t popular with the recording companies or the radio stations is that it doesn't address the loudness issues, just the peak loudness issue.  In effect your amplifier/speaker circuit already acts as a dynamic limiter as the maximum SPLs produced are simply a function of the output power and speaker sensitivity, and we all know roughly speaking that 100dB is bloody loud in an enclosed space whilst 40dB is average background noise during the day with the volume turned down, so one could argue that except for the odd quick transient peak (a loud drum thwack for example) that 60dB range is plenty enough for the average listening room (which happens to coincide with maximum LP dynamic range).  However, applied to broadcasts, it doesn't help sales or airplay to limit the top, they want compression and everything to be LOUD.

Studio engineers when mastering , mixing and recording for CD to Red book should aim for 90dB range depending on the material being recorded and by definition you'll never get (or want) some of the amplitudes experienced in the studio or at a live event in your living room, so a degree of rolling off is already practised in recording music.  Peaks are usually defined in relation to audible threshold, and the value of that threshold determines theoretical peak volume for the recording.  The actual value when replayed depends on your gear and how much wattage you feed the speakers and only when a certain wattage is attained to you realise the full dynamic range in your listening room. For most purposes 60dB range is more than enough unless hearing damage is the aim!

Docfoster

Agreed.
And thanks for fleshing out the figures - meat of the sandwich and all that...
I drew the distinction between compression and limiting (and agreed - it's a spectrum) to frame them in terms of their different purposes: compression for loudness and; limiting for making recordings usable (eg in the case of vinyl to avoid the problems that you describe).
I guess I'm just restating what's been said above. I'm good at the bleeding obvious mind.

RobHolt

I don't automatically view dynamic compression as bad and I think it gets an undeservedly bad press usually.

I do think it gets used inappropriately at times but we must always consider the type of music, the intended audience and the conditions where it is likely to be played.

Some dynamic compression is probably required for any domestic replay. For example, a live band playing in your living room just as they would in a normal live venue would make quite uncomfortable listening. It might be accurate in the strictest sense of the word, but are you actually enjoying the result?

Squash it down a little and suddenly it becomes more palatable. The trick is to do this skilfully so as not to kill the music, and that's where I think things have gone adrift in recent years.

More generally, having music sound enjoyable on portable systems, or in noisy environments such as cars does require considerable compression. Of course it would be great to have the option to apply this via the hardware, leaving the original intact, that's surely the ideal but sadly not the reality.

Alan

I don't automatically view dynamic compression as bad and I think it gets an undeservedly bad press usually.

I do think it gets used inappropriately at times but we must always consider the type of music, the intended audience and the conditions where it is likely to be played.

Some dynamic compression is probably required for any domestic replay. For example, a live band playing in your living room just as they would in a normal live venue would make quite uncomfortable listening. It might be accurate in the strictest sense of the word, but are you actually enjoying the result?

Squash it down a little and suddenly it becomes more palatable. The trick is to do this skilfully so as not to kill the music, and that's where I think things have gone adrift in recent years.

More generally, having music sound enjoyable on portable systems, or in noisy environments such as cars does require considerable compression. Of course it would be great to have the option to apply this via the hardware, leaving the original intact, that's surely the ideal but sadly not the reality.

This is a nice point, well made too (sorry I missed it for so long).

An appropriate amount of dynamic compression is absolutely necessary with most recordings, it gives an appropriate sense of dynamic 'perspective' and scale. But when it is used to equalise the finished piece to an artificially loud (and often distorted) level then it gets in the way of believability, rather than helping the listener to suspend disbelief.

Some mastering engineers do this beautifully, often the best examples are slightly older recordings. For instance, listen to George Massenburgs work with Lyle Lovett on 'Pontiac', and 'Joshua Judges Ruth'. He now runs his own firm making parametric equalizers and dynamic range controllers, as well as teaching at various universities as a professor of recording arts and sciences. There is a list of people he worked with here, all names that care about their SQ.

PAC

It is a good point and very well made.  Dynamic compression is essential in certain circumstances.  Consider LP replay.  LPs are limited to about 60dB headroom, the grooves would be too wide and cut through to the adjacent groove tracks if amplitude was increased further so they require some degree of dynamic control, but go too far ironically and with headroom reduced to below 20dB, a similar thing happens as "loudness" increases so mastering for an LP is a skillful business.  Looking at CD, headroom is close to 90dB, and if a system floor is say 45dBA, the resultant 135dBA on transient peaks would be uncomfortably loud plus most domestic set ups would struggle to reproduce it, so dynamic compression to a 60dB range is more the norm for a good classical CD recording (for example).  Some music doesn't have a particularly great dynamic range so it becomes less important.  The problem arises when (mostly) digital recordings are compressed to 20 or 30 dB range, killing the ambience of the event (studio or live) and making everything too compressed and loud and fatiguing...all IMHO of course.