Discussion in 'Mastering' started by audiokid, Oct 22, 2012.
Ian Shepherd discussing the 2012 Green Day Remaster .
I started listening to this on laptop speakers, and of course heard very little difference. There is a lesson in that. So many listeners today are using low fidelity playback in rather noisy environments. The compressed version is arguably more useful/provides a better listening experience/is "better sounding" in that situation. With a switch to consumer-level headphones the differences are easier to hear, but they aren't huge. I certainly prefer the uncompressed version, but I understand the consumer appeal of the compressed version.
Another takeaway is that the mastering for this song is very well done. It may be compressed more than I like, but there is still some life and definition there. Yes, this type of music makes it easy to tuck a little clipping into the mix without anyone noticing, but whoever did this is better at it than most. Most squashed to hell records song a whole lot worse that this. I wonder if this "victory" in the loudness wars won't turn into a defeat? Are artists going to listen to this and say, "I can get a 5dB louder record and sound THAT good? I'm in!"
The one that gets me most excited is being penalized if your master goes over the mentioned threshold, and how the brick walled master drops in level compared to the one with more transients. If that was enforced, what a victory. But I wonder if someone would find a way to fool the censor. I doubt it would be an analog function that takes care of that.
But wouldn't this be wonderful.
I don't know about you but I have spent hours trying to get something just slightly louder, knowing damn well I'm killing it to get there. And what a joke putting all the energy into the volume over sonic mastering. Like you say, a bit of rasp sure hides some of the ugly.
Something that I've noticed (that could turn into a really interesting topic) is how much louder electronic music can get with less distortion in comparison to a track with even one "real" acoustic instrument in it. The more first generation acoustic information in a song, the harder I find it is to compete without killing it. There isn't the transients in electronic music like their is in acoustic sounds. And the more times you sample it, including ADDA, it starts looking like a brick, or sounding more digital per-say. In a resent topic over at Track Talk, I proved in a Null test to myself that I could make my analog mix sound digital in a few conversions. But you can't do it the other way around.
Spend enough years as I have with samples and you soon realize why electronic music is an attraction. The wall of sound at volume setting 2 sells. And like you say, most people don't hear the transients so the next question is, are we sonic martyr's or business men. How do the purest compete when this is what sells.
Two decades of samples being perfected or recycled, no wonder music is all washed together today. But the new generation doesn't even know this has happened. Its loud but there isn't anything real left. And so, the music keeps evolving, out with the old, in with the new. Its very interesting to me, including the psychology of sound.
I would love to hear what others think about " electronic vs acoustic mastering loudness levels?
I am a not a Mastering engineer or even close to being a professional recording engineer. I sure hear a difference on my shure headphones. The difference is funny because you still think the clipped version has more impact on the total sound of the track. I think we have been brainwashed by the loudness wars. It's as if your getting a quick one w/ little regard to your emotions or your ears. In the long run you will feel used and you will suffer a sonic head ache. That's how I feel about electronic music too. It's not fulfilling to me in the long run but at first is mesmerizing! I am not saying it is bad, but just that after some time I need the old school approach again to get my bearings back again.
I have never mastered any track that was more than -11db RMS ever and never will make anything ever louder than that... That was my rule a while back and has been ever since. If I see a brick wall forming then I feel I am only being lazy and not committed to the listener. If the listener does not agree well then maybe they are deaf by now? Even at -11db I see a wall forming and that wall is much better than a wall at -6db.
I usually wind up with masters around -15 to -11db and find that to be a scale of normalcy in a reality based production sense. I am afraid to join the loudness wars and actually used the topic in a public speeking class in college. You should have seen the people's faces listening to my public speaking exercise!! They had no clue at all what I was trying to explain but it was a great time spent analyzing the loudness wars. I think I made at least one person in that room go research the loudness wars and that was enough for me. That one person was involved with film and that is probably why it was even an interesting speech to them.
I am really happy to see something in the industry that is taking into account what is the rule of thumb to mastering for loudness. Guidelines should be set to keep the dynamics of music still alive. Electronic musicians would agree too but it will only make sense to them in about another ten years.
electronic vs acoustic mastering loudness levels
Regarding my comment on electronic vs acoustic mastering loudness levels, have you ever compared your master's (example: metal, jazz) to a remix of club music?
Sent from my iPhone
i can hear the difference on my laptop speakers and it's not small.
i prefer the sound of the HD less compressed/less clipped version although the results are nothing new for me. i have always thought that smashing masters to get louder ruined them sonically.... on the other hand anything Billy Joe Armstrong writes and sings is tops to me and i would listen to it no matter how smashed it was. rock on Green Day and thanks to Ian Shepard for doing this and thanks to Chris for posting it for our enjoyment. this made my day ... so don't shoot me just yet.
The difference between small and not small is all a matter of perspective...or haven't you ever had a woman tell you that.duh
wadda mean there Bob?
seriously, it wouldn't take a trained ear to hear it. i think the difference is very plain to even a novice. i guess my laptop speakers must be good ones! :biggrin:
NOW you can just shoot me!!!
mmmmmmph! arrrrrgh! suicide ahhhhhhhhzh*t!
re Remy's Neve on pruemix.net: http://recording.org/pro-audio-gear/53526-remys-old-neve.html
I found this scrolling through and thought this was appropriate to this thread:
And to add, my life changed after I got the Dangerous Monitor ST. You don't know what you're missing until you get the Monitor ST or second best, the SPL 2381.
Yes I did. It is not always the case (lots of variables) but yes, in my experience, electronic music could often get louder (apparent loudness) compared to "real instruments" music without affecting the original sound, shapes and vibes.
"Even Green Day sound better with dynamics" Nice.
I agree totally with the above thinking, and led me to believe that new producers and engineers have to be well educated about real sound and psycho- acoustics. It is a shame how the industry is becoming more un-aware of basic concepts about sound engineering.
Please do not get me wrong as I know there is a lot of good professionals around, but there is a lot also, of amateurs doing pseudo electronic music with mono thematic ideas with in the end are simple seven minute loops with little music in them really. With no feel what so ever.
About the acoustic vs electronic question, I really think it is hard to compete with really fast envelope sounding kicks, and there is a lot of already mastered and processed material that has to go to a master bus again to be compressed again during mastering and so on. Acoustic sounds have to be treated as such. Good miking techniques and all that.
Blumlein array, fig 8, MS recording... sadly this is all becoming sort of unknown things.
I have done a lot of different styles of music as a producer and artist, and sometimes I have been really disappointed with the mastering sound itself, even form well respected mastering houses. It appears to me that the only thing I received in return for the money was a lot of boost on hi mids and a lot of soft limiting.
I personally love analog vintage like equalizing, like E.A.R stuff, Neve, etc (who does not?) and I think plugins are really away from that "sound", but sadly it is becoming a standard. Since they are easily accessible, there is an existing grow of myth and black mastering and mixing techniques that is spoiling the art of sound.
Best regards to all
While watching this CLA vid of a 48 track master session "Holiday",one can clearly see he definately like to hit the compression hard enough to produce some 2nd order harmonics from the gear.Also it looks like some of the audio is being run through 1/4" tape at 7 ips on an old pioneer reel to reel.This guy is definately about driving the gear to the brink to bring out distortion.
I think this may be part of the reason why "American Idiot" gets picked on,as there is deliberate audible distortion on alot of the elements in the mix that only get more emphasis during overdriving brick wall limiters during mastering.
While it's clear to see the clipped the peaks in the CD master in Wave Lab,that's not at all unusual these days
He's skeptical about there being an improvement moving from 16 to 24 bit?! For a stereo mix??!!
For the A/B test he'll downsample a Green Day (you wouldn't expect pop-punk to sound loud & trashy?) 24 bit file (remastered at higher resolution exactly how?) to 16 bit... then resample the 16 bit file by attenuating it by a non-integer dB/bit value? Then upload the results to youtube for one to mull over...
16 bit is indistinguishable from 24 bit?
If you mix at -24dBtp with no compression on the mix bus at all... oh you'll hear the difference alright.
In 16 bit maintaining a strong level is especially important considering that 1/2 of the resolution of LPCM is in the loudest 6dBs below clip, you get quite a bit more resolution if you use the loudest possible levels. As you get quieter, the amount of quantisation becomes increasingly harsh and distorted.
You get nasty obviously audible artefacts creeping in at around 11 to 8 bit, and becoming very pronounced around 6 or 7 bit, so if you're frequency masking those artefacts out with loud broadband sounds such as heavily distorted guitars, you may not be able to audibly pinpoint these things. Those artefacts do sound rather like the distortion you get from a transistory marshall amp, the "valvstate" series. However, just because the meter level is reading a big level, doesn't mean that quiet elements of the mix aren't being pixelled out. 16 bit running at -48dBtp the bitrate is only 8 bit, so any sounds running close to that threshold in the mix will have those artefacts, and be audibly horrible. Complex sounds such as reverb tails should be present in a mix at those levels *relative* to the loudest possible sound, so even peaking at 0dB resolving that level of detail is off the cards. Running 24dBtp headroom and having clean subtle reverb at -40dBfs is just not an option at 16 bit.
Single decimal figure bitrate – 9 bits or less – have a weird distorting/noise-gate effect. Hip-hop producers from the 80s and early 90s used the low bitrate (12bit) of the EMU SP12 / SP1200 sampling drum machines for the "crunch", and the foot-room resolution issues in that machine were such that it required each voice output to have an analog voltage-controlled envelope filter to filter out the mashing hiss on note tails (also used creatively to put "fatness" into the beat). You can use 14 bit or 12 bit as a processing effect on snare drums, it takes out all the reverb and makes it sound like there's more snare chain and crunch in the sound, and rig up a synth-filter to do the job of taking out the digital ripping noise in the note-tails. Artefacts can be employed.
Can't hear the difference between digidesign 882s and 888s?! 16 bit is *just fine?!*.
Here's some artefacts that I just generated by nulling a jazz record (this isn't copyright infringing... the music is *gone* if you do this, bet y'all $50 that you can't guess what track this once was) against a low-bitrate version at the same level (5 bit – right in the sweet spot of nasty), so that only the aberrations are left over. X Desk that thing that sounds like a whiplash is actually a solo piano and later in the sample a bit of piano and double bass.
The advantage of 24 bit is that these harsh, crunchy aftefacts are not audible on the resynthesis and playback systems of today, since they're masked out by a typically 80dB signal to noise of the analog circuitry and speakers and room the converter is liable to be running into. 24 bit converters have "foot room" beyond the typical dynamic range, keeping the artefacts from being heard. There is much less audible tizzy crunch. Also, the number of audio levels, and hence the amount of detail that the mix can hold is a substantially increased – specifically every bit of 16 bit has 256 steps in 24 bit. That's 256x the detail. Even with a pristene 110dBfs of dynamic range, that's 18 bits. At the very floor of detail that can be resolved, the resolution at the noise floor is 6 bit. You don't need to dither 24 bit, since the noise in the analog stages will most likely take care of that for you.
Try getting a decimator on a linear 24 bit mix with no compression, and A/B against 24 bit setting the decimator to 20, 16, 12 bit, and then 8 bit, and see what happens. A slow, dissonant piano or guitar piece with a lot of loud and quiet bits will be particularly revealing, but it should also be quite obvious on drums – or anything with dynamic range. In both the case of amplitude resolution and sample frequency, as resolution increases, the improvements are subject to the law of diminishing returns.
However, I'm sure that if you could give me some 24 bit PCM converters that had 256x the sample rate – 11.2MHz – they'd sound better than 44.1.
96k is about as much better than 44.1 as 17 bit converters would be better than 16 bit converters. It's still relatively close to the threshold of audible artefacts (in the case of sample rate having a Nyquist limit in excess of the highest frequency people can hear).
Ideally one wants to be many orders of magnitude over that threshold, so that there are say hundreds of datapoints per cycle of a 20kHz wave, allowing one, for example to represent multiple non-periodic high frequency waveforms over the top of each other, with different phase differences through each channel of playback. That sort of detail is important for sound localisation and separation of one sound source from another in the mix.
I'm of the opinion that there is an argument for pushing levels on 16 bit or CD because of the resolution issues and what happens to the quietest parts of the mix. 16 bit is not a format with a lot of foot-room. Use dither? That hissing least significant bit arguably just makes matters worse, to really override those issues and get a consistent hissing background you have to use so much dither that it makes compact cassette appear to have a really desirable S/N.
Obviously there's no good in pushing stuff into the sausage shape, and obliterating all peaks, but 16 bit should not really be considered a high fidelity format, it hasn't got the dynamic range. There will be a compromise at one end or the other, quiet or loud, it's a balancing act.
But y'know. CD is an 80s format. In digital-years that's worse than dog years. CD is related to the SNES and ZX Spectrum. It's very surprising that it's still around. Just goes to show you who runs the electronics / media companies, and the trickle-down of technology is heavily influenced by the drive to retain control. Customer needs and product quality are *secondary* to maintaining corporate territory. Hence a lot of stuff happens that makes no sense for the customer or the market http://en.wikipedia.org/wiki/Rent-seeking http://en.wikipedia.org/wiki/Vendor_lockin http://en.wikipedia.org/wiki/Protectionism etc.
Of course, Yetnikoff is singing a different tune these days http://www.youtube.com/watch?v=bAzuvFxrkP0
Ahem and whatever happened to http://en.wikipedia.org/wiki/Direct_Stream_Digital ...?
In my opinion, it's all in the context of the audio - the style, the intended usage, etc.
For those balls-to-the-wall club/house tracks, where the main goal is simply to jam a dance floor, there's no real need to protect any dynamic range, because there isn't really much dynamic range there to protect to begin with.
On the other hand, take a track by a singer/songwriter artist - someone like James Taylor for example - and it's all about the dynamics.
I've just finished a project where I agreed to assist a friend to "master" a project from a band he is in; it's a bluegrass thing (actually more of a "newgrass" style), and the reason I put the word master in quotation marks in the last sentence is because:
1.) I'm not a true mastering engineer, (and I told him up front that I wasn't, and that that the best thing they could do would be to have a real M.E. do the work, but they didn't wanna pay for it, sooo...) and
2.) I didn't have access to the gear that most authentic mastering engineers and mastering labs have.
In the end, against my friend's strong suggestion - and mine as well - the rest of the band opted for a hyper-limited, super squashed track... because to them, their only criteria was that it be loud and that they be able to put it next to a professional release and have it equal in volume.
Sonic integrity - like EQ and dynamics - meant absolutely nothing to them. They couldn't have cared less how it sounded in terms of tonal attraction or dynamic range that is inherent with acoustic instruments.
All they cared about was the volume. If getting the volume to their acceptable level meant sacrificing tone and dynamics, so be it.
In the end, you have to do what the paying client wants (and they didn't pay much, by the way LOL).
Although, I did ask that they not use my real name on the album liner notes, and I had absolutely no hesitancy in telling them why.
In my humble opinion, of course.
People should be pointing the finger at poor quality systems with limited dynamic range and headroom, which requires people to turn stuff up.
Whether it's poor quality car sound systems, CD's 50dBfs dynamic range, weak headphones amplifiers in various devices, rubbish boomboxes, rubbish computer sub & satellite speakers, rubbish plastic hi-fi systems with 12W 4 inch woofers etc, the issue is quality and quantity of sound, both are very limited for the typical listener. The solution of today is to do work on the master to override these crippling limitations, which isn't the ideal place to do the work. Half the time there's some extra processing being done by the playback device, "Mega T bass" etc. The restrictions are largely based upon decades of manufacturing cheapness.
Were everybody to own a decent quality playback system, they'd not find such an improvement in the overcooked master. As it is, 95% of the people receiving mainstream material are feeding low-res digital files into really rubbish playback systems which don't go loud enough (and have no bass).
It is actually a result of the chronic lowering of costs and increasing of profits in the electronics manufacturing industry. The people who own even reasonable quality playback systems are a minority.
Dance music is a different matter. PAs are really high quality these days. If you overcook a master that's going out to be played through d&B, Funktion One, Meyersound or EAW, it won't sound better for having a lot of compression, it'll just sound muddy, hurt the audience's ears and end up getting turned down.
Since owning my hybrid system and hearing such beautiful sounding mixes now, my worst nightmare (and where all my time is being wasted) is when I have to upload my mixes in a pool of other crap tracks that have been smashed to zero transients.
If the majority in the pool are all 8 db louder than my pristine track, will the audience take the time to turn up my song? If they do, it will clearly sound better but if they don't, it sounds duller at the low volume my sound system is set to for "the majority".
Regarding a PA system.. If a sound system is set-up and eq for the "majority" and you toss in a song full of dynamics and full bandwidth, will blow the system per-say or sound out of place, whacked etc. Am I wrong I'm wrong?
There needs to be a standard because we are being forced to accept the lowest common denominator. And sound systems and general population are evolving all over the world to accommodate smashed and compressed music. The radio in my town sounds like ass. Everything reminds me of the days of AM radio all over again. Lets hope for someone to start "FM" again. Remember that?
To shift into a related discussion about limiting now and to talk about what I am learning.
Has anyone tried Pro L from FabFilter? The limiter is awesome, best I've used. But, its forces me to think like the "majority" once again. It appears to be designed for smashing. No analog limiting system is as fast as this is.
So when I actually take advantage of its capabilities, it does some weird stuff to the transients of lush analog full bandwidth masters. Too much information there.
But if I think ITB and design my sound with the intention to smash my music, plan ahead during my mixing stage with this limiter on ( example: DAW>DA>analog> AD to second computer which is the Mastering DAW> Pro L Limiter on the 2-bus> studio monitors) , smashing my music ends up sounding better "louder". You are able to hear everything they way it will sound smashed online.
So what I'm saying is, I have been able to make my masters louder ( better sounding) using that plug-in active early in the mixing stage while monitoring the master at the end of the chain. Its terribly painful doing this. I hate it. I hate how we are being forced to think "smart stupid" . Its a trend not only in music, but with everything around the world now.
At the end of the day,
All the lush gear I have invested in all gets turned into crap online digital music. I'm mean, who is playing even CD's anymore, and that was bad enough! How do we sell music to the world?
Playing the stupid game and learning to be one step ahead is sort of fun but its wearing off.
The greatest advantage having this all now becomes personal pleasure. Its like driving a Ferrari on a highway set at one speed with no hills or corners though.
FWIW, the more analog you go, the further away you place yourselves between the end result. Whichis why hybrid is choice for hi Fidelity sound.
If music doesn't change for the better in the next few years, however, I may be selling everything and buying a bigger boat.
The point is that the gramophone behemoths since the advent of digital technology have been a sinking ship. Their disagreement with digital was lossless copying, and that's when they logged out of the future. The idea that profitability or value in the record business is related to profit margins from a manufacturing perspective is the fallacy which is killing them. The record business should not be confused with the music business.
As we move to pure media – records that don't need pressing – and the encoding system of the playback device is running *software* stored as binary data, it will be possible to think of the 'record' being shipped as a piece of software, not a piece of data, and include the EQ and dynamics as working code, rather than having to render that into the data, and for the "record" to adapt itself to the playback device intelligently.
One thing I really believe is that music does not respond well to mediocrity. Unfortunately if music isn't *really* good, captivating, then it's boring / annoying. These days the valuable, quality musical experiences that people have are very, very rare, and that is killing the music business.
Thing is, for there to be quality sound, you have to eliminate the playback bottlenecks.
At the moment, the best market for audio in that sense is in the live setting, and the majority of "live" or "club" music performances are actually of recorded music. The capabilities of a contemporary PA exceed anything we've ever had before, especially from a domestic speaker system.
The other good market (where they care about sound quality) is film.
The domestic market, 'rock/pop' is reaching rock bottom. One of the best indicators of what's happened becomes visible looking at Altec Lansing speakers from the 1970s compared to today. They could have made today's speakers in the 1970s, but nobody would have bought them. In the manufacturing scene, today you'll find that pretty much every single studio converter box, CD player, car headunit or domestic hifi is actually a cocktail of IC chips made by one of three or four companies. What we're really seeing today is stealth monopolies, which is why it's possible for the quality to drop massively, the prices go up, and sales continue (allegedly) unabated. The manufacturing is where the money is. The customers don't (currently) have any available alternatives which will provide better quality sound in their local shopping center, the entire shelf of consumer electronics says "sorry, it's gonna sound pretty dire" and if you take the stuff apart, you'll find the same culprits within. It's like the mafia, they've taken over the whole neighbourhood. The myriad and hugely powerful influence of private equity firms and semiconductor conglomerates manufacturing in the far east is the root issue. This isn't just happening to domestic sound playback devices. It's happened to most major consumer markets. From medical drugs to your dinner to your shoes, the catalogue of seemingly different brands belie a monopoly grip on the raw materials and capital goods which go into producing them. Thing is, if big business operates by gouging out money–value-profitability that way, diminishing the quality across the board by an aggregate of monopoly control over resources and profit extraction from each brand via various company-vampirising financial instruments or "investment funds" (bain capital, Mitt Romney, I'm talking about the likes of you), it doesn't kill just one company, it kills the entire market. It isn't just the music business, the music business is but a tiny microcosm of what is happening to every "industry" or marketplace.
Nonetheless, there are notable examples of record companies that avoid pandering to the frankly crippling issues of post-80s "consumer-grade" ie "rubbish" or "junk-food" sound quality. Whether it's Chesky records or Warp records (who sell 24/96 Wav files via their marketplace website X Desk), you won't find them shrinking or losing double-digit percentages of their market share every quarter.
It should be noted that the sort of labels that care – and whose customers care – about sound quality are also typically interested in the artistic quality of the actual music. They're in the music business.
Separate names with a comma.