If texting has taught us anything, it's that people can quickly figure out abbreviated, misspelled, or otherwise mutilated words pretty quickly. But more impressively, our brains automatically fill in the gaps and omissions, which is also what happens when we listen to compressed music. Because in order to shrink a file, you need to throw things out, which was exactly the breakthrough of the MP3.
Or, "breakthrough," according to many sound engineers who feel we've lost a great deal in digital. Here's a presentation given Saturday morning by Andrew Scheps at the Grammy futureNOW symposium in Los Angeles. Scheps has engineered and mixed for bands like Metallica, Red Hot Chili Peppers, Weezer, and U2, among others.
The first is a visual representation of what a high-quality, master recording might look like, if expressed in written form.
Compress a little...
A little more...
Casey Saturday, June 23, 2012
I no doubt have different opinions than many, but I personally feel people willing to purchase digital music are entitled to .wav files if they choose. It doesn't have to be master recording quality, but why are people still to this day not getting CD quality audio from buying legal music downloads? They are paying full price and only getting half the product. Actually not even near half. There is no reason for it.
Visitor Sunday, June 24, 2012
You are correct they are not getting near half.
But what is much worse is that the compressed file people think they are buying is often just a license to use the song on select digital devices.
Jeff Robinson Monday, June 25, 2012
Are you suggesting a drop in sales price/licnesing fee of the consumer product based upon how much audio has been removed from the original master?
Petros Monday, June 25, 2012
The only reason I can see why when you purchase a legal download you receive an MP3 file is down to the size of the file which is small, making it easier/quicker to download... but with that said, we are in an age where we have fast internet so you're right that a high quality version of the file should be optional to the purchaser. MP3 download related more in the dial up days where you'd leave the computer on and downloading overnight.
Things in the music industry need to evolve with the rest of the world, no use clinging on to ancient methods if we are to make progress in the future.
Anthony Polis Saturday, June 23, 2012
That was probably my favorite speaker of the day. Glad people are addressing this issue...music should be enjoyed to its fullest extent!
Jos Smolders EARLabs Sunday, June 24, 2012
that's an aweful lot of hair. Could do with some cmprssn. Totally agree with the story (as deduced from pics).
Clintone Sunday, June 24, 2012
If one thing the cassette has tought us is that people will always chose convienence over quality. The vynal record sounded so much better that tape. This is the exact same thing.
wallow-T Monday, June 25, 2012
The tradeoff between fidelity and portability has been ongoing since the dawn of sound recording.
Vinyl was a compromise when compared against the live concert experience. And before cylinders and vinyl, music at home was a privilege of Royalty. :-)
The Nyquist theorem guarantees that CD audio (.wav or FLAC) will yield a perfect reconstruction of the wave form between the sampling limits. My belief is that there were about 8 years of tinkering with the necessary filters to get the high and low end removed from an analog signal before digital encoding. If you want to use the old Harman-Kardon argument that response > 20,000 Hz is important, go for it. But "pieces are missing" from CD audio is a crap, anti-science argument.
Vinyl apologists gloss over noise, record and stylus wear and tracking error. (I own about 1000 LPs.) I've seen no sign that the bulk of today's vinyl enthusiasts have any concept of cartridge alignment issues.
Cliff Baldwin Sunday, June 24, 2012
If anyone on this thread, or the speaker with all that hair, or any normal music fan, could hear the difference between a modern 256K AAC file and the original master (or the CD, or the .wav for that matter) then this discussion wouldn't be a complete waste of time. I'm sure when bandwidth, battery life, and CPU on smart phones etc. gets to a point where it is more practical to deliver lossless, everyone here will celebrate by dancing in the streets. But y'all still won't be able to hear any difference or tell them apart. By the way, did you know that some people can hear in colors? Those are the same people who can hear the difference between 256 and 320. There are two of them in the world. Maybe three.
Casey Sunday, June 24, 2012
If you use piece of crap iPod headphones, you won't probably be able to tell the difference. If you actually care about how music sounds and buy an even semi-decent pair of headphones/ear-buds, you will be able to tell the difference betwen 256kbps aac and .wav.
paul Sunday, June 24, 2012
Hair discussions aside, Scheps actually conducted a listening laboratory right after this discussion. He used a pair of $120,000 speakers, and offered listening breakdowns between all the various formats and codecs for all types of music. I was sitting next to two sound engineers who said they could tell the differences between all the gradations - master to CD to high-end Spotify to iTunes, etc. Maybe I've blasted my speakers one too many times, but I often had trouble distinguishing between the stairsteps.
In many cases it was blatantly obvious. Shifting from CD to YouTube, for example, or even iTunes to YouTube. Other times I thought I was hearing quality differences but couldn't really be sure if I was simply biasing my opinion based on knowing the compression specifics beforehand.
The really interesting part is that some formats seemed to bring out different parts of the music - vocals, bass, etc., though in some cases, music is specifically engineered for vinyl, etc. So that's a complicated investigation.
To me, I think there's a worthwhile discussion to be had about what the audience actually hears or cares about when it comes to quality.
Aaron Monday, June 25, 2012
It would be nice if one of these 'listening tests' could be conducted blind for once. Ben Goldacre does a brilliant job of explaining blind and double blind experiments in science (and why not doing so is essentially a waste of time.)
Aaron Levinson Monday, June 25, 2012
Cliff is referring to synesthesia and many people experience it including my friend mike Fossenkemper at Turtletone mastering.
To suggest that only 2 or 3 people in the world have this is totally absurd. I can tell the difference between different compression levels and in general the person that wrote the original comment is an idiot.
mdti Wednesday, June 27, 2012
The difference is SO OBVIOUS... except for deaf people or those who have earing problems.
Of course, not if every point of your listening chain is crap too.
Bree Sunday, June 24, 2012
In this case you can def give me .flac - and my AKG's please. It depends on the delivery system as much as the file. Overall if you listen to crap you get used to crap. If you're in too much of a hurry to focus on quality in life, then that's your experience. Is there hope for the future of your ears? It depends on where you been takin them.
R.P. Monday, June 25, 2012
Flac/Lossless doesn't sound so bad. Just download in Flac/Lossless if you're a true audiophile and require it. Fact of the matter is not everyone has $100,000+ home systems to really grasp the difference. Even Beats by Dre phones add more bottom so you can't even differentiate at times.
Visitor Monday, June 25, 2012
Part of the issue with digital audio is that it is always a lossy format. Even FLAC files are losing information compared to an actual analog sound
As a result, listening to digital music is inherently more work for your brain because you are constantly filling in the gaps subconsciously. And the lower the bit rate, the more interpolating the brain has to do. This makes listening to music fatiguing.
I imagine most people who make the argument that no one can hear the difference have never actually taken the time to listen to the difference. And once you replace a high quality product with cheaper lower quality ingredients on a mass scale, people forget and begin to feel an ego attachment or bonding to the new inferior replacement. The ego can't cope with the possibility that it has been duped into accepting something inferior so the rationalization takes over and all of a sudden the people who point out that "the emperor isn't wearing any clothes" are branded the weirdos.
Digital music is ok for people who don't care about quality. It is not a really great way to listen to music. But it sure is convenient. Just like McDonalds
J Red Monday, June 25, 2012
I really love the McD's analogy. You get a big mac and it looks, feels and tastes (sort of) like a cheeseburger. It has something that represents meat, cheese, bread, and veggies. When you eat it you are full, but it is no where near the same experience as scarfing down a big juicy burger from a nice resturant. Yes you have to walk into the resturant and Yes you have to wait for them to grill it up, but YES it is worth it! Well to me at least.....
- J Red
Aaron Monday, June 25, 2012
Plain text is a pretty terrible way to represent an audio file.
A more fair comparrison, taking psychoacoustic effects in the algorhithyms into account, would be to place all the text from the works of Shakespeare on the screen and then remove the odd punctuation mark and see if anyone in the audience could spot the difference. Yes if you read it all in detail you'll notice the subtle diffence, but most of the people in the audience definitely wouldn't.
AScheps Monday, June 25, 2012
As the guy with all the hair I figured I'd chime in. These are all my opinions and not meant to be stated as facts. I didn't have an agenda for the listening room, I'm just trying to get some awareness going out there.
First of all, for those that came Saturday, thanks. The listening room was a big success, and it was eye opening even for me. I went in expecting it to be harder to hear the difference between high bit rate lossy and lossless files, and easier to hear the difference between 44.1/16 and the high res material. It turned out that the lossy to lossless comparisons were night and day (regardless of bit rate) to most of the people in the room, and the low res/high res lossless comparisons were more subtle.
Another interesting side note was the incredible range of quality for the vinyl. Some sounded amazing and some sounded terrible. I guess it's obvious that this would be the case but I wasn't prepared for quite how bad some of it was.
Anyway, to address a couple of the comments:
1) We used the $125k PMC speakers, because, well, we needed to use something that could fill a room with 50 people in it, and they generously offered to bring them. It certainly makes it easier to hear the differences, but I could tell the difference between the lossy and lossless formats when I was checking my source material on my laptop speakers the night before. That said, I wouldn't advocate filling up your iPod with wav files.; too big and probably not worth the payoff playing back on a portable device.
2) These sorts of experiments HAVE been done double blind in a controlled environment by Harmon (the details were in my presentation). The quick rundown of the results is that 70% of the time the listeners preferred CD over mp3.
3) I used an external D/A for playback. I can still hear the difference on the laptop's built in audio, but it goes without saying that the higher quality playback devices you use, the more stark the differences.
4) While I know that my plain text analogy isn't exactly correct in terms of scale, it was the best analogy I could come up with to get the point across. Making people understand the concept was my one goal, and I think I achieved that. Also, I would refute the argument that the actual process is like taking some punctuation out of the complete works of Shakespeare. Mp3 files are, on average, 10-20% of the size of the 44.1/16 wav files they are made from. Therefore you are intelligently removing around 80% of the information in the original file. That looks a lot more like my second example than Shakespeare with some commas removed.
5) The whole point of the presentation was to get the conversation started, and I've succeeded on some small level to do that, so I'm happy.
For anybody that wants to test it, just get a CD and rip it into iTunes as a 256 aac or mp3 and then as a wav, AIFF or ALAC and compare the two files. If you don't hear the difference, then cool, you're good with mp3s and other lossy formats. If you do hear it, then buy up CDs while they're still around!
mcrein Monday, June 25, 2012
for what it's worth, i think you're 'writing' analogy did exactly what an analogy should do. illustrate a point that is not easily quantifiable. though my hair is not as long as yours, it is definitley as grey. i've been around for the whole technological development from early 70's to today. i distinctly remember looking at a graphical representation of a digital sound wave back in 85 and wondering where all the information of the original sound wave went. even at 24 k per second, that still meant a lot of information was being skipped over.
in those days we could really hear the difference. digital sounded, well colder was the general term. was that because we weren't getting all the information? maybe.some of it for sure was due to the fact that few engineers back then could mix properly for digital. the early digital recordings sounded terrible, by and large. of course today that has changed. i personally can't distinguish between full rez lossless and audiophile analog. but i certainly can hear the difference between mp3 and 44.1/16, especially 44.1/24. so much so that i can't really listen to mp3 or even radio.
that having been said, most people don't care. they don't listen to music for music's sake. music is the backdrop to their lives. in fact if it intrudes to much, they'll turn it down. back in the day it was really no different. 45's were played on little portable record players sometimes with a darning needle as stylus. as long as people could hear the lyrics, sing along, dance and whatever else they wanted to do, the recording served its purpose. for them at least.
digital compression has made music so much more convenient. 5,000 songs on one little phone. think of it! is the quality there? no one cares. i hate mp3 but i'm forced to use it when pitching for film or commercial spots. why? because that's what they want. everyone is looking for that loud compressed sound. they're so used to it. so that brings us to the point: you can't miss something you never had nor have no need for.
so, i wonder sometimes why we put so much effort into recording and mixing. no one really cares about all the subtleties, as long as it's loud. people don't really know what things sound like in the real world anymore. i've had pieces come back from music supervisors telling me the violins or brass sounded fake. when it was real violins and real brass, recorded with real musicans. they wanted that big hollywood movie sample library sound. now i work with sample libraries.
thanks though for your effort. nice to know some do notice.
cipher Tuesday, June 26, 2012
I agree with what you say...my history with audio goes back to 1951 so have been involved with the progress of technology in it's different formats for a long time.(my first recorder was a German wire recorder).One cannot help arriving at certain opinions rightly or wrongly. It is a vast subject with countless variables. In my experience here are a few..vinyl pressings after about 10000 start to deteriorate. Hence, number 2000 pressings quality is better than say a number 12000 pressing. One German company would only press 6000 records before replacing the master however they were in the minority.Vinyl quality reproduction requires quality equipment for best results especially the turntable. It is difficult to put a dollar figure on it but lets say $10000 minimum. There is music and there is entertainment music...most of us like being entertained so quality is no always important...then there are music lovers who demand quality... they are discerning...usally own high end equipment and play their music in a dedicated environment.A niche market.Vinyl is tops but at a price...A poor quality pressing played on poor quality equipment ...sorry to say this but reproduction is crap.We move on to modern digital and compression. Put simply..for the dollar spent and convenience mp3 and the like are fine...we are entertained in trains, planes, walking, running, standing on our head or wherever we chose.That cannot be achieved with vinyl.Lets take it as it is..want quality with either vinyl or digital.. spend the dollars, have your ears checked and build a dedicated environment.Or,go the other way for it's convienience and price be entertained enjoy what is coming through the low price equipment and live a little. I would like a new Porsche but have to settle for a standard boring sedan...but gee I have got some great sounding equipment...think I'll trade it for a Porsche...as they say no fool like an old fool.
pm Tuesday, June 26, 2012
Finally people are waking up and hearing the difference...... Digital does not necessarily mean high quality, the Ds are not a guarantee for this. Using 24 bit resolution for recording and taking advantage of the dynamic range available is good enough for current CD reproduction 16 bit 44,1 or 48kHz on DVD. As a PRO AUDI0 professional, I do not listen to MP3s or other compressed formats, therefore my ears are still well conditioned to enjoy good music. IT IS ALSO A MATTER OF MUSICAL TASTE, which gets killed by compressed formats.
BE CAREFUL WTH HEADPHONES !!!!!!!
cipher Wednesday, June 27, 2012
Here is a little "something" that has no scientific proof... is only an observation and has no merit whatsoever...but to my ears or imagination is real. Compressing music generally "takes away" frequencies the human ear cannot hear...However, I believe that when certain frequencies are deleted the sound leaving stereo loudspeakers...lets say...are in a skeleton state.Uncompressed sounds when leaving the stereo loudspeaker are full bodied. Here is the part where all the audio experts are going to shoot me down...these sounds (although we cannot hear them) when they leave the speakers "mingle" and create a nuance that our ears hear therefor giving music a "fuller" sound. There it is folks..is it real or not ...I do not know. Maybe sound is in the ears of the beholder and differs from person to person.
Andy Wednesday, June 27, 2012
Wasn't there, thanks for this overview. Got a lot out of Mr. Scheps talk in Vancouver a while back, and more recently on Pensado's Place.
Wondering what everyone here thinks about the "Mastered for iTunes" development? In the right hands, does it significantly improve the quality of sound of iTunes-purchased music? Are you seeking out Mastering Engineers who offer it?
Andy Wednesday, June 27, 2012
whoops, I see there's been some coverage here from last September:
Maybe the better question (if it's not too off-topic), is: how has MfiT been stacking up since then?
Brian Cauley Wednesday, June 27, 2012
i think this presentation is somewhat misleading. yes, compression takes out a considerable amount of the detail in audio, leaving holes for our brain to fill in. yet, those holes are much smaller and spread out than taking letters out of words. the problem is that this example confuses the way we hear. unlike reading, where each letter carries almost equal weight, certain frequenicies are more subtle. so, you loose texture when you compress, like taking out entire words or phrases in a story (think Tolkien with half the descriptors). you get the idea, but you miss nuance. compression uses alograithms the way your Mac sumarizes speech. i feel most people reading this forum probably know this, i just can't stand by while someone explains things poorly.
ivan rosenberg Wednesday, June 27, 2012
Congrats on an important concept presented very well. As a recording engineer and musician who can definitely hear the difference, the analogy I use is to compare an original high-quality 300 dpi or better digital photograph (taking up several megabytes) to the same digital photo after it's been compressed to 200k so you can email it to your friends. From afar, and at a small viewing size, the compressed photo conveys the same general visual information. But as soon as you zoom in on the compressed pic, instead of seeing more detail as in the original photo, all you see are blurry pixels. Mp3s and other compressed files lose the detail of the music, lose the peaks and valleys in terms of volume (the dynamics), and therefore lose the overall emotional content of the song when compared to the original un(or less)-compressed track.
Pristine analog vinyl is theoretically better, but as others have pointed out, most listeners have a lot of weak links in their vinyl listening chains. I think that, for practical purposes in today's digital music age, simply offering higher quality (WAV, flac, etc.) downloads is probably sufficient to at least preserve a reasonable amount of a song's integrity. iTunes and many other media players can now play 24-bit tracks, so that's a step in the right direction.
buckbaran.com Wednesday, June 27, 2012
In 2010 I decided to produce my music and purchase Pro Tools, etc. My system quality is way beneath an Audiophile’s. I generate Wav and MP3 masters. Hearing the difference is difficult. The only issue is mastering for the laptop. Egad. Where did the bass guitar go? In my research and testing sending MP3s to my bass player I’ve learned to roll-off the bottom (-24dB cut @ 20Hz) and high-end (-18dB @ 16 kHz). The end result is more separation in the tracks and a little more bottom. I guess rolling-off the ends makes it easier for the MP3 compression algorithm.
Sounds are created analogically but eventually converted into chemical-electrical signals, kind of like digital, where the brain then translates the impulses of energy into recognizable sound patterns. Is the brain being fooled?
Regarding downloads, how much time do you have to spare? IE: 4:40 song length. 4.32 MB MP3 vs 47.7MB Wav. We’re talking roughly a difference 10x file size. Anyway, with mobile listening devices quality goes out the window; reminds me of the good old days of transistor radios vs. home Hi-Fi. By the way, I’m 60, embrace digital and I think vinyl blows.
Tony Cardenas-Montana Friday, July 06, 2012
This is so spot on... Kudos to Andrew!