Most pirated music acts on the internet graphic

[smark21]

The Beatles are #1. Not many R and B acts on this graph.

http://blogfiles.wfmu.org/KF/2011/08/31/Most-pirated-artists1.jpg

[vcq]

Yikes...feeling my age, since most of these bands don't ring a bell. I could make out Marvin Gaye. Some of my faves like Bat For Lashes and Florence + the Machine are on the list. Loads of indie / emo / techno, no surprise there. I *was* surprised to see so many classical composers and jazz / easy listening, though [[I wouldn't have considered that the "target audience" for pirating).

I work in graphics / publishing layout; give that graphic artist a gold star and a raise! I shudder at the thought of how long that must have taken to set up with so many layers and varying font sizes...

[soulster]

Oh I know it!!!

Even though they are illegal, I look around at what is available for high-quality vinyl rips, and all I see is rock, blues, and jazz. There is almost no R&B. I hate to say it, but it's true: most of the people who seem to care about high-quality sound are white males, so the kind of music most of them gravitate to are going to be that demographic. Besides, that group also has more disposable income.

On the other hand, when I look around to see what kind of R&B is available, all I find are crappy vinyl rips with skips and scratches, played on Mickey-Mouse turntables, done at low-bit-rate mp3s, with all the noise left in, or CDs ripped at low bit-rates. The rock and jazz is done with gear/software that costs an average of $2000, and ripped to lossless FLAC. The rock guys are adding all the graphics, liner notes, and technical information of what they used and did. The same thing that the record labels should be doing. Not even HD Tracks offers all of that!

All this tells me that most R&B lovers who get their music illegally don't care about sound. That makes me all the rarer.

[delta]

As the creator of the original graphic, I've looked around to see where it's popped up, and thought I'd sign up here to share a few thoughts, particularly in light of soulster's comments.

It may interest you to know that the graphic's original form was actually made for a 48" x 36" poster for the PAF [[People's Arts Festival) in 2010, which is held annually down at the Russel Industrial Center in Detroit. After bringing photography to the show for three years running, I was sick of not actually selling anything. I had some excellent pieces, but photography is one of the hardest artworks to sell, because you're not just competing with other photographers, but in an era of democratic digital photography, you're also competing with the customer - if it takes me ten shots to get the perfect picture I want to share with others and sell, and it takes a potential customer 1,000 shots, without an overhead of film costs, the only difference between 10 and 1,000 is persistance. Further, at the PAF there is always plenty of photography for sale, and most of it very good.

Shortly before the 2010 show, I was trying to think of something else, something involving data visualization. I was already doing a statistical analysis for the world's largest private tracker annually for a few years, so I had this fantastic rich dataset already compiled and begging to be used. I also had access to wide-format printing, and due to my previous experiences with photography, I decided to keep my costs as low as possible by printing on traditional bond paper as opposed to card stock or mylar [[translucent mylar is drop-dead gorgeous, but expensive), so that literally anyone who was interested in the poster could afford to go home with it.

It's great at starting a few different kinds of conversations, about: music, demographics, the culture of piracy, statistics, and typography & design.

Soulster, you've got a good eye for one of the most important considerations - demographics play a key role in why the statistics are vastly different than record sales from the same period. The source of this data is listed as "the largest private music tracker in the world", and the demographics aren't the same as the general population. Now since this is "only" the top 10,000 of over 300,000 artists downloaded, even being on the list in the smallest typeface represents a good deal of downloading, but it's true that a lot of stuff is completely out of synch with sales and radio play.

Certain demographics come into play. While the tracker is not US-only, it is heavily dominated by the natively English-speaking countries, and those countries for whom English is a common second language [[Scandinavian countries and Slovenia are, in particular, very well-represented). The age of the users is in a left-weighted bell distribution, heavily favouring 18-28 year olds. The gender balance is heavily in favour of males. Due to the costs of broadband, computers, and hard drives [[data hording is a hobby), the demographics lean heavily middle class and suburban - the interests of urban poor are under-represented, as are the interests of the rural poor - both R&B and country are both under-served. For vinyl enthusiasts, as you point out, there's even a bigger money hurdle to jump through to get "good" quality rips, and sites like this have pretty exact rules about achieving good audio quality.

It turns out what people are most interested in now is what is new now - not just new bands, but new releases by older bands. The release of The Beatles' box sets was the largest music piracy release event in history - and probably will remain that way. While this may sound like an astonishing statement, here's my thoughts on the subject: there is no other band in the world with as many old AND new fans whose CD-era releases have been so poorly released by its record label - about half of The Beatles' CD-era releases were mastered very poorly, and the box sets correct this. No other band with a similar stature can quite match the reach of The Beatles or has been as long-under-served. Led Zeppelin's releases on CD have sounded great since their early 90s box sets. Most of The Rolling Stones' releases have been mastered well. Pink Floyd's catalogue is looking and sounding good for a while, and though it's due for a re-release later this year [[and it's going to be another big piracy event), there isn't as much pent-up demand. Dave Thompson's book "I Hate New Music" [[http://www.amazon.com/hate-Music-Classic-Rock-Manifesto/dp/0879309350) has interesting insights into another issue. Though the book is mostly a classic rock screed, and intended to be subjective and funny [[and it is, even though I highly disagree with its main premise), it makes a great point about how before 1980, the music wasn't as self-concsiously separated by genre. It was not unusual for people to have Joni Mitchell, The Beatles, Otis Redding, and Perry Cuomo all represented in a collection. Today, it's not only genres which are important, but micro-genres - stuff like dubstep, minimalist techno, ambient drone metal are so much more specifically defined than the genres they came from that even the language of sub-genres becomes very opaque to those not following it pretty closely.

Interestingly, The Beatles' poorly mastered CD releases are important to why a lot of people still think that vinyl sounds better than CD. I think it's fairer to say that well-mastered vinyl sounds better than poorly-mastered CDs, and the solution is to use good masters and not engage in the loudness wars where songs are brickwalled to sound "loud" for radio and mass-market CDs.

Another interesting facet of this is that when people DO download music from earlier decades, it's almost never the music that was most popular in that decade. People don't download 80s Madonna as much as they do King Crimson. They don't download as much 90s Pearl Jam and Nirvana as they do Aphex Twin. It gets REALLY interesting in the 1940s and 1950s. Since the sample set [[number of downloads) is much smaller, it's harder to make broad generalizations, but in each decade there are about ten artists who do better than Frank Sinatra, none of which were top 40-selling artists during their day. It turns out that people don't download Perry Cuomo or the Andrews Sisters [[and why would they, Cuomo has aged as poorly as Sinatra, but he has less of a cult of personality around him since he was a better human being on all fronts - Sinatra's ugliness as a person keeps his legend alive), they download classic jazz [[vocal and instrumental). Miles Davis is a great example of this era. Some of his albums released in that timeframe still get downloaded directly [[new anthologies and box sets are calculated as the year they were released, which pushes even a lot of classic artists into more recent decades). Take the comparison with Elvis, though - most of his records from when he was at the peak of his powers weren't Elvis sitting in a studio thinking about presenting a 45 minute album with a specific running order - his label would record singles and b-sides, and when it had enough, it'd put them onto one LP - the same as many [[most?) artists of the era. Miles Davis's "Kind of Blue", however, remains a cohesive artistic statement that doesn't lend itself to chopping into bits for a "best of" anthology. So, it turns out, people aren't downloading "Jailhouse Rock" - they're downloading Elvis compilations - but "Kind of Blue" remains a distinctively relevant artistic statement as a whole.

I actually made decade-specific posters for the 1940s to the 2000s [[2000-2010). I'm not quite as happy with them, but they do have different stories to tell other than the main poster. One drawback of all the posters isn't obvious to people not familiar with the source, however. Most private torrent trackers are ratio-driven. This means if you upload 200 MB and download 100 MB, you have a 2:1 ratio. These sites usually require [[or give incentives to) users to have a specific minimum or optimal ratio. Since this is a kind of "closed economy", much like the Federal Reserve, sometimes you need to pump credit into the system. This is done by bonuses on some sites, and freeleeches on the other. This site doesn't use bonuses, but does use freeleeches. For a freeleech, your download isn't counted, but your upload is. So, if you download a freeleech album that "costs" 300 MB, and upload 200 MB back to other users downloading the same album, only your 200 MB upload counts. So, what happens is that some albums which are freeleech are downloaded by people whose only interest in the album is to download it and get credit for seeding [[uploading) it to other users, and who will never listen to it. This is why some artists are unusually well-represented, such as Creedence Clearwater Revival [[though their career box set was getting a fair amount of "normal" attention beforehand).

Other artists [[Vampire Weekend, Animal Collective) are as big as they are due to being darlings of indie rock journalism. It's fair to say that Pitchfork Media is as influential in rock circles as Rolling Stone Magazine was in the 1970s. If they champion your work, you're going to get a lot of record sales and downloads.

[delta]

The forum software needed me to shorten to 10,000 characters, so here's the remainder:

As far as FLAC vs MP3, I'm more interested in transparency [[if you AB-X test something, transparency means that you cannot determine which signal [[A or B) matches signal X any better than 50% - guessing) than anything else. Since the site bottoms out at 192kbits per second, this is transparent for anyone with cheap equipment. The LAME mp3 codec's v2 variable encoding is about 190-220 kbits/sec, which is transparent for most people with typical consumer-grade equipment, and the codec's v0 setting is fantastic - it's effectively transparent to CD/FLAC in all but the most specific circumstances [[the only time I could successfully ABX v2 from FLAC was quiet cymbal work at excruciating volume, and then couldn't ABX v0 from FLAC in the same section). Then, there's a large FLAC contingent which is by definition transparent to CD, and transparent to vinyl in all cases except when the record is in bad condition or the recording equipment is cheap or improperly-configured. In fact, one of the original reasons I was doing these statistics was to discover how Ogg Vorbis was doing compared to MP3 and FLAC [[and even AAC, AC3 and DTS) [[answer: extremely poorly indeed), to compare interest in vinyl to CD and other less-common formats, lossy to lossless, etc.

If any of you have any questions, I'd love to answer them. I'll post links to the decades-specific poster later after I get back from some last-minute moving tasks.

[vcq]

Delta,

I just learned about wordle [[http://www.wordle.net/) since I'm in ed tech [[I do some simple graphic layout as a curriculum developer w/ InDesign CS5 and I've tinkered around with Photoshop for over a decade). Could you give a bit of insight into HOW you created your graphic?

Much appreciated,
VCQ

[smark21]

Hi Delta, I found your graphic posted at a board for The Killers and decided to share it over here. It was very interesting and well designed. Good job. Over at the Killers board, the fans considered an odd accomplishment that their favorite band seems to be pretty popular with those who prefer to pirate rather than buy their music.

I'm wondering if there is any correlation between acts with fanbases who put a lot of stock and emotional investment in sales and chart position and being pirated? I know some of Madonna's most die hard fans love to trump how successful she's been on the charts over the years. Here on this board, the most sales and chart position obsessed members tend to be Diana Ross and/or Supremes fans. On the other hand, there are music fans, especially of indie acts, who look down on sales and charts--would they be more likely to pirate?

[soulster]

Interestingly, The Beatles' poorly mastered CD releases are important to why a lot of people still think that vinyl sounds better than CD. I think it's fairer to say that well-mastered vinyl sounds better than poorly-mastered CDs, and the solution is to use good masters and not engage in the loudness wars where songs are brickwalled to sound "loud" for radio and mass-market CDs.


A lot of people disagree with you on this. I thought the 2009 Beatles remasters, [[both stereo and mono) are so good that I got rid of my Beatles vinyl and Dr. Ebbits needledrops.


The forum software needed me to shorten to 10,000 characters, so here's the remainder:

As far as FLAC vs MP3, I'm more interested in transparency [[if you AB-X test something, transparency means that you cannot determine which signal [[A or B) matches signal X any better than 50% - guessing) than anything else. Since the site bottoms out at 192kbits per second, this is transparent for anyone with cheap equipment. The LAME mp3 codec's v2 variable encoding is about 190-220 kbits/sec, which is transparent for most people with typical consumer-grade equipment, and the codec's v0 setting is fantastic - it's effectively transparent to CD/FLAC in all but the most specific circumstances [[the only time I could successfully ABX v2 from FLAC was quiet cymbal work at excruciating volume, and then couldn't ABX v0 from FLAC in the same section). Then, there's a large FLAC contingent which is by definition transparent to CD, and transparent to vinyl in all cases except when the record is in bad condition or the recording equipment is cheap or improperly-configured. In fact, one of the original reasons I was doing these statistics was to discover how Ogg Vorbis was doing compared to MP3 and FLAC [[and even AAC, AC3 and DTS) [[answer: extremely poorly indeed), to compare interest in vinyl to CD and other less-common formats, lossy to lossless, etc.

If any of you have any questions, I'd love to answer them. I'll post links to the decades-specific poster later after I get back from some last-minute moving tasks.

FLAC, is the now-standard lossless format. 320 mp3 is the next lossy standard along with lossy AAC. Ogg Vorbis beats them both, but was never accepted as a standard in the U.S..

Even on modest stereo gear, I can hear the deficiencies of 320 kbps mp3 and AAC. Why? as musician and audiophile, I have trained my ears to hear the artifacts and/or sound degradation.

[delta]

I just learned about wordle [[http://www.wordle.net/) since I'm in ed tech [[I do some simple graphic layout as a curriculum developer w/ InDesign CS5 and I've tinkered around with Photoshop for over a decade). Could you give a bit of insight into HOW you created your graphic?

The program I used is based on Wordle. I needed what Wordle does, but more so - more words and a higher level of DPI. I had a friend write me some new code with those design goals in mind. I knew I wanted the top few artists horizontal, and the rest mixed between horizontal and vertical. Unless they've changed Wordle from when I was experimenting with it, you don't have placement control. But one thing computers are great at is doing the same thing over and over. I had my system work overnight and create a few hundred layouts, knowing one would be exactly how I wanted it, and then I chose the one that fit best. Most of the actual work was getting the data, but by the time it occurred to me to create a poster, that part of the work had already been done. The first time I did it, the data generation took me about 7 weeks - mostly web scraping and then normalizing data, and much of that either maximizing my network input [[data scraping) or my CPU [[normalizing). The last time, due to more streamlined techniques on my part and faster computing power, it took about four days. Turning that data into statistics takes a few more days; I'm actually using Excel for a lot of that. Though I used to program in Access, it's been such a long time that I determined it would probably take me longer to re-learn Access enough to do it right than to just use brute force in Excel.

I also wanted to make a statement about typography. The typeface used is Constantia, developed for Microsoft by Tiro Typeworks for Microsoft in 2003-2004, and released with every copy of Vista, Windows 7, and Office 2007. It is a tasteful modern hybrid of classical and contemporary styles, and raises the bar for default type options in Microsoft products. It includes glyphs for Eastern European languages, and Cyrillic and Greek alphabets. Regrettably, Constantia does not include glyphs for Hebrew, Arabic, Chinese, Japanese or Korean. This is a minor issue for this project: although entries exist for artists rendered in all five of those scripts, there are also Latinized translations, and most of the artists’ names in those characters would have been too small to fully appreciate.

One reason I chose Constantia is that not only is it one of the first time Microsoft commissioned something new in typography that was actually good, but it almost makes up for Comic Sans and Arial. Almost.

[vcq]

Thanks, Delta :-) I want to try experimenting with Wordle, but I wanted something more powerful [[and with more options re: font, leading, etc.).

[delta]

I'm wondering if there is any correlation between acts with fanbases who put a lot of stock and emotional investment in sales and chart position and being pirated? I know some of Madonna's most die hard fans love to trump how successful she's been on the charts over the years. Here on this board, the most sales and chart position obsessed members tend to be Diana Ross and/or Supremes fans. On the other hand, there are music fans, especially of indie acts, who look down on sales and charts--would they be more likely to pirate?

I just double-checked The Killers - they've never been a staff pick freeleech, so that band isn't getting an "artificial" freeleech boost. That's about the only time you see the fans synthetically altering download behaviour. The thing is, only staff and occasional other users [[VIPs, contest winners) get to pick an album for a staff pick freeleech. So, even if there were a few thousand Killers fans wanting to help them game the system, if one of them wasn't a staff member [[and there aren't that many), there wouldn't be a way to convince anyone to do so. The people who pick albums for freeleech are usually trying to promote favourite artists and under-appreciated gems that they feel very strongly about but don't often get a lot of critical attention. Sometimes, torrents that would otherwise be very popular get this treatment, too, [[The Beatles come to mind), but in the end, it's the whim of the person making the pick.

The Killers do pretty well, and while I can't think of one particular reason, it helps that they are a current act, that they tour regularly, that the site's demographic overlaps pretty well with that of the Rock Band video game, and that they had a very catchy song included as a pack-in [["When We Were Young"). The site has their album "Day & Age" on a list of the top 50 albums that have never been freeleech [[except when the entire site goes on freeleech, which doesn't favour any particular albums or artists). They overlap the site's demographic interests pretty well, have a large male audience and a large female audience without pandering to either gender. If this site [[ http://rockbandscores.com/popularity.cgi ) can be believed, their hit "Mr. Brightside" is in the top 20 most downloaded songs for Rock Band as well. So, pretty popular on all fronts, it doesn't suprise me that they get downloaded a lot.

[delta]

A lot of people disagree with you on this. I thought the 2009 Beatles remasters, [[both stereo and mono) are so good that I got rid of my Beatles vinyl and Dr. Ebbits needledrops.Actually, the "bad" masters I mean were from the early 80s. The 2009 remasters were gorgeous, and I've heard other Ebbits' fans say the same thing as you're saying. It's the first time some of those albums sounded right on CD. So, my point is that when mastered from a good master, CDs can sound every bit as good as well-treated vinyl. A lot of people with a vinyl bias were rightfully scared off by early CD releases not because CDs are inherently worse, but because many of those early 80s reissues were rush jobs.

One thing that [[occasionally) comes into play is that the CD was originally designed to be a 14-bit medium, but before the first consumer product was made, the specs changed it to a 16-bit format. Some of the old CD reissues sound sub-par because some companies did their digital conversion to 14 bit, then changed to 16 bit, either releasing a 16-bit source, or a 14-bit source which was treated with pre-emphasis, which was a scheme to increase the signal to noise ratio of the high end of the audio spectrum. For 16 bit releases, it simply wasn't an issue.


FLAC, is the now-standard lossless format. 320 mp3 is the next lossy standard along with lossy AAC. Ogg Vorbis beats them both, but was never accepted as a standard in the U.S..

Even on modest stereo gear, I can hear the deficiencies of 320 kbps mp3 and AAC. Why? as musician and audiophile, I have trained my ears to hear the artifacts and/or sound degradation.

I won't get into an audiophile "golden ears" argument; I've only known one person who has ever changed their mind on the issue. The one thing I will leave it at is this: if you're not ABX blind testing, you're just confirming existing biases.

The LAME v0 format is designed to conform to a psychoacoustic model, and is able to go to higher bitrates when it will preserve content, then go to lower bitrates for frames where doing so will not audibly alter the content. As such, v0 sounds just as good as 320, but is a bit smaller. I've seen this particular statistic follow through on most audiophile-level trackers that aren't FLAC only: v0 is by far the most downloaded. FLAC tend to be about 18% of downloads, MP3 tend to be about 80%. v0 is about 65% of all MP3 downloads [[about 52% of total downloads), and 320 is about 19% of all MP3 downloads [[or about 15% of total downloads). Over the time period, v2 downloads have gone down, and both v0 and 320 have gone up. Pirate "scene" releases are usually uploaded at 192kbit or v2, but are frequently lossy-to-lossy transcoded, and as such those tend to go away and be replaced by better options as time goes on.

Ultimately, though, quality is about transparency - and MP3 and Ogg Vorbis both get to transparency for either "almost everyone" or "everyone" at a sufficiently high bitrate [[depending on your opinions about lossy formats - I assume you'd agree with the former and me with the latter). There is a very vocal and small contingent of Ogg Vorbis supporters on the site, and while Ogg is allowed, it's only allowed at q.8. Most people who prefer Ogg Vorbis tend to actually dislike q.8 because the quality/size tradeoff is so close to 320 cbr MP3 as to not be worth their time. I sympathize. But, that's a format war that was already over before Ogg was created. The vocal contingent of Ogg supporters tend to be very stubborn, very technical, and in my view more committed to the format of the music they listen to than the music itself. The Ogg Vorbis fans aren't any more successful in Europe than in the US - the site in question gets a lot of activity from both groups. For all of MP3's flaws, we as users of the format kind of lucked out - had FhG been more in-tune with the interests of the RIAA and had the RIAA been as successful as they had been in killing Audio DAT and the early sales of the very first commercially successful MP3 hardware player [[Diamond Rio, which they got an injunction against but later lost), they probably would have built DRM into the format.

I suspect that even those people who are 100% against piracy aren't for the kind of world DRM creates - a world where you buy the song once on physical media, again for your portable player, buy it again for your computer, buy it again when you get a new computer, buy it again for your second portable player, etc. This "toll booth" model is the sort of model the RIAA fought for early on, and thank goodness they didn't have a tool which was better equipped for such use than MP3, or later, FLAC and even Vorbis despite its lack of success as an end-user audio format [[games developers love it for good reason, since it's easy to support both with regards to implementation and licesinsing - or the lack thereof). Ultimately, the FhG were creating an audio format for video digital media, where the DRM is taken care of at a different place than the audio stream level. Thank goodness.

Here's a quickie that will make you wince, though. I have a friend who swears he can't hear the difference between MP3 encoded at 96 kbit/sec and 192 or above, so he lossy-to-lossy transcoded his entire music collection to 96 kbit/sec. Ouch.

[delta]

Thanks, Delta :-) I want to try experimenting with Wordle, but I wanted something more powerful [[and with more options re: font, leading, etc.).

There's an implementation written by the same guy but distributed by his former employer, IBM. What you can use it for is different - read the license - but if you're using it for your own personal project rather than profit, the license may not be an issue for you. This is more similar to what I am using than wordle.net itself:

http://johnlaudun.org/20110405-using-ibms-word-cloud-generator/

http://johnvictoranderson.org/?p=105

Unfortunately, the IBM link is dead as IBM likes to change stuff willy-nilly and old links end up dying. The best approach is to use this old version of the page, click the "download" button, and make a userid on the page it directs you to, then it should download for you:

http://web.archive.org/web/20090419150144/http://www.alphaworks.ibm.com/tech/wordcloud/download

As far as using it, it's moderately user-unfriendly, more so if you're not used to command-line utilities, and even more so if you're not used to Java. If you're doing print work, you'll want to increase the resolution, and if you do that, you'll want to drastically increase the amount of memory heap. When I first tried this incarnation on an XP computer, 4 GB was only good enough if I had most programs closed. Under Win 7 with 12 GB, it's still pretty greedy but I don't have to shut down things for it to work. For design for screen at about 72 dpi, there's less fussing with memory involved.

[delta]

Here's the decade specific posters. Like I said, sample sizes from 1940s to 1960s make the data behind this visualization a lot less "useful" than for more recent decades where the downloads are a few orders of magnitude higher. But, I think they still show interesting trends and there surely is a story to them:

http://imgur.com/a/OZkdb#TMler

If you click the + magnifying glass on the right hand side, it'll link you to a full 1920 x 1200 wallpaper-ready version.

It just occurs to me that the 2000-2010 version is actually a special version I made for a friend - it doesn't have the band "Phish" in it, as per his request, which would have been about twice as large as the next largest artist.

[soulster]

Actually, the "bad" masters I mean were from the early 80s. The 2009 remasters were gorgeous, and I've heard other Ebbits' fans say the same thing as you're saying. It's the first time some of those albums sounded right on CD. So, my point is that when mastered from a good master, CDs can sound every bit as good as well-treated vinyl. A lot of people with a vinyl bias were rightfully scared off by early CD releases not because CDs are inherently worse, but because many of those early 80s reissues were rush jobs.

The first CDs weren't exactly a rush job, it's just that they didn't have good converters back then. The tapes were transferred flat. The new stereo CDs were goosed a bit in places. The mono yapes were pretty much transferred flat.


One thing that [[occasionally) comes into play is that the CD was originally designed to be a 14-bit medium, but before the first consumer product was made, the specs changed it to a 16-bit format. Some of the old CD reissues sound sub-par because some companies did their digital conversion to 14 bit, then changed to 16 bit, either releasing a 16-bit source, or a 14-bit source which was treated with pre-emphasis, which was a scheme to increase the signal to noise ratio of the high end of the audio spectrum. For 16 bit releases, it simply wasn't an issue.

All true. However, many early CDs didn't use the best tape sources. And, again, the converters weren't the greatest.


I won't get into an audiophile "golden ears" argument; I've only known one person who has ever changed their mind on the issue. The one thing I will leave it at is this: if you're not ABX blind testing, you're just confirming existing biases.

Some people do not need ABX testing. If one is honest with themselves, they will recognize when there the power if suggestion may exist and will attempt to verify it. I do. At one time, I was almost convinced that there was a difference between FLAC and the source, but I did the null test.


The LAME v0 format is designed to conform to a psychoacoustic model, and is able to go to higher bitrates when it will preserve content, then go to lower bitrates for frames where doing so will not audibly alter the content. As such, v0 sounds just as good as 320, but is a bit smaller. I've seen this particular statistic follow through on most audiophile-level trackers that aren't FLAC only: v0 is by far the most downloaded. FLAC tend to be about 18% of downloads, MP3 tend to be about 80%. v0 is about 65% of all MP3 downloads [[about 52% of total downloads), and 320 is about 19% of all MP3 downloads [[or about 15% of total downloads). Over the time period, v2 downloads have gone down, and both v0 and 320 have gone up. Pirate "scene" releases are usually uploaded at 192kbit or v2, but are frequently lossy-to-lossy transcoded, and as such those tend to go away and be replaced by better options as time goes on.

I only deal with lossless or 320 kbps encoded with LAME V0. 256 VBR will do in a pinch, but that is it. These days, I also deal with hi-rez all the way up to 24-bit/192kHz. It's all about the sound quality.


I suspect that even those people who are 100% against piracy aren't for the kind of world DRM creates - a world where you buy the song once on physical media, again for your portable player, buy it again for your computer, buy it again when you get a new computer, buy it again for your second portable player, etc. This "toll booth" model is the sort of model the RIAA fought for early on...

I could introduce you to plenty of industry people, or more accurately, those who were in the industry, who absolutely would go for that kind of DRM model.


Here's a quickie that will make you wince, though. I have a friend who swears he can't hear the difference between MP3 encoded at 96 kbit/sec and 192 or above, so he lossy-to-lossy transcoded his entire music collection to 96 kbit/sec. Ouch.

Bad hearing, bad gear, or he doesn't know what to listen for.

[vcq]

Here's the decade specific posters. Like I said, sample sizes from 1940s to 1960s make the data behind this visualization a lot less "useful" than for more recent decades where the downloads are a few orders of magnitude higher. But, I think they still show interesting trends and there surely is a story to them:

http://imgur.com/a/OZkdb#TMler

I'm surprised that my 1940s idol Frankie Sinatra was downloaded less that Django! I own his amazing 12CD Columbia box set [[1943-1952), in addition to the Tommy Dorsey / Frank Sinatra box The Song Is You and his early work with Harry James. Sadly, most of Frankie's early recordings are now out of print.

[delta]

Actually, in the 40s and 50s, literally every artist who is downloaded more than Sinatra sold [[much) less than he did in his own era. A comparison to Perry Cuomo is instructive. Perry Como was a phenomenally successful artist. Meaningful statistical data about 1940s and 1950s record sales are very difficult to obtain, so while I am still looking for evidence for and against this belief, what I have managed to find supports the following conclusion: Perry Como was more popular to more people in his own day of pop stardom than Sinatra in his, or even Presley and The Beatles in their own. Frank Sinatra has a much bigger hold on popular culture because of his controversial personality, private life and public life, so in the later stages of their careers, Sinatra became the staple crooner of new listeners, whereas Como didn't have a "bad boy" image to lure in a new young audience. So, not suprisingly, Como fares much worse than Sinatra in this exercise, but Sinatra is outshined by artists he greatly outsold in his own era. In fact, part of the reason for The Beatles' continued popularity isn't that they had a wide age appeal in their heyday, like Perry Como, but precisely because it was almost strictly a youth culture phenomenon, ushering serious pop music criticism and journalism. Since The Beatles weren't your parents' music, you could "keep" them and grow with them, and their percentage of population that considered themselves fans actually grew as years went on, whereas Perry Como and the crooners became less and less important to people as time went on.

It makes perfect sense. If you're a fan of rock and popular music, and you don't know either Perry Como's or Frank Sinatra's work, it's not much of a handicap, as rock and modern pop is far removed from crooner singers. In fact, you could get a very good job in music journalism and never have to write a single word about either of those artists. Compare this to Miles Davis, who does so well on the list - unlike today's pop and the crooners, Miles Davis is in a much better position: if you don't know Miles Davis, you literally do not know jazz. The 1940s and 1950s were rough years for serious popular music that stands the test of time, whereas the masters of instrumental and vocal jazz were making records which still retain relevance and influence other artists today. So, when people download albums encoded from 1940s and 1950s originals, it's not Sinatra, Como, or the Andrews Sisters so much as the masters of 20th century jazz.

When I grew up, my father listened to the "oldies" stations - CKLW in particular. I HATED listening to it, and it wasn't helped any by my father's tone deaf singing. A few years ago, though, I took up a project of listening to the top 100 songs of each year from 1936 to present. As someone whose musical tastes tend to veer pretty far away from popular tastes, it's a project I often describe myself as "uniquely unqualified" for. I enjoy a lot of minimalist electronic music, some early 90s brit pop, shoegaze, industrial [[Throbbing Gristle-style industrial noise and Front Line Assembly-style industrial dance), Japanese noise, IDM, etc. So, most of what i listen to has very little connection to anything before about 1975, with the exception of avant garde composers and electronic musicians like Edgard Varese and Karlheinz Stockhausen. By the time I got to the part of the project where rockabilly and eventually actual "rock and roll" starts hitting the charts, something amazing happened.

I finally "got" it and understood the context of this music - the oldies I grew up hating. Now, this isn't to say I actually like it any more than what I did before, but now I understand why stuff like Bill Haley and his Comets and Elvis and their peers changed music. What they were doing sounded like a breath of fresh air after getting through a few decades of crooner songs, western swing, and _god-awful_ novelty songs. Haley and Elvis and Cook [[when his blues singles finally appeared in popular charts) were an absolute relief compared to what came before. For jazz, you get off the hook a lot earlier, but the mid-century masters of jazz weren't selling numbers near the top-selling artists of the era, so they don't appear in the top 100 tracks I was listening to for each year. It was an interesting exercise, and gave me a lot of context.

[smark21]

I just double-checked The Killers - they've never been a staff pick freeleech, so that band isn't getting an "artificial" freeleech boost. That's about the only time you see the fans synthetically altering download behaviour. The thing is, only staff and occasional other users [[VIPs, contest winners) get to pick an album for a staff pick freeleech. So, even if there were a few thousand Killers fans wanting to help them game the system, if one of them wasn't a staff member [[and there aren't that many), there wouldn't be a way to convince anyone to do so. The people who pick albums for freeleech are usually trying to promote favourite artists and under-appreciated gems that they feel very strongly about but don't often get a lot of critical attention. Sometimes, torrents that would otherwise be very popular get this treatment, too, [[The Beatles come to mind), but in the end, it's the whim of the person making the pick.

The Killers do pretty well, and while I can't think of one particular reason, it helps that they are a current act, that they tour regularly, that the site's demographic overlaps pretty well with that of the Rock Band video game, and that they had a very catchy song included as a pack-in [["When We Were Young"). The site has their album "Day & Age" on a list of the top 50 albums that have never been freeleech [[except when the entire site goes on freeleech, which doesn't favour any particular albums or artists). They overlap the site's demographic interests pretty well, have a large male audience and a large female audience without pandering to either gender. If this site [[ http://rockbandscores.com/popularity.cgi ) can be believed, their hit "Mr. Brightside" is in the top 20 most downloaded songs for Rock Band as well. So, pretty popular on all fronts, it doesn't suprise me that they get downloaded a lot.

Thanks for the info. I know Killers fan are very devoted and loyal to the group. I've been to several of their concerts and indeed they do have a wide demographic appeal from young kids to people in their 60's, with men and women about equally represented. Makes for a good crowd at a show. And Mr. Brightside is one of the best rock songs of the past 10 years.

[smark21]

Actually, in the 40s and 50s, literally every artist who is downloaded more than Sinatra sold [[much) less than he did in his own era. A comparison to Perry Cuomo is instructive. Perry Como was a phenomenally successful artist. Meaningful statistical data about 1940s and 1950s record sales are very difficult to obtain, so while I am still looking for evidence for and against this belief, what I have managed to find supports the following conclusion: Perry Como was more popular to more people in his own day of pop stardom than Sinatra in his, or even Presley and The Beatles in their own. Frank Sinatra has a much bigger hold on popular culture because of his controversial personality, private life and public life, so in the later stages of their careers, Sinatra became the staple crooner of new listeners, whereas Como didn't have a "bad boy" image to lure in a new young audience. So, not suprisingly, Como fares much worse than Sinatra in this exercise, but Sinatra is outshined by artists he greatly outsold in his own era. In fact, part of the reason for The Beatles' continued popularity isn't that they had a wide age appeal in their heyday, like Perry Como, but precisely because it was almost strictly a youth culture phenomenon, ushering serious pop music criticism and journalism. Since The Beatles weren't your parents' music, you could "keep" them and grow with them, and their percentage of population that considered themselves fans actually grew as years went on, whereas Perry Como and the crooners became less and less important to people as time went on.

It makes perfect sense. If you're a fan of rock and popular music, and you don't know either Perry Como's or Frank Sinatra's work, it's not much of a handicap, as rock and modern pop is far removed from crooner singers. In fact, you could get a very good job in music journalism and never have to write a single word about either of those artists. Compare this to Miles Davis, who does so well on the list - unlike today's pop and the crooners, Miles Davis is in a much better position: if you don't know Miles Davis, you literally do not know jazz. The 1940s and 1950s were rough years for serious popular music that stands the test of time, whereas the masters of instrumental and vocal jazz were making records which still retain relevance and influence other artists today. So, when people download albums encoded from 1940s and 1950s originals, it's not Sinatra, Como, or the Andrews Sisters so much as the masters of 20th century jazz.

When I grew up, my father listened to the "oldies" stations - CKLW in particular. I HATED listening to it, and it wasn't helped any by my father's tone deaf singing. A few years ago, though, I took up a project of listening to the top 100 songs of each year from 1936 to present. As someone whose musical tastes tend to veer pretty far away from popular tastes, it's a project I often describe myself as "uniquely unqualified" for. I enjoy a lot of minimalist electronic music, some early 90s brit pop, shoegaze, industrial [[Throbbing Gristle-style industrial noise and Front Line Assembly-style industrial dance), Japanese noise, IDM, etc. So, most of what i listen to has very little connection to anything before about 1975, with the exception of avant garde composers and electronic musicians like Edgard Varese and Karlheinz Stockhausen. By the time I got to the part of the project where rockabilly and eventually actual "rock and roll" starts hitting the charts, something amazing happened.

I finally "got" it and understood the context of this music - the oldies I grew up hating. Now, this isn't to say I actually like it any more than what I did before, but now I understand why stuff like Bill Haley and his Comets and Elvis and their peers changed music. What they were doing sounded like a breath of fresh air after getting through a few decades of crooner songs, western swing, and _god-awful_ novelty songs. Haley and Elvis and Cook [[when his blues singles finally appeared in popular charts) were an absolute relief compared to what came before. For jazz, you get off the hook a lot earlier, but the mid-century masters of jazz weren't selling numbers near the top-selling artists of the era, so they don't appear in the top 100 tracks I was listening to for each year. It was an interesting exercise, and gave me a lot of context.

YOu are very serious about your music history to subject yourself to the 100 most popular songs of each year from 1936 onward! Any horrors of the time that you still can't shake off? Especially some of those novelty songs or songs that abused double tracking production?

[delta]

YOu are very serious about your music history to subject yourself to the 100 most popular songs of each year from 1936 onward! Any horrors of the time that you still can't shake off? Especially some of those novelty songs or songs that abused double tracking production?I mis-typed - I meant 1946 onward, from when sales data starts becoming a bit more available and reliable, and the predecessor to what later became the Billboard Top 100 started [[which wasn't 100 songs long until about a decade later).

It's not specific songs that get stuck in my head, but certain kinds of songs which were popular in specific eras. The worst, for me, are novelty songs with made-up words or self-concsious real slang. Often, they're older artists either parodying or pandering to the language of a younger generation by exaggeration. It's about as bad and off-kilter as when people try to use rap to "speak to the kids", for example in the public service announcement propagandamercial "Don't Copy That Floppy".

Specific examples of what I'm talking about are:

1946, Tex Beneke & The Glenn Miller Orchestra “Hey! Ba-Ba-Re-Bop”
1946, Perry Como “Dig You Later [[A Hubba-Hubba-Hubba)”
1947, The Andrews Sisters & Danny Kaye “Civilization [[Bongo Bongo Bongo)”
1947, The Andrews Sisters “Toolie Oolie Doolie”
1951, Debbie Reynolds & Carleton Carpenter “Aba Daba Honeymoon”

etc. Really fatiguing stuff on several levels. Another phenomenon which crops up in 1949 on the pop charts and recurs more strongly in the "classic" oldies era is the song where the dance you're supposed to do with the song is the name of the song, and the lyrics of the song instruct you on how to dance the dance. Another thing which goes away for the most part by the time the rock era starts is what I refer to as "aw shucks country pop", where the song is about uneducated rural people and makes fun of the stereotype, either directly or by demonstration. Sadly it raises its head again in post 1990's "new country" pop. Then, there's a few outright racist moments, and "Open the Door, Richard" pops up a few times by different artists - it's so blatantly offensive to modern sensibility that I can't even imagine a black man singing it unless he was wearing blackface makeup, it's very nearly a minstrel piece. There's a few later pet peeves, such as The Steve Miller Band's "The Joker" which has always irritated me beyond belief, but on the whole, it's the pre-rock era that really hurt the most.