Definitions
Sampler
Many teenagers love sampler CDs - sets of 1 or 2 CDs, filled with lots of
"chartbreakers", e.g. "Bravo Hits", MTV something, "Best of the Best" or
how they are all called. Samplers give them the songs they want (because
they know them from the radio and like them), for a seemingly good price.
Unfortunately, many people are not aware of the gotchas with these CDs.
There are also samplers that carry popular songs from only one artist or group, but for the sake of this document, they count as albums.
Album
A CD with many different (usually 8-14) songs from one artist or group.
Maxi or single CD
"Maxi" and "single" are practically synonyms, the latter originating at vinyl times, I think.
A maxi is a CD for one song. It usually contains other tracks as well, most
often several other versions or remixes of the same song or sometimes one
or two other "B-side" songs. Playing time rarely exceeds 30 minutes.
Various reasons
There are many reasons to buy the albums or maxis of the artist and let samplers rot on the store shelf.
Artist compensation
I think that artists don't get much money from the sampler CDs. The
big music/record labels, which work all day to screw the rest of the world
(including artists, music lovers and even normal computer owners), cash almost
all of the profit for these CDs.
Artists earn the most with album CDs, although it's still only $0.5-$1 per CD!
Work as a whole
Some (not many) albums are a work as a whole. They have a certain theme and sometimes even evolve throughout the CD.
Good unknown songs
On good album CDs, the "rest" of
the songs (those which you don't know yet) are not "fillers", as many people like to call them. Many of the
songs are just as good as the "hits" or "chart breakers" which you can hear
on the radio. I found many songs that I really enjoy on some albums, although
they probably never hit the radio.
The difference between a good song and a "chart breaker" is that the latter
is heavily "promoted", i.e. the artist pays lots of money to e.g. the radio
station, so that they play the song n times per day and you know about the
song at all.
Similarily, on a good Maxi, the remixes or other versions are just
as good and enjoyable as the radio version. In fact, some good versions would be unsuitable for the radio.
Sound quality
Another important reason to not buy Samplers, and the reason for this webpage,
is the sound quality. I wondered why some of the CDs sounded so terrible,
almost worse than on the radio. Then I compared a song from a sampler with
the same song from an album or maxi. The difference is utterly obvious, and
you can even see it on the computer.
Graphs or "Cutted"
If you load the song into a normal music (WAV) editor, it will display a
visual graph for the song. Here is how the 2 versions of the song look like:
This is the same song, from 2 different CDs. Both files are MP3 files from
Weblisten.com, loaded as WAV into audacity. But you can observe the same
with your own, uncompressed CD rips and e.g. CoolEdit.
On the left: File 2. Here, you can see how a song from an album or maxi usually looks like.
The graph has fringes. This is normal and good, it represents part of the
character of the song.
On the right: File 1. Here, you see how a song on a sampler usually looks like.
The graph has, in this view, almost no character. If you zoom in, you will
see more differences between songs from samplers, but the fact remains that
many peaks hitted and exceeded the maximum of what the WAV/MP3 file can represent
(100%) and are thus cutted. In other words, part of the sound of the song
is gone. The information is lost.
Listening tests or "flat"
If you have good loudspeakers connected to your computer, and a somewhat
reasonable equipment (soundcard, amplifier), you can make the test yourself.
It is so obvious that even untrained ears should easily be able to tell the
vast difference.
Below, I have 35 seconds of the song "Madonna - Open your heart", from 2 different
CDs and additionally in different encodings. For the sake of testing Samplers,
you only need to care about the files 1 and 2.
(See remarks below)
Sampler Test
- Download files 1 and 2.
- Listen to the file 2, ideally 3 times
- Listen to the first 15 seconds of file 2 again.
- Immediately after that, listen to file 1 from the beginning.
That hurts, doesn't it? :-)
Encoding test
(This is not relevant to Samplers, just to sound quality on the computer. Feel free to skip.)
Compare files 2, 3 and 4. You will probably find differences between 2 and
3: The spacial (stereo) resolution is extremely poor. Various details (which
are part of the fun with music) that can be clearly heard and located with
my own MP3 are not only poorly resolved, but heavily distorted and sometimes
even on the wrong side!
A quick EncSpot check suggests that weblisten.com used the Fraunhofer encoder. That encoder
is reported to have problems with Joint Stereo (used for 128k MP3). From
my tests, this sounds like an understatement. :-(
I find it hard to hear differences between file 3 and 4, i.e. between a relatively good MP3 encoding and the original.
Nevertheless, I don't want the hassle of re-ripping each time I do notice problems, so I just keep all my CDs flac-encoded
on my harddisks. Flac is lossless: after decoding, you get exactly (bit-for-bit)
the same file that you once encoded. Its compression ratio varies between
50% and 70% (usually more towards 70%), depending on the song. 400MB for
an album CD are fair IMO, given that you can buy 80GB disks (storing 200
full CDs) for 100 US-$ these days (Aug 2002).
Some detail remarks to the listening test (feel free to skip):
I tried to maintain the highest quality of the files, e.g. ripping from CD with cdparanoia and cutting the MP3 files to 35 seconds with mpgsplit.
Please forgive me that in this case, this is not actually a song from a sampler vs. a song from an album. In fact, CD A is True Blue (I think) and CD B is The Immaculate Collection. I chose them, because I had them handy and because they illustrate the problem which I saw (in the past) with samplers vs. albums.
Final words
Intentional
The bad sound quality of samplers is not an accident or unfortunate side
effect, it is deliberate. The people, who are creating the samplers, intentionally
do it this way, because that makes the song louder and it thus sounds better
at the first glance to most people. But as soon as you listen a few seconds,
it sounds terrible. What's wrong with the volume control on my amplifier?
So, I guess the real reason is that the music labels intentionally drive
down the sound quality. Of course, nobody (from them) tells you that. Talk
about malice!
BTW: The radio stations usually do the same, enabling "bass boosters" and similar crap.
Results
Unfortunately, many people never find out about the sound quality of some
good songs. On these sampler CDs, very sophisticated songs/artists sound
the same as songs/groups ala Backstreet Boys.
Sampler CDs are like radio on demand. But way too expensive for that. You
can get that on the Internet cheaper (review of legal mp3 sites to follow).
Even 128k MP3 files can have way higher quality than sampler CDs.