I don't think it is as clear cut as it seems. When you figure in everything that goes into making a CD that is not necessary for making a digital stream (the case, booklet, physical CD, packaging on the CD, shipping of the CD to the retailer, the energy cost associated with keeping the CD in storage or on the retail floor, transporting the CD home, and even the cost associated with spinning the disk drive, etc.). Basically, a lot goes into a CD. Streaming is much simpler. The files get sent to the online service and from there are streamed. The only clear environmental costs associated are the costs associated with the data centers, the costs associated with the creation of the software and necessary operating procedures, and the device a person listens on. Unlike CDs, transportation is free (environmentally) when it comes to streaming. The internet networking infastructure is already in place, consuming the same amount of energy regardless of what passes over it. Physcial trucks do not operate that way.

Technically, the two products being offered are streams and CDs. The costs associated with the making of the products is the cost of running the servers (which make the streams) and the cost of making the CDs and their necessary cases, etc. Figuring out the costs associated with making the servers (not running them) would mean you have to figure out the costs of making the equipment that makes CDs. That is going to be nearly impossible to calculate.

Now data centers consume a lot of energy. It is not just the servers and networking devices, but the cooling is also a very large factor. It is possible that the environmental cost for streaming is greater than CDs. But it won't stay that way. Servers are getting more energy efficient on a continuous basis, networking devices are moving more data per watt, and virtualization is reducing the amount of servers needed. Streaming will only get more efficient. CDs are probably not going to get more efficient in the near future and probably never as the amount getting produced continues to drop.

"Streaming or downloading 12 tracks, without compression, just 27 times by one user." Highly possible. But songs are not streamed without compression. And no one is going to download tracks 27 times. Granted you technically download a track every time you stream a track, but an actual file download from iTunes only needs to be done once, with compression, making is much more efficient than a CD. Streaming songs with compression could result in more energy usage than a CD, but being as the music is compressed it would be much more than 27 times. And as I was very clearly made aware of in a different debate, people rarely listen to a track that many times.

Just my take.

Casey, 

On your last point, the report asserts that compression could be a temporary solution to current bandwidth restrictions.  And, non-compressed delivery could become a reality in the future.

"While there’s a huge difference between energy consumed by one streamed MP3 (0.16 Wh) and that of an uncompressed track (1.21 Wh), advances in network infrastructure speeds in excess of 100 mbps (wireless) mean that audio files would no longer need to be compressed in future."

Actually, Jimmy Iovine and other audiophiles are among those that would like to see this reality.  

I see something even bigger: audio-only itself is a function of bandwidth restrictions, but this could eventually be replaced with far broader media packages that include video, more metadata, etc., to be accessed if the listener decides.

/paul

Streaming in the future might also include more than 2 channels. I was never tempted by the surround sound releases, but now that I can stream Spotify through my Onkyo home theater receiver, I would pay a premium for 5.1 channels, which would be ~2.5 x 1.21Wh per track without compression.

What's the purpose of streaming 2 channel audio in 5.1?

http://en.wikipedia.org/wiki/5.1_Music_Disc
http://en.wikipedia.org/wiki/Super_Audio_CD
http://en.wikipedia.org/wiki/DVD-Audio

Spotify's cache defaults to 10% of your hard drive. My hard drive is 320gb, so that's ~3,200mb of Spotify storage. Supposing a song is 10mb ea (compressed), that's 320 songs stored locally, or ~26.5 albums. Supposing I don't have more than 26.5 albums in rotation, those songs are being pulled locally. At least that's my understanding of it. And the cache can be increased to 10gb. So this idea that 324 (27x12) non-compressed streams (~19 hours at 3.5 mins/song) equals one manufactured cd is misleading because we're still in the compression era, & a "stream" isn't necessarily a stream. It's an extreme oversimplification.

The comparison is made tougher because you're comparing consumer habits to manufacturing habits here. Plenty more CDs are manufactured than sold, & all of the overrun has to be factored in.

Additionally, the ever increasing affordability of recording equipment drives the creation of more content than there is demand for, & the number of projects released grows by leaps & bounds each year. This is offset somewhat by online-only releases. Albums that are released & never stream have a negligible impact on the environment, when these albums would've otherwise been pressed, even if only in runs of a thousand or so... thousands of runs of a thousand really add up.

great...

Learn something new EVERY day :-) Never thought the beauty of music could potentially harm the envirionment.

Paul,

You state -

Quantifying the problem remains a difficult problem itself.  In the physical realm, measurement is easier: x-number of CDs, with a certain percentage ending up in landfills, etc.  But in the digital realm, complexities balloon to include plants, hardware, electricity consumption, different types of media, and even energy costs related to creating all this content.  That is a question for a lot more research, though the answer could be a depressing one for environmentalists.

I have a problem with this argument.  It seems that you are comparing a portion of a costs associated with physical to anything and everything involed with digital.  If you are going to incorporate the waste created by i Pods and smart phones, you need to take in to account walkmans and home stereos of yesteryear.  There is also the fuel cost of shipping all that plastice across the country and world.  There are other things as well but i believe this articulates my point.  I'm not trying to hate on you, I just want a fair comparison is all.

Jared Kuenstler

CEO Nasty Recordings

NastyRecordings.com

I don't think the hardware player to play streamed music should be included for a fair comparison.

Because hardware exists for old formats too.

The difference is probably that chinese workers were not killed by explosions of turntable factories, contrary to some high end tablet we all heard about.

What could be included, is the cost of electro-sensibility for wireless devices. But the number of people truely sensitive to wifi radiations is too low and not taken seriously. May be in a few decades when western science will admit that electromagnetism can cause health damages (like it was found in other, not western cell phone producing countries).

nb:

wifi: understand, wifi and any type of wireless connection, microwave or not....

at what frequency do you produce pop-corn again ?

That was kind of my point.  Maybe I didn't express it well enough, but I'm glad we agree. 

J Red

i agree with the difficulty to make an accurate global picture....

I never heard of a progress in any domain that does not imply a step backward of any kind.

data can be stored in DNA cells.

All the information on the net today can be encoded in a DNA abled device that is smaller that your smallest finger nail.

http://www.extremetech.com/extreme/134672-harvard-cracks-dna-storage-crams-700-terabytes-of-data-into-a-single-gram

The question is "in how long" and whether this time lapse is shorter/longer than what is takes for vynil to disappear in nature (hundreds of years) ?

I disagree.