July 04, 2006
Ubiquitous VoIP Communications - Audio File Storage Issues
Assuming that VoIP communication becomes ubiquitous, with or without SEDs (Service-Enabled Devices), recordings of VoIP communications are likely to become compulsory for businesses, for a variety of reasons. Email communications are currently archived, and it's likely VoIP conversations will be, too. The result is a need for ever larger storage devices.
Consider for a second that a one-minute recording in .WAV format at CD stereo quality (44.1 Kilohertz = 44,100 cycles per second) takes about 10.1 Megabytes. One employee that communicates by voice for at least half a regular work day thus spends 4 hrs/d x 60 min/h = 240 min/d, give or take. In a 5-day work week, that's 1,020 minutes x 10 Mb = 10 Gb/wk = 40 Gb/month. Five employees doing the same thing amounts to 200 Gb/month. [Note: MP3 format at the same quality takes exactly the same amount of space.]
Of course, not every company is going to have the same amount of VoIP audio storage requirements. But any company that intends to use VoIP for live support - say a call center or an online sales department - is going to need much more than that.
Say a call center operates 24 hrs/d and has 50 employees on at all times, 7 days a week. Their storage needs (at 44.1 Khz WAV) amount to 50 e x 24 h/d x 60 min/h x 7 d/wk x 10 Mb/min = 50 e x 10,080 min/wk x 10 Mb/min = 5,040,000 Mb/wk = 5.04 Terabytes/ wk, or about 20 Tb/mth = 1/4 Petabytes per year.
You can see that if VoIP recordings need to be kept "live" for several months, there is a considerable requirement in storage space. While archival could still be done on CDs/ DVDs, space is still an issue, as more companies adopt VoIP for applications. Some businesses may actually need to keep recordings live, instead of archiving them.
Not only will a demand for high-capacity SANs (Storage Area Networks) increase (consider buy company shares :), but so will the demand for more efficient audio file compression techniques.
I won't get into it much here, but one such technique is Wavelet Compression, which is suitable for some audio and video recordings. Wavelet theory was devised by quantum physicists in the early 1970s, and but was only applied recently to file compression techniques, sometimes resulting in space savings of up to 75%. Wavelet compression outperforms the slightly older fractal compression techniques.
Both techniques are "lossy"; that is, a slight loss in quality is sacrificed for the larger compression ratios. Unfortunately, not all audio and video compresses well with this method. But with ubiquitous VoIP, and possibly ubiquitous Video over IP, on the way, optimum compression techniques and massive storage units are going to be in high demand.
--
Did you enjoy this post?
« Ubiquitous VoIP Communication and SED Service-Enabled Devices | Main | Vocera and the Road to VoIP Application Ubiquity »
