Your figures are spot on Bert. My employer used to use a Polycom ISDN videocoference system that had 2 versions - one used 384 K (6 ISDN channels) and could conference 5 sites on a 30-channel Primary rate connection, and the other used 128k - 2 channels or one basic rate connection (these are Euro standards, not US). Picture and sound quality were both good, especially on the 384K links.
We now use a networked system which uses a lot more network bandwidth - though the sound and pic quality don't seem a lot better than the old Polycoms....
My god. Why don't we spend time actually making the tech we already "have" work? I have yet to experience a video-conferencing system that worked even remotely close to expecations. Rather, we swear and curse trying to get it set up, swear and curse again when it drops out, or the audio garbles up or starts echoing, or the video freezes, or who knows what else. I think too many of these systems are designed in a "lab" environment with extremely high speed, low latency, and robust connections, and ignore the real world that is nothing like that.
PS: Web-site developers suffer from the same mis-conception. SURE your page is all flashy and cool...but not if it is too bloated and slow to run on the connections that many of us have.
True, David. ISDN limited itself to 8 KHz sampling frequency, like just about all digital telephony standards, so the audio limit was 4 KHz. But the reason I mentioned Skype was, why the slavish adherence to that inadequate standard?
When videoconferencing can be made to use broadband Internet links, and not circuit-switched telephone links, the equipment in the use premises should allow for decent audio quality.
And it's not like you need to send that audio uncompressed, like in a CD. DVD movies have very decent sound quality, it uses perceptual compression, and only takes up maybe 384 Kb/s of transfer bandwidth.
Sure, cell phones use maybe 9.6 or 14.4 Kb/s for voice. But a videoconferencing system that will dedicate maybe 4 Mb/s or more, to H.264 compressed HD video, can certainly afford to dedicate 384 Kb/s to the audio? (Maybe they do offer decent audio, and it simply was not covered in the article.)
Good point Bert, when you can stream CD-quality audio with even a bad broadband connection, you wonder why someone doesn't update the system. I guess the answer is standards - the 4 KHz phone bandwidth standard has hung around longer than most.
It's a pity they didn't up it to 10KHz when ISDN came in, but I guess bandwidth was still too valuable then to waste it.
Back in the 80-ies when I was a student I already worked on video conferencing. I never understood the attraction. A good audio link with screen sharing is much more useful, cheap and available many years.
A tip for those who like video, hook up a couple of skype or other cameras and a couple of scrrens and you can get widescreen video.
Very cool, IMO. I've had fun creating panormaic photos by attaching a camera to a tripod, then taking a series of pictures, swiveling the camera between each pic. The results are super. This combination camera does the same thing automatically.
I agree that a better view and better video resolution is sorely needed to make videoconferences better, but you know what? Always forgotten in these videoconference improvement discussions, and I mean always, is a discussion about the audio. I would venture that what is needed even more than the better images is to get beyond the 4000 Hz telephone quality audio in just about every teleconference system I've used. Heck, if nothing else, Skype sessions, in which the audio is better than telephone quality, ought to give a hint, no?
Teleconferences would be immeasurably better if the audio portion would include at least up to 10 KHz acoustic frequencies. IMO, this is even more important that the panoramic view, to give the impression of "being there." And if you're carrying HD video anyway, surely something at least aproaching hifi sound won't take up a lot of extra bandwidth?
I'll grant that people who limit themselves to smartphones and tablets won't notice the difference. Then again, I don't see how anything approaching a "being there" experience can possibly be achieved when videoconferencing with the handhelds anyway.
All you have to do, to witness the huge difference audio makes, is to listen to a typical call-in show on a decent sound system. The guests in the studio, or even guests in remote studios linked into the program with high capacity feeds, sound like they're right there with you. The ones calling in on the phone sound pathetically bad, sometimes even hard to understand. Unfortunately, all too often, the audio quality achievable with the ancient 4 KHz analog telephone standard is still being used as the reference standard. And digital audio made to emulate that primitive quality level.
Blog Doing Math in FPGAs Tom Burke 15 comments For a recent project, I explored doing "real" (that is, non-integer) math on a Spartan 3 FPGA. FPGAs, by their nature, do integer math. That is, there's no floating-point ...