"However, Intel Widi uses 802.11n and 802.11n can support upto ~600Mb/s (quad-band, 40Mhz)."
That's absolutely false information. The best consumer grade equipment might reach 90 Mb/s under IDEAL conditions. Most of the time you can divide promised speed by 6, and you get close to real speeds with ideal conditions.
Btw, 40 MHz is just not happening with compliant chipsets in 2.4 GHz spectrum. Too much interference, and compliant chips back off. 5 GHz might be - that's when you get that 90 Mbit/s. But at 5 GHz walls block a lot more, the range is not great.
it's not just about hd. the most common use i think will be sharing lower def content on a common screen (tv), like when you sitting around with friends and want them to check out something cool you saw on youtube. or quickly pull up the most recent news or weather video on the internet.
You are correct that 802.11g would have a hard time with 1080p broadcasts - even the crapiest quality 1080p broadcast uses ~25Mb/s. It "might" be possible under great conditions, but I'd suspect the transmission would drop in and out quite often.
However, Intel Widi uses 802.11n and 802.11n can support upto ~600Mb/s (quad-band, 40Mhz).
H.264 using HiP profile (the standard HD digital broadcast format, also used for HD Blu-Ray) "only" use ~300Mb/s at its highest level - definitely usable and more realistic for what is suggested in this article and the video included.
Even compressed 1080p streams consume over 20 Mbps. Max. throughput for 54 Mbps 11g wireless is 24 Mbps, assuming perfect signal and no one else using the same channel.
While 11n might achieve almost 90 Mbps under ideal conditions, it's often no better than 11g. I've seen less than 5 Mbps in noisy environments. Worse, it might be 0 Mbps for a second or two. To cover this, it'd need to buffer for a few seconds, which means equally many seconds of latency.
The only way around is to aggressively recompress the signal with little consideration of quality. I bet this is what Intel is doing. Sure you can get 1080p down to 5-10 Mbps, but it'll look like shit. Assuming 30fps video, you need 25 Mbps to get a good quality and much, much more if you have realtime requirements (say, encoder that adds less than 500 ms of latency), because you can't predict as much from "future" I or P frames. No backwards prediction is possible when you need sub-33ms encoding latency.
"However, Intel Widi uses 802.11n and 802.11n can support upto ~600Mb/s (quad-band, 40Mhz)."
That's absolutely false information. The best consumer grade equipment might reach 90 Mb/s under IDEAL conditions. Most of the time you can divide promised speed by 6, and you get close to real speeds with ideal conditions.
Btw, 40 MHz is just not happening with compliant chipsets in 2.4 GHz spectrum. Too much interference, and compliant chips back off. 5 GHz might be - that's when you get that 90 Mbit/s. But at 5 GHz walls block a lot more, the range is not great.
it's not just about hd. the most common use i think will be sharing lower def content on a common screen (tv), like when you sitting around with friends and want them to check out something cool you saw on youtube. or quickly pull up the most recent news or weather video on the internet.
You are correct that 802.11g would have a hard time with 1080p broadcasts - even the crapiest quality 1080p broadcast uses ~25Mb/s. It "might" be possible under great conditions, but I'd suspect the transmission would drop in and out quite often.
However, Intel Widi uses 802.11n and 802.11n can support upto ~600Mb/s (quad-band, 40Mhz).
H.264 using HiP profile (the standard HD digital broadcast format, also used for HD Blu-Ray) "only" use ~300Mb/s at its highest level - definitely usable and more realistic for what is suggested in this article and the video included.
N
Yeah, this is going to happen. Not.
Even compressed 1080p streams consume over 20 Mbps. Max. throughput for 54 Mbps 11g wireless is 24 Mbps, assuming perfect signal and no one else using the same channel.
While 11n might achieve almost 90 Mbps under ideal conditions, it's often no better than 11g. I've seen less than 5 Mbps in noisy environments. Worse, it might be 0 Mbps for a second or two. To cover this, it'd need to buffer for a few seconds, which means equally many seconds of latency.
The only way around is to aggressively recompress the signal with little consideration of quality. I bet this is what Intel is doing. Sure you can get 1080p down to 5-10 Mbps, but it'll look like shit. Assuming 30fps video, you need 25 Mbps to get a good quality and much, much more if you have realtime requirements (say, encoder that adds less than 500 ms of latency), because you can't predict as much from "future" I or P frames. No backwards prediction is possible when you need sub-33ms encoding latency.