bodosom wrote:
jimpal wrote:
I think we both may be right. My experience clearly shows the problem begins with Netflix overload, and that's my opinion, but it may be that ATV and iPad apps are more vulnerable to that overload than some other products.
This supposition doesn't explain why multiple people have found an alternate device (PS3, TiVo etc.) can consistently get 5800 while the ATV3 can only get anywhere from 235-2350. These tests have been run at the same time and in my case I only use results where it's the same time and using the same content distribution IP address. All of my media devices that can be wired are wired to the same gigabit switch.
I.e. I've held everything I can control constant across platforms and the ATV consistently gets < 3000 (typically < 2000) while the other devices always get 5800 on the test and report superHD on all superHD content
Unless you're defining Netflix overload to mean any time an ATV3 gets less than superHD.
Actually I think my supposition does explain it. If you read it again, in a nutshell I said that it seems that at a period of high use, the Apple software may not handle the diminished stream from Netflix well. That would explain why I got 235 kbps on Netflix's Example Short last evening, and 5800 kbps this morning (clearly a period of low use). Obviously Netflix high usage is a factor. I think there may be a problem in both houses.
I just called Netflix. While very polite and helpful, the rep told me the problem can't be Netflix at all, but must be my internet supply. (It measures 70 Mbps on Speedtest reliably.) He said "it's impossible for Netflix to bog down at times of high usage because of their new Cloud Server system." And he told me to (guess what?) unplug and replug my modem and router. I really don't believe what he said, but at least he was nice about it.
I define Netflix overload whenever the picture gets crummy, such as 235 kbps last evening.
So, I will stick with my statement above. Thanks for your helpful input to all of us who are trying to figure this out.