I have Comcast home internet in the bay area (Speedtest is around 30mbit down/8 mbit up) and every time I try to watch anything on Netflix it buffers for a minute and the resolution usually stays unacceptably low (320p?!) while usually having to buffer again every 2-5 minutes, sometimes for the whole run time of the show. Meanwhile, Comcast's own XFINITY streaming service delivers instantly (no buffering) at full HD resolution.
I actually never understood what the whole net neutrality discussion is about, as it's very obvious that we're already getting screwed over by ISPs.
Contrariwise, I have Sonic's "Fusion" bonded ADSL in SF. My modem and Speedtest.net report around 12/2 Mbits down/up, and both Netflix and Amazon Prime consistently stream flawless HD for me.
I've had the exact same experience in Houston with Comcast + Netflix at slightly faster speeds. Eventually I dropped Netflix because it was unusable everywhere except my phone. Oddly I don't have the same issues with Amazon Prime, YouTube, or MLB/MiLB streaming. Given that Prime is also a direct competitor to Comcast streaming I'm not entirely convinced this experience with Netflix is the result of some deliberate plot on the part of Comcast versus some other bizarre issue.
That isn't to say I disagree with network neutrality - I think it's vital. I'm just not going to attribute to malice what is quite possibly the result of incompetence.
You'd be wrong. About 2 years ago, Netflix switched to using Level3 as their primary ISP. The peering between Comcast and Level3 became saturated, and Comcast declined to increase bandwidth at the peering point without being paid, and also declined to install netflix's CDN boxes somewhere on their network without being paid. The question of "who pays who" is kinda complicated between ISPs, payments are rather high, and comcast managed to reverse it's position/relationships in that space, because they have a monopoly on the users in their territory. They also hate competition from netflix more than from the others you mention, it's a more complete replacement for their video offerings.
They also hate competition from netflix more than from the others you mention, it's a more complete replacement for their video offerings.
So they've never actually browsed Netflix's offering then, huh? Because it's hardly stellar or offers much in the way big-ticket items that compete with anything Comcast shows in their on-demand or via cable channels.
I wonder if some form of traffic obfuscation would help in your endeavor? For instance, setup OpenVPN on a VPS or something, and tunnel to connect to Netflix. It would be a shame if you end up with better quality.
I've had streaming issues with both Netflix and Youtube while using Verizon FiOS in the past. I then connected to my University's VPN and instantly enjoyed smooth-as-butter streaming on both sites.
There's a good chance you'll have the same experience with a VPS, depending on what networks the VPS's traffic routes through.
The problem is that traffic from you to Netflix travels through several networks, and at each link between networks it will face varying amounts of congestion depending on the size of that link. There is a link somewhere between $BAD_ISP and Netflix that, due to political/business reasons[0], hasn't been upgraded to accommodate more traffic.
By connecting through a VPN, you may be able to circumvent this as long as the route from you-->VPN-->netflix doesn't travel across the problematic peering point.
[0] Screwing you over; Promoting the ISP's own competing services (TV, for example); Attempting to extract payment from Netflix since its "their traffic"[1], despite the fact that you are already paying them for this connection.
In the bay area, do look at sonic.net; there are real limitations to their last mile connectivity (ADSL line shared including in Remote Terminals, or ADSL2 from Central Offices, and a couple small Fiber installs), but they have very good connectivity once you get in.
I could rant about network neutrality... Comcast may not be (and probably isn't) doing anything to degrade Netflix's ability to stream to you; but they're probably also not making as many upgrades to peering or transit capacity as they could for that traffic.
It's hard to tell how often this is getting messed up by content delivery partners in the middle, versus intentional traffic shaping, but it sucks either way.
I have Verizon Fios. When I go to watch a Youtube video on my Android, I often notice the video won't load. I then turn WiFi off, and the video streams flawlessly over 4G. I don't know when Net Neutrality died, but it did.
I'm in exactly the same boat. I'm >14000 wire feet from the xo, so dsl barely works, either. Net neutrality is needed as long as consumers don't have real, competitive choices.
I have a similar issue using Brighthouse in Florida. For the longest time I was able to stream HD quality from netflix through my roku, until recently. I always suspect that my ISP is somehow the cause, the only problem is I can't imagine how I would prove it, or what I could do even if I could prove it.
The fine print: "These ratings reflect the average performance of all Netflix streams on each ISPs network from Nov. 2012 through Sept. 2013 and average performance during prime time starting in Oct. 2013. The average is well below the peak performance due to many factors including the variety of encodes we use to deliver the TV shows and movies we carry as well as home Wi-Fi and the variety of devices our members use. Those factors cancel out when comparing across ISPs, so these relative rankings are a good indicator of the consistent performance typically experienced across all users on an ISP network. - See more at: http://ispspeedindex.netflix.com/usa#sthash.ufsSY1Bi.dpuf"
Right! I even entered /canada manually but it seems like they just don't measure it. Though humiliation means nothing to the Big Three but it is still fun to watch how horrible our internet providers are.
I actually have an amazing ISP (start.ca, uses Rogers' cable network, but with their own peering infrastructure behind it)
I always hear about people saying "I can't play YouTube videos!" or like the comments above this "Netflix keeps buffering!". I've had absolutely zero issues like that.
But, I'm starting to think that people's issues with their home internet comes down to their network setup. I have the fast Asus RT-N66u with Tomato firmware, good QoS rules, Apple TV wired and not wireless, etc.
Maybe some of these folks have the crappy gateway from the ISP, or really bad wireless environments, or bad cabling.
Case in point, I was at a person's house the other day who wasn't even aware that their DSL connection from Bell was disconnecting every 5-10 mins for 30secs or so (probably due to bad phone lines in the old building).
But they went on watching Netflix and checking their email totally unaware. So Netflix drops it quality or buffers a little then continues on it's way and people go "Oh man, my ISP sucks! Netflix is always buffering!"
I bet there's a good business in going around to people's homes to fix network issues.
Yeah. I agree. I was with start.ca Rogers cable on the 150/10 plan and at that point you really need gigabit or 802.11ac to feel the speed of your home internet. Luckily I've both, and switched back to Rogers to have a go with their 250/20 plan. I sometimes feel like I'm the only one on HN satisfied with my ISP. Even if I'm paying too much for it ($125.99/mth), consistent downloads of 270-320mbps and an uplink of 20mbps with a terabyte bandwidth cap is nothing to sneeze at. I suppose I've Teksavvy/start to thank for it. I still highly recommend start.ca for those that aren't as insane about speed as I am :) But don't get me started on people who use the ISP's free email address and so cannot switch...
I don't see that this tells us anything of interest at all. It's certainly not saying how fast a given ISP can go if you pay for it. I think it's largely reflecting a combination of which ISPs have a higher-proportion of subscribers on cheap plans, and who has users that are pulling content through marginal WiFi connections.
Would be more useful if it controlled for what plans people are on.
You really need a scatterplot here, or to look for clusters at certain speeds. Say you're looking around 15 Mbps. At that rate, which company has all the subscribers coming in at 16, and which ones at 13 or 14?
Since there are a relatively small number of speed options, it should be possible to notice these sorts of clusters in the data.
Also, how sensitive is a certain company's speeds to time of day? Whose average dips the most during peak viewing times?
It seems like even a little data analysis would go a long way here...
Remember this is the speed (in theory) that you get from Netflix on that ISP. Netflix is not measuring your total ISP bandwidth. They can only measure how much bandwidth the ISP provides you for Netflix.
The list of speedtest.net servers is public and I don't doubt that Comcast uses it to drive traffic shaping decisions. Speed tests are meaningless except to establish an upper bound on last mile speed. They are not representative of bandwidth to real services like Netflix, YouTube, Dropbox, Steam, etc.
I can't argue with that. I just checked actual scp speed to and from ec2 and I'm getting slower downloads then uploads right now. At 7PM my time. 6.3mbps up, 4.5mbps down.
speedtest.net shows those exact speeds now as well. That's nice, 80% of bandwidth going away in a few hours :-)
Doesn't the 20-25Mbps speedtest focus on "speed boost" where Comcast gooses the speeds for initial connections (so web page loads are faster, but sustained downloads are slower)?
> the average performance is well below the peak performance due to many factors including the variety of encodes Netflix uses to deliver the TV shows and movies as well as the variety of devices members use and home network conditions. These factors cancel out when comparing across ISPs.
Speedtest results are completely irrelevant because Netflix uses Level 3 as CDN and bandwidth between Level 3 and Comcast is always 100% utilized at prime time.
While others have correctly pointed out that 25 Mbits/s does in fact equal 25 Mb/s, (though, I always use Mbits/s, just to eliminate the confusion) - I'd also like to note that bandwidth is always measured in SI Mbits - I.E. 25 Mbits/s of bandwidth is always 25,000,000 bits/second.
Any time you see someone call it out explicitly as 25 Mib/s, double check and see if that's what they actually mean - 90+% of the time it isn't.
We have Cox in my state, and I have been incredibly impressed with how reliable they are for Internet. It's refreshing to just take my connectivity for granted and not even remember the last time I've had issues or slowdowns.
The only real downside is that when I recommend Cox to friends, I have to be very careful how I say it.
After living in a Comcast stronghold for > a decade, I was flabbergasted at how ... decent Cox is, in terms of both quality and customer service. I'm still waiting with baited breath for them to realize that they can get away with more and pull one on me.
They have Norway, Sweden, Denmark, Ireland, and Finland listed on that site but not Canada? All of which have a combined population smaller than that of Canada.
It looks like this is just measuring the average bitrate that people are streaming at. It won't reflect whether or not someone is capable of watching higher quality streams than Netflix is actually giving them. I don't think their highest quality is more than 6 megabits/sec, and it's only available on certain devices and movies.
Forgive me if i am not reading this correctly, but surely we need to know what bandwidth netflix is streaming at?
For example Virgin in the UK offers 20mbps+ on their fibre network for most customers, so this is telling me that netflix is able to stream at a maximum bitrate (of assumably 3mbps)?
What this is not telling me is that Virgin's average network speed.
Out here in the flyover states, complaining about Frontier's terrible service is right up there with deer hunting and watching old episodes of "Hee Haw".
Now we see AT&T and Verizon are doing even worse than Frontier and somehow a cable provider is getting flattened by DSL. What's up?
How is ClearWire considered "fixed"? Anyway it's not called that anymore.
I guess this info might be useful for public policy, but there's so much variability within a single provider that you wouldn't want to choose an ISP based on this alone.
I just assumed that "fixed" was in contrast to "mobile". Anyway it seems a bit too generous to the actual "fixed" providers to compare their performance with a WISP.
My guess is because latency is not relevant when it comes to streaming videos from Netflix. A consumer will not be able to notice their video took 40 milliseconds longer to start streaming. They will, however, notice when the video stops to buffer every 10-15 minutes because their ISP can't keep up with the throughput.
Latency matters depending on the video player. At 1080p I find FLV streaming services with encryption tend to fall over unless your Internet connection is silly fast, including latency. Any delay, due to poor protocols or poor playback software, will cause the video to stop and rebuffer. Then again, perhaps the video is coming from sufficiently far away that in a live stream, I'd be right on the borderline of watchable if my ping on the wider internet spikes too much. Eh, what can I say. Every time I can boost the speed tier on my cable line, it seems to slightly improve ping and dramatically improve the reliability of my connection. Perhaps the longer things are transferring on the internet the more likely I'll see variability during the transfer?
of course. but not greatly. Generally (depending on the client implementation). The actual download of the data (and the images) far eclipses latency issues.
30ms vs 160ms delay is irrelevant when it takes you 3 seconds to download all the image tiles :D
I actually never understood what the whole net neutrality discussion is about, as it's very obvious that we're already getting screwed over by ISPs.