CA3LE

So the SB6183 (16/4 vs the SB6121 4/4 channel) increased my speed but not totally. Ramp up to full speed takes a second or two when before it was nearly instant. Note how the middle variance went up and TiP minimum is much lower. Hurts the overall score. Above is actually an exceptional result right now... below is more typical... Sure... it ends up faster but because of the weak start it scores worse on TMN's linear test. Multithread and Mercury tests it performs totally different all together. Here's the best I've seen... (using sites.google.com Mercury Speed Test) This is not typical... I watched it burst when I got this result. My network controllers confirm bursts of up to 33.2 MB/s (266 Mbps) but once engaged transfers are chugging along at an average of 23-24 MB/s (~185 Mbps) Same Mercury test a moment later that didn't burst... (in getting this screenshot it sustained the faster speed for the longest I've seen... usually just see short bursts, like at the beginning of the graph above.) Linear results are showing me that there's room for improvement, the other modem initiated quickly... why can't this one? TMN didn't change. Nothing else changed... Looking at the levels.. Thanks again tjritter79 for letting me know about that modem. I'll get it dialed in. If this modem initiated as fast as the other one I'd expect to see 170-185 Mbps consistently with the linear (single thread) test. Again, this is why linear is the default. Multithread is important too but can make things appear better than they really are. The classic test is still the true benchmark to beat. IMHO these results clearly illustrate that. Downloading files off usenet (Giganews, 50 threads) Much faster than before.

You'll still be able to test with Google's servers just not the exact same way. You'll have to use the Mercury Speed Test system. Here's one that works awesome for me all the time. lh3.googleusercontent.com Mercury Speed Test Result I just ran on my laptop... Using that same test on one of my commercial connections I was able to see nearly 600 Mbps, again... confirmed by the client network controller. Google has very fast bandwidth and very quick responding servers, ideal for this test. You can search the Mercury index for other Google Speed Tests or make your own to nearly any web server you want to test against. Not many people realize that TMN does this. Find one that works good for you and bookmark it. If you like the Google tests, go to the Google Speed Test page, scroll down and click "Custom Index"... then when you visit testmy.net/mercury it will show your favorites. You can custom index by city, state, country or ISP (that the test is hosted on). Happy Testing!

I didn't realize that 16 channel modems were out, I'm going to pick one up right now! http://surfboard.com/products/sb6183/ SICK! A Comcast sales person came to my door a few weeks ago and was trying to get us to upgrade to the 150 package. I was like, "what?!" I thought I was paying for the fastest? --- wanted me to resign a contract. I shut the door. --- looking at the website... WTF! Been putting off the call to Comcast to see what's up but just now called and the girl tells me that the website is totally misleading and that I already have the Extreme 150. The sales person who came to my door was a total scammer! Just wanted to get me back in contract. Mother f&*%&*! So NOW... I'm pretty sure that if I swap for this 16 channel down modem it will bring the speed over 150 Mbps. My current modem is 4 down 4 up. Thanks for cluing me in.

You should always purchase your own modem. Don't buy it from your provider either, don't even let them give you one. Always do your research on the modem, even if you're not going to use it's full capability... in the future you may. Motorola owns. If you get the best model (usually for around $100) and pair it with a great N-600+ or AC router (Netgear owns here IMO) you can't go wrong. So side from saving money why should you always purchase your own? Especially with Comcast... Because Comcast USES their customers who have leased modems. If you have one of Comcast's modem/router combos you're not only broadcasting your own private wifi but also Comcast's public wifi network. People can hop on your modem and draw your bandwidth. They don't access the Internet from your IP but they do draw bandwidth from your modem. You can now call 1-800-XFINITY or visit http://customer.comcast.com/ to disable this "feature". But in my experience, even without that... their modems suck. They're huge and you don't want to combine your router and modem. I change my routers out more often than my modem needs to be changed... More info > Comcast sued by customers for turning routers into public hotspots You can get much better supported routers in the open market that will give you a much richer feature set and technologies behind them. IMHO it's always a dumb move to get modems and routers from your provider. ESPECIALLY if they're a combined unit. YUCK! No thanks. If they gave one to me... I wouldn't even sell it or give it away... I'd disassemble it for fun, see if any parts could be salvaged. Get that public wifi out of people's houses! It's a disgrace and they should be ashamed for doing it without really telling people. Hey Comcast, install that crap out in your pedestals at the street and pay for your own electricity!

So you can still browse the site but when you go to start a test the page is blank?

No problem, glad that you've got it all figured out. By the way, you've earned enough trust now so you no longer have to wait for posts to be approved. Happy Testing!

Thanks... I'll look closer and see what they're doing over there.

depends on the bitrate.

It must be add-ons or something specific to your machine. I test as expected using SeaMonkey. I could try on windows too but I'd bet money it works as expected. :::.. Internet Speed Test Result Details ..::: Download Connection Speed:: 128011 Kbps or 128 Mbps Download Speed Test Size:: 111.8 MB or 114432 kB or 117178368 bytes Download Binary File Transfer Speed:: 16001 kB/s or 16 MB/s Upload Connection Speed:: 19323 Kbps or 19.3 Mbps Upload Speed Test Size:: 8.4 MB or 8640 kB or 8847360 bytes Upload Binary File Transfer Speed:: 2415 kB/s or 2.4 MB/s Timed:: Download: 7.323 seconds | Upload: 3.663 seconds Tested At:: http://TestMy.net Test Time:: 2015-05-25 18:37:46 Local Time Client Location:: Pike Ntl Forest, CO US https://testmy.net/city/pike_ntl_forest_c Target:: Dallas, TX US https://testmy.net/multithread Validation:: https://testmy.net/db/23HGQ781.bJvVznKQ TiP Measurement Summary (Download):: Min 60.79 Mbps | Middle Avg 129.18 Mbps | Max 133.58 Mbps | 16% Variance TiP Data Points:: 104.21 Mbps, 133.58 Mbps, 132.94 Mbps, 131.9 Mbps, 131.48 Mbps, 132.1 Mbps, 133.15 Mbps, 129.65 Mbps, 126.91 Mbps, 128.27 Mbps, 131.07 Mbps, 130.66 Mbps, 125.02 Mbps, 127.49 Mbps, 104.21 Mbps, 132.1 Mbps, 133.15 Mbps, 132.31 Mbps, 60.79 Mbps Client Stats:: https://testmy.net/compID/4602014672148 User Agent:: Mozilla/5.0 (Macintosh; Intel Mac OS X 10.10; rv:36.0) Gecko/20100101 Firefox/36.0 SeaMonkey/2.33.1 [!] Fresh install, 1st attempt.

Here's how it works. You modify your query to show the way you want... then click "Share This Query" under the graph. The resulting link will display the page to others the way you see it. Here's what I get when I do max results. https://testmy.net/stats/?&t=u&l=100&q=dpitman You only have 86 results right now so max results are 100. As you test more the link above will only show the first 100. Change 100 to 1000 and you can extend that. Hope this helps

As long as you don't mind clearing your settings and addons... this is what I would do at this point. Go to Chrome settings, scroll down and click on "Advanced Settings" then scroll to the bottom and click "Reset Settings" to Restore settings to their original defaults. This is what I would do at this point. Whatever is causing the issues if it's affecting the way the test loads it's most likely affecting the way other sites load as well. Like I said, sometimes just the fact that TMN loads at all is itself a test. It's designed to load the same on any device, any browser... something that's been tested by millions in every scenario under the Sun. So when it doesn't load... it's a strong indication something isn't right. Let us know how it goes.

Sorry, TMN doesn't have an API yet. Maybe released with the next version.

Love it! Thanks for the post!

The Google CDN is being depreciated so that will no longer be an option soon. Just a heads up. I'll work on more alternatives. Bummer.

Hi ohbothersaidpooh, sorry you're having problems. Whether in disappointment, "Oh bother." or in anger, "Oh bother!" it was truly Pooh's way of politely saying "Damn it". Eeyore was always my favorite. Watching and reading it with my daughter now and he still cracks me up. You're making me blush. Do you have wifi and a smart phone or tablet? Any other device really... laptop, other computer. Get connected to the same network and test. If you're able to test (which I bet you will be) then we've established Are you running through a VPN? These can cause issues. TMN is capable of testing VPNs but I've seen odd behavior with certain types. Especially web based anonymizers. That's actually the only time on a healthy connection I've seen things like you're describing. When those are disabled things instantly return to normal. Whatever is causing it... it's a clue to an issue that is most likely affecting your speed in some way. Sometimes when I test a device I don't even need to see the result. I can feel by how the test loads if the device is running great. The progress bar itself loads in different ways depending on how the data is flowing and how the device is processing. This was not intentionally programmed to do this... it just does it. If you use TMN enough you'll have a feeling for what it means when that bar reacts differently. It's very hard to explain but I'm sure others experience can back me up. Just seeing the progress bar (not even having to know how much data is being transferred) you can see fluctuations, jumpiness in the updating... in the linear download test this directly correlates with the data being loaded. You want to see a nice smooth transition across the entire progress. Software and hardware issues can cause the progress to jump. Multithread's progress is always jumpy, because multiple elements can finish simultaneously. Speaking of multithread. Try a multithread speed test. It may run when the linear test won't. Let me know. And let us know if you figure out what's holding you up. It may be causing unexpected behavior on other sites and you may not even realize it yet.

When you're a member some settings that you modify on TMN only affect the computer or device that you're on others are global and affect your entire account. I think it's a great idea to give you linear/multithread selection on the auto test page. Then have it lock on that selection for the duration of the test even if your global settings change. It should do the same for the server selection. I'll keep this in mind while I'm building the latest version. Thank you! Great suggestion.

Hi editorsean! Thank you for pointing that out. What you're seeing is a long standing quirk... a bug but it's funny why it's there. The original design of the database was not well thought out. Instead of doing it over an accepting that I did it wrong the first time I instead wrote around the issues. One of the issues was that the field for test size was originally stored in kB... but the saved information was only allowed to be 5 characters long. So to express anything over 99,999 kB was impossible. Altering the massive database was not an option at the time, it would have required too much downtime. I also didn't want anyone to lose their results. So I started storing by MB in some instances and by kB in others. So 200 MB will save as "200" -- 400 kB saves as "400" 50 MB is saved as "51200". The newer databases are correctly storing the data by kB but they're largely behind the scenes right now. Eventually it won't have to store in tricky ways. ...When the TMN database (mySQL driven) was first started I had absolutely no knowledge of databasing... I'm sure I still do it the wrong way but it's working on a large scale and takes little resources so I can't complain. Keep learning and keep improving.

Damn, that sucks. Post up a pic of your bill where it says the date. I know it's lame but just pay the bill. They have the best Internet in your area, right? If you want to escalate the issue you may be able to get a credit for the inconvenience. But be nice... if you're not getting what you think you're due ask to talk to a manager. Still getting nowhere, send an email http://frontier.com/helpCenter/contact-us. If they are fair I'm sure they'll make it right, they shouldn't disconnect you 2 days early. That's BS.

That's pretty 1337. Imagine the L next to the K is an extension of the K... another K cloaked behind.

You're welcome. Thank you for visiting. I thought your imaginary friend had responded and I just couldn't see it for some reason. So, first... are you positive that you don't have something going on during those hours. Something within your network that might be drawing bandwidth at that specific time each day?? The 2nd picture with the more variable graph... is that during peak? What kind of result do you get on the multithread speed test while you're having the issue (also get a baseline multithread score when it's running great). Hopefully this website helps you accurately illustrate the issue to your ISP and they do something about it.

This topic should help https://testmy.net/ipb/topic/30875-hard-copy-speed-test-results-save-results/?hl=export

HAHAHHAHAHAHA -- if it were done on purpose it would have just been "BUMHOLE" ... I've seen a few funny ones myself. It's completely random. $tid = ""; $possible = "0123456789ABCDEFGHIJKLMNOPQRSTUVWXYZabcdefghijklmnopqrstuvwxyz"; $i = 0; while ($i < 7) { $char = substr($possible, mt_rand(0, strlen($possible)-1), 1); if (!strstr($tid, $char)) { $tid .= $char; $i++; } } Let's make this thread the official funny TID thread!! Okay, they're rare. Just went through pages of recent test results and can't find anything funny like this one. When I find one I'll know where to post it. Just have to happen upon them naturally.

Wouldn't really slow results offset the really fast ones? In theory yes, in reality... no. Example (5+5+5+5+5+5+6+6+6+6+6+6+150+0.01)/12= 18.00 ... vs 6.6 if the high and low were removed. You're right. They should still be included and calculated but what I could do instead is make a new calculation we could call the middle average (or something). The new average will take the heaviest outlying numbers out of the equation. I'll give people the ability to toggle this as a setting when I open the preferences panel to the public. ... the new algorithm will most likely end up being the default.

That's because some guy was blowing 100 Mbps+ upload results recently (in relation to your query). Look on the stats page on the right under the graph. It's considering the last 50 of each type. You can change that. The results page is averaging more than the last 50, currently set at 250 (haven't changed that in years but it may in the future). The results page is lagged a little from the actual stats pages (which are live). The results page can be switched to live mode and I'll definitely make that an option in the next versions (which will have an options control panel). A long while back I made a program that runs through every couple minutes and pre-runs all the numbers so when people load the results page it doesn't actually have to do the calculations every time for everyone. Made results load much faster... but to be honest on this server I can switch it to live mode and barely tell a difference. Where the difference matters is on scale. If it were the old way when the massive amount of reddit traffic hit I think the server would have overloaded under half the traffic it handled. Those calculations add up. But thinking about it, I can make it an option people could toggle. As long as the wide scale is on the more optimized version it's all good. I just looked at the code and it's just a simple variable to switch it to the old live version. ...why not let people access that? :-/ So to answer your question... it's both. Both are right, the question asked is what changed. --- but also realize because the results page stats run a couple minutes behind so even if you set the stats page to 250 results it won't match. You'd have to refresh the page at the exact time the compiler program comes around to run calculation on the corresponding database. Pretty much impossible timing. Less popular stats pages will match because less changes over the period of time.

See the title above the graph "Compare to Choopa..." (your detected host) -- below the details you'll see a button that says "My Graph" -- clicking that will remember your choice for subsequent tests.