[mpancha]

(before anyone says anything, this isn't under applications b/c its only for 10.4 and is more of a system tweak)

http://www.apple.com/support/downloa...ndtuner10.html

I just saw this on macslash, its basically a utility that is suppose to give you a speed boost on high latency broadband connections, especially the 5mbps and up connections. I have a 6mbps connection at home, not the best, but the best available in my area. Does this seem like a worthwhile install?

[zerostar]

I was wondering the same thing, I am on 12MB/2MB right now but my latency is around 45-55ms so I don't think that is high, is there an uninstaller provided?

[rickey939]

Originally Posted by zerostar

...is there an uninstaller provided?

Yes.

[cmeisenzahl]

I saw that and was wondering about it myself. When I think of high-latency connectuion, the first thing that comes to my mind is satellite based net access.

Chris

[ShotgunEd]

Originally Posted by mpancha

(before anyone says anything, this isn't under applications b/c its only for 10.4 and is more of a system tweak)

Just because you don't think its an application doesn't mean you should try to supercede the topic.

http://forums.macnn.com/showthread.php?t=276958

If it in were in the wrong place it'd have been locked by a Mod and the poster taunted inexorably.

[ghporter]

Yeah, it sounds like they should have said "The Broadband Tuner allows you to take full advantage of very high speed Internet connections that have a high latency," and then added "like satellite Internet."

Really large TCP numbers don't help that much with lower-latency connections and can sometimes hurt. But satellite kind of defines "high latency," doesn't it?

[badidea]

What exactly is high latency and how do I find out what mine is?

(8MBit/1MBit DSL connection)

[jcarr]

Originally Posted by ghporter

Yeah, it sounds like they should have said "The Broadband Tuner allows you to take full advantage of very high speed Internet connections that have a high latency," and then added "like satellite Internet."

Has anyone tested this with a connection like DirecWAY? Is there a noticeable improvement?

[olePigeon]

Unless this software program somehow breaks the laws of physics, it will be impossible to get low latency on a satelite internet connection. Light can only travel so fast, and when it comes from a satelite in low orbit, it's going to take a several hundred milliseconds to get here.

[gururafiki]

Originally Posted by badidea

What exactly is high latency and how do I find out what mine is?

Yes, somebody please answer badidea...

I have a 6Mbps cable connection...does Broadband Tuner apply to me?

[ghporter]

The idea is not to get "lower latency," which, as badidea points out is impossible. The idea is to make the best of a bad situation and to reduce the "waiting for reply" time. Once a packet is sent, the computer waits for an acknowledgement. If you send small packets, you will spend a while waiting for acknowledgements for every one. If you send HUGE packets, you'll still wait for an acknowledgement of each one, but you'll have a LOT more data at the other end, doing its thing (page requests, etc.).

[JHromadka]

So this would be better for desktop users right, as laptop users would switch among different connection speeds?

[Cadaver]

There's several comments posted on xlr8yourmac's homepage regarding the performance loss/gain with the Apple Broadband Tuner.

Here are my informal tests (Speakeasy's Speed Test):
With = 6090 kbps down, 356 kpbs up
Without = 6094 kbps down, 335 kpbs up

So, for me, no difference. Cable modem -> Apple Airport Extreme basestation -> wired gigabit switch -> G5 (i.e., the G5 is not a wireless client).

For comparison - my Windows XP Pro box (AMD Athlon 3200+, Intel gigabit PCI):
6171 kbps down, 339 kbps up

[Don Pickett]

For the record, looks like this is an implementation for sysctl -w.

[ghporter]

Cadaver, cable Internet tends to have fairly low latency, so I's be surprised if you found much difference. DSL is typically similar, so that would be another place where I would skip the tuner. Basically the only (widely adopted) place most people would find the Broadband Tuner useful would be with a satellite connection.

[Kyle Dreaden]

I have a pretty basic DSL line at home (3mbps/1.5mbps). I went ahead and installed the tweak and did notice a nice little boost in speed.

Nothing massive, but a nice little bump none the less.

[Millennium]

Originally Posted by badidea

What exactly is high latency and how do I find out what mine is?

(8MBit/1MBit DSL connection)

There are two numbers which come into play when determining the speed of an Internet connection: bandwidth and latency. Bandwidth is the more widely-known of the two, and it's the one that you mention in your post: how much data can get through in a given amount of time.

Latency, on the other hand, is how long it takes to get a single packet of data from point A to point B. Let's say that I put 10,000 DVDs into a big box and shipped it to you overnight. The bandwidth of this transfer would be huge: even though it would take a day, 40,000GBytes in a day averages to almost 5Gbits per second. However, the latency would also be enormous, because it took a whole day for even the first byte to get through.

On the other hand, let's say that I had a teleportation device that worked on very small objects. I've thought of a way to encode a byte of data on these objects, but I can only use the teleporter once per second. This kind of transfer would be the ultimate in latency: there's no time delay whatsoever between the time I send an object and the time you get it. But the bandwidth would be very tiny: only 8 bits per second.

These are extreme examples -the second one isn't even realistic at the moment- but I'm trying to illustrate the concept. You could think of bandwidth as a measure of how wide the pipe is, and latency as a measure of how long the pipe is.

Now, what does all of this mean for the average person? High bandwidth is good, and low latency is good, but you may or may not need one more than the other. High bandwidth is good when you need to transfer large amounts of data, but when it gets there isn't critical. Downloading files is an example of just such an application: it doesn't matter whether a byte of data gets there one second or twenty seconds after it's been requested, as long as it gets there in a reasonable amount of time.

Low latency, on the other hand, is best when a given piece of data has to get there as quickly as possible, but there isn't much data that needs to get there at any given time. Online games provide the best example here: it doesn't take much data to tell other people where you are and what way you're facing, but that data needs to get to everyone very quickly or you'll experience lag.

You have a DSL connection, which has relatively high bandwidth and relatively low latency, so you have a well-rounded connection. You probably don't need to use this optimizer. The sorts of people who would benefit from it are people on wireless connections, most notably DIRECWAY and other satellite Internet services. Each piece of data takes a relatively long time to get from point A to point B, so if you send more data at once then it will all get there faster. That's the point of this optimizer.

Why wouldn't someone on a low-latency connection want to use it? Because there's a flip side to the idea of sending more data at once: if something goes wrong with the packet (and this happens all the time, even though we're usually not aware of it) then more data has to be re-sent, and that slows things down. If you have a high-latency connection, then it's better to send small amounts of data, because that way not as much data is wasted when a packet fails to get through.

[ghporter]

Millenium, you have a career ahead of you as a technology instructor. Your analogies are great! Very well explained.

[mpancha]

Originally Posted by ghporter

Millenium, you have a career ahead of you as a technology instructor. Your analogies are great! Very well explained.

agree, that's the best explanation for latency I've come across so far.

Originally Posted by ShotgunEd

Just because you don't think its an application doesn't mean you should try to supercede the topic.

http://forums.macnn.com/showthread.php?t=276958

If it in were in the wrong place it'd have been locked by a Mod and the poster taunted inexorably.

thanks I guess?

for the record i started the topic b/c at the time of my posting it, i did not see the other one. My disclaimer was for those who like to get trigger happy and ridicule those who put a topic in the wrong thread instead of giving constructive criticism. Thanks for your "criticism'.

[ShotgunEd]

Originally Posted by mpancha

for the record i started the topic b/c at the time of my posting it, i did not see the other one. My disclaimer was for those who like to get trigger happy and ridicule those who put a topic in the wrong thread instead of giving constructive criticism. Thanks for your "criticism'.

I had thought you had seen the other topic and had thought, ah fark it, I'm putting mine in the OSX Forum because the other one is in the wrong place.

I thought wrong.

Apologies.

[gururafiki]

Thanks Millennium, now I get it.

[mpancha]

Originally Posted by ShotgunEd

I had thought you had seen the other topic and had thought, ah fark it, I'm putting mine in the OSX Forum because the other one is in the wrong place.

I thought wrong.

Apologies.

no problem.

[badidea]

Yeah, thanks Millennium - this is how I like to get stuff explained!

So is latency the same as the ping time?

[oni]

I have a 6mbps connection at home, not the best, but the best available in my area

Damn internet sucks where I live, im just upgrading to 2Mb/128k and thats really expensive. Looks like no broadband tuner for me

[Kristoff]

Originally Posted by olePigeon

Unless this software program somehow breaks the laws of physics, it will be impossible to get low latency on a satelite internet connection. Light can only travel so fast, and when it comes from a satelite in low orbit, it's going to take a several hundred milliseconds to get here.

Satelites do not transmit light...they transmit radio waves.

[f1000]

Originally Posted by Kristoff

Satelites do not transmit light...they transmit radio waves.

Radio waves are electromagnetic waves just like visible light, so they're subject to the same speed restrictions in a vacuum. Moreover, some military satellites do communicate using light in the form of lasers.

[ghporter]

Actually, radio waves travel a bit slower than light... It's all pretty complex. In any case, using the "speed of light" as a basic speed for radio wave propagation works well for a close approximation. Radio waves are how satellites typically communicate with ground stations-your TV satellite receiver, your XM or Sirius satellite radio, a GPS receiver, or satellite Interned dish.

Further, the satellites used for such things as TV, satellite radio and satellite Internet are in stationary orbits at an altitude of approximately 23,000 miles (GPS satellites orbit at about 11,000 miles, but that's not relevant to this). That's a LONG way for signals to go... and it translates to about a 0.13second propagation time EACH WAY, or about a 0.25second round trip.

[f1000]

Originally Posted by ghporter

Actually, radio waves travel a bit slower than light...

Maybe through air, but not through a vacuum.

[Kristoff]

Originally Posted by f1000

Radio waves are electromagnetic waves just like visible light, so they're subject to the same speed restrictions in a vacuum. Moreover, some military satellites do communicate using light in the form of lasers.

Radio waves and visible light waves have vastly different properties.
You can not distort light in the same ways you can distort radio signals and vice versa.

And current production military satellites do not use lasers. There are experimental systems in development that use them for inter-satellite communication only, but not for Earth bound communication.

[mpancha]

Originally Posted by f1000

Maybe through air, but not through a vacuum.

as true as that may be, the signal from satelite to satelite and back....its not through a vacuum. For purposes of this discussion, what ghporter said sums it up:

Originally Posted by ghporter

Further, the satellites used for such things as TV, satellite radio and satellite Internet are in stationary orbits at an altitude of approximately 23,000 miles (GPS satellites orbit at about 11,000 miles, but that's not relevant to this). That's a LONG way for signals to go... and it translates to about a 0.13second propagation time EACH WAY, or about a 0.25second round trip.

[f1000]

Originally Posted by mpancha

as true as that may be, the signal from satelite to satelite and back....its not through a vacuum. For purposes of this discussion, what ghporter said sums it up:

For all practical purposes, the speed of radio waves through air is very nearly close to the speed of light. ghporter's .25s calculation doesn't have enough significant digits to show the difference caused by air versus vacuum.

EDIT: Nevertheless, ghporter's post was still spot on.

[f1000]

Originally Posted by Kristoff

Radio waves and visible light waves have vastly different properties.
You can not distort light in the same ways you can distort radio signals and vice versa.

Of course not, because condensed matter selectively interacts with photons according to frequency. Radio and visible light photons differ in terms of frequency, so you can't use glass optics to "bend" radio waves. That doesn't mean that radio is qualitatively different from visible light. Both are electromagnetic waves that propagate as photons, and their group velocities are still c in a vacuum.

Both light and radio waves can be bent using gravity, so there's at least one way to "distort" both in the same way.

And current production military satellites do not use lasers. There are experimental systems in development that use them for inter-satellite communication only, but not for Earth bound communication.

In other words, I'm right.

[Kristoff]

Originally Posted by f1000

Of course not, because condensed matter selectively interacts with photons according to frequency. Radio and visible light photons differ in terms of frequency, so you can't use glass optics to "bend" radio waves. That doesn't mean that radio is qualitatively different from visible light. Both are electromagnetic waves that propagate as photons, and their group velocities are still c in a vacuum.

Both light and radio waves can be bent using gravity, so there's at least one way to "distort" both in the same way.



In other words, I'm right.

I learned all that crap in my QED course years ago....but you're not right. You're going off on some tangent.

There are no satellites that use Lasers or light to send signals to earth that anyone would benefit by installing the Apple Broadband Tuner 1.0.

[ghporter]

Originally Posted by f1000

Maybe through air, but not through a vacuum.

A satellite shot's got about 200 miles (round trip) of atmosphere to go through, and there are some interesting effects on all antennas that relate to actual propagation velocity. In a practical sense there isn't much difference, but there is some.

[f1000]

Originally Posted by Kristoff

I learned all that crap in my QED course years ago....but you're not right. You're going off on some tangent.

You're the one who went off on a tangent. Olepigeon wasn't implying that commerical internet satellites used visible light to relay data. If you had read his post carefully, you'd have understood that he was asking how a software fix would speed up signal propagation when radio waves (and all EM waves) are limited by the speed of light. Your snappy retort was that satellites didn't use light, which completely missed the point of his question. ghporter understood what olepigeon was asking and explained that the software didn't bend the laws of physics but adjusted software parameters to compensate for latency due to the extreme physical distances involved.

[drnkn_stylz]

I use wireless networks at school and I have a cable connection at home. I can see that I might get a little better performance for the wireless, but none for cable. However now I am wondering if this will have a negative effect on my cable connection?

[ghporter]

drnkn stylz,

Your cable connection will supply all that it can and nothing more. The cable company provides a specific connection speed, and you use it the way you want. Basically, if you don't find the the Tuner improves your connection (and I'm betting you won't; cable has pretty low latency) then uninstall it. You can't hurt anything mechanically with the Tuner, and you can't "draw too much data" from your cable connection. Nor can the cable company come down on you and say you're doing something wrong.

[olePigeon]

Originally Posted by Kristoff

Satelites do not transmit light...they transmit radio waves.

I meant that you can't go faster than light, whatever signal is being transmitted. Besides, radio is light, just on a different spectrum.

[olePigeon]

Originally Posted by Kristoff

Radio waves and visible light waves have vastly different properties.
You can not distort light in the same ways you can distort radio signals and vice versa.

Yes you can, you just need different materials to do it. For example, shooting an argon laser through an extremely cold, but condensed gas will actually physically slow light down. Depending on temperature and the gas used, light can even be brought to a virtual stand still.

Pretty fascinating. I haven't seen any pictures, but I've always wondered what visible light looks like (if it can be seen) when it's moving slowly.

[f1000]

Originally Posted by olePigeon

Pretty fascinating. I haven't seen any pictures, but I've always wondered what visible light looks like (if it can be seen) when it's moving slowly.

There's no way to slow photons down; photons always travel at c through the vacuum between atoms.

What you're talking about is the delayed re-emission of absorbed photons. This has the effect of slowing down an optical wavefront propagating through a medium. We're definitely on a tangent again.

[B Gallagher]

is there any way to find out what the latency on my mac is?

[waffffffle]

Never underestimate the bandwidth of a station wagon full of tapes hurtling down the highway.