Talk:Bandwidth (signal processing)

Professional sound production Start‑class

	This article is within the scope of WikiProject Professional sound production, a collaborative effort to improve the coverage of sound recording and reproduction on Wikipedia. If you would like to participate, please visit the project page, where you can join the discussion and see a list of open tasks.Professional sound productionWikipedia:WikiProject Professional sound productionTemplate:WikiProject Professional sound productionProfessional sound production
Start	This article has been rated as Start-class on Wikipedia's content assessment scale.
???	This article has not yet received a rating on the project's importance scale.

Telecommunications Unassessed

	Telecommunication portal This article is within the scope of WikiProject Telecommunications, a collaborative effort to improve the coverage of Telecommunications on Wikipedia. If you would like to participate, please visit the project page, where you can join the discussion and see a list of open tasks.TelecommunicationsWikipedia:WikiProject TelecommunicationsTemplate:WikiProject TelecommunicationsTelecommunications
???	This article has not yet received a rating on Wikipedia's content assessment scale.
???	This article has not yet received a rating on the project's importance scale.

old question

Aren't bps-type bandwidths and frequency bandwidths really the same thing? A coax cable can only carry a certain frequency range, or a certain number of bits per second. how do you convert between them? - Omegatron 21:59, Jul 29, 2004 (UTC)

This is given by the Shannon-Hartley theorem. - Omegatron 18:27, Aug 10, 2004 (UTC)

typo?

English is not my native language, but don't you say "to a great extent" and not "to a great extend"? (in the second to the last paragraph).

Yes, "extend" is the verb, "extent" is the noun.--Orthologist 13:36, 26 February 2007 (UTC)[reply]

bandwidth vs. datarate

I think this mixup of "bandwidth" and "data rate" here is quite bad. Bandwidth in communications is a fixed term measured in Hz - it is related to (but not always 1:1) to the datarate (which you could measure in information/time or Bits/sec).

Yes, the distinction between "bandwidth" and "data rate" can be crucial in some (cascaded) settings, as I have illustrated at...

"Bucketing Bandwidth" by Paul Niquette

...to which I would be pleased to see an External Link on this page.Paul Niquette 00:59, 7 December 2006 (UTC)[reply]

Bandwidth vs Throughput

Bandwidth is the frequency, and throughput is actually the Kb/s,Mb/s, etc. If no one disagrees, I will make that change after this discussion posting.

Hmm... I think the term can be used for either. Is that "throughput" before or after encoding? Before or after compression? — Omegatron 19:57, 4 October 2005 (UTC)[reply]

Throughput is the amount of data that can be sent over the wire at any particular time. Yes everyone uses the term bandwidth, but it's just not factual. If you are referring to the OSI model with your question, I would think that it would be atleast level 3 or lower, but I am not positive. ymmotrojam 04:40, 12 October 2005 (UTC)[reply]

Hi! I do agree that the two terms are not the same though describe related processes. Throughput is indeed the amount of data that can transferred via particular channel. But here is one more confusion regarding the information rate, bandwidth and modulation rate. I would appreciate if anybody read the following:

1. Imagine, the road. The road is your bandwidth. The wider the road the more cars can travel at the same time. One car is a bit of information. So the wider the bandwidth the information bits can bit transferred at a time. We got this one. Cool. Here is one more thing you are to comprise. 2. Assume a road with 4 lines, but we have 2 cars on the garage only! So two lines will be used at a time. Turn on your imagination: a car = a truck with some load. The load is the information we want to transfer. The information rate = the rate of those two trucks. We have more space though! So what we can do is to go to a car rent place to get some more cars. Car rate station is our modulator. Truck is a a carrier we load our information at.We did! Hurray! Now we have to divide the load we are to transfer between 4 cars. As we have 4 lines on the road and 4 trucks the load will be delivered faster as twice as we double our speed. Well, in fact the delivery will be any faster... So when we increase modulation rate.. What I know is that we can adjust signal parameters to recall with particular channel parameters. But how about increase in speed?

Equations

Bandwidth/SNR, or Bitrate/SNR theoretical limits -- anyone know where they are on wikipedia, and why they're not on this page? -- is it called Shannon's theory or something? Ojw 14:53, 18 October 2005 (UTC)[reply]

what bandwidth will i used

Graph

It is poor writing style to present a graph with unlabeled axes. I imagine the graph shown has frequency as the horizontal axis, but what does the vertical axis measure? Could the author of the text please label these axes?

--> Vertical axis should be gain in decibel.

Digital bandwidth section

I removed a clause about "with optimal encoding" in the example about a 32-bit data bus. Digital data buses never use a Shannon-Hartley optimal encoding, since that would require reducing the contrast between symbols to account for the high S/N in the bus -- unless the bus S/N ratio is very bad indeed, bilevel binary logic is less than optimal from the standpoint of channel design.

I also removed the phrase "by the uneducated" because it is smarmy and POV. I, too, wish that the bandwidth/capacity channel were not present in common usage, but it is. zowie 01:11, 23 April 2006 (UTC)[reply]

Fractional Bandwidth

I believe the second sentence of "A commonly used quantity is fractional bandwidth. This is the center frequency of a device divided by its bandwidth. E.g., a device that has a bandwidth of 2 MHz with center frequency 10 MHz will have a fractional bandwidth of 2/10, or 20%." should read "...This is the bandwidth of a device divided by its center frequency...". This makes the definition in agree with the example in the third sentence.

Split proposal

Whoever proposed it should say here what they have in mind doing about it. I think it's not a bad idea, if we can agree on sensible boundaries that everyone likes. Dicklyon 06:36, 1 September 2006 (UTC)[reply]

Agreed if we can get some ideas out and about they might be implmented.
Tokyo Michael 17:09, 5 October 2006 (UTC)[reply]

Yes, certainly split the page. -G photo 17:59, 24 October 2006 (UTC)[reply]

Oppose. Having thought it over some more, there's not much here. How about expanding on the various meanings, or the application of the concept to the different fields, and see if it gets large or unwieldy. If so, then split it at that time. Dicklyon 05:00, 27 October 2006 (UTC)[reply]

I don't know what should be done with the article, but bandwidth as it relates to computers needs to be examined in much greater detail - e.g. the physics behind data transmission, the economics of bandwidth, the engineering involved etc. Perhaps a separate article like Bandwidth (computing) could be created for this, since it could easily dwarf the article. I still think this page should have a general scope though as it does now rather than being a disambiguation page. Richard001 06:12, 10 December 2006 (UTC)[reply]

Split: Bandwidth means fundementally different things in the analogue and digital spheres, though they both equate approximately to how much data can be stuffed through a media. Would make more sense to make this page a disambiguation page and have separate articles.

Split - Bandwidth has important mutually exclusive meanings. Data rates and capacities versus frequency ranges. --JJLatWiki 17:31, 23 April 2007 (UTC)[reply]

Split - Bandwith means many different things. I would have this page either be the bits-per-second definition or a disambiguation page listing the bits-per-second meaning, the linear algebra furthest-distance-from-nonzero-to-diagonal definition, and the signal-processing range-of-frequencies definition. Either way, somewhere we should clearly state "Bandwidth has many distinct technical meanings." —Ben FrantzDale (talk) 02:05, 23 January 2008 (UTC)[reply]

Simple English

The article linked on "Simple English" is not just simple; it's completely wrong. Does anyone here who understands anything have an account there to fix it? Dicklyon 15:18, 14 November 2006 (UTC)[reply]

Disambig

I support, for bandwidth can also mean how much a server can take. —The preceding unsigned comment was added by -Slash- (talk • contribs) 20:08, 19 December 2006 (UTC).[reply]

Declining Bandwidth Costs

Can someone provide a web page or information on the historical decline of bandwidth, say from early '80's through today (perhaps with projections to tomorrow)?

Thank you! R/ Porter Clapp Porter Clapp 19:59, 5 March 2007 (UTC)[reply]

To what kind of bandwidth are you referring? Data rate capacity, data count allotments, frequency ranges, or some other definition? And what do you mean by "historical decline"? --JJLatWiki 20:56, 23 May 2007 (UTC)[reply]

Hi! I would add the following link to the additional resources section: [1] It contains some very nice and detailed tutorials on telecommunication basics

Confused in "Digital BW"

I can't understand this sentence:"a digital data bus with a bit rate of 66 Mbit/s on each of 32 separate data lines may properly be said to have a bandwidth of 33 MHz and a capacity of 2.1 Gbit/s " It's because how can 66Mbit/s*33MHz=2.1Gbit/s?? It's there any wrong ? In my opinion, the number of bit of each data should be 66 bits and the bandwidth should be 33MHz then the data rate is gonna be 2.1Gbit/s.. If i get wrong concept on this , pls give me the answer~~ Appreciate! Thanks! —Preceding unsigned comment added by Ryan siow (talk • contribs) 02:41, 1 October 2007 (UTC)[reply]

The writer meant that 66 Mbit/s can be sent through a band of 33 MHz; true, but utterly irrelevant to the meaning of bandwidth in this context, as your confused interpretation shows. The data rate is 32*66 M = 2.1 Gbps. Dicklyon 02:48, 1 October 2007 (UTC)[reply]

Thanks your answer.But i still can't make it clearly. Would u mind tell me what's that mean of "66Mbps"? Thanks! —Preceding unsigned comment added by Ryan siow (talk • contribs) 05:36, 1 October 2007 (UTC)[reply]

66 Mbps or Mbit/s is the bit rate on each wire, in megabits per second. Dicklyon 06:16, 1 October 2007 (UTC)[reply]

computer network bandwidth should be split

The sections of this current page about computer network bandwidth belong in their own separate article. They should be split out, and a disambig. link put at the top of this article. --jacobolus (t) 22:54, 18 October 2007 (UTC)[reply]

Oppose. You mean moving one short paragraph to a separate article about Digital bandwidth in bit/s? It is important that people from analogue electronics, traditional telecommunications and physics become aware about today's widespread usage of the digital definition. It is also important that computer science people are aware of the original analogue definition in Herz, and are encouraged to use less ambiguous terms such as gross bit rate, throughput or channel capacity. Mange01 00:29, 30 October 2007 (UTC)[reply]

I agree with jackobolus that they should be split into:
Analog Bandwidth being a range of frequencies utilized in an application such as an RF communications channel
Digital Bandwidth being the data rate witch differs from throughput (witch should account for transmission overhead) --Douglasmwilliams (talk) 20:43, 13 December 2007 (UTC)[reply]

Support These are different concepts and should be split Lbgrowl (talk) 00:42, 31 December 2007 (UTC)[reply]

Oppose. I believe there is room on this page to support all the meanings of bandwidth. Like Mange01, I think the reader should be able to see the various meanings side by side and so become aware of the similarities. Binksternet (talk) 03:01, 10 January 2008 (UTC)[reply]

Support Analogue bandwidth is a complex topic, with links to control theory. It needs exposition of some analogue concepts totaly irrelevant to someone looking for digital bandiwdth. Digital bandwidth does not deal with similar topics anymore. With multiple digital lines and transmission media, analogue bandwidth in terms information theory and digital technology is not relevant. Where it is (i.e. chanel coding?), a link would be appropriate. 134.225.216.197 (talk)

Support – the modern proliferation of things called bandwidth (the digital stuff) is pretty much distracting in an article that wants to try to communicate the traditional meaning. Make a new article to split off the digital stuff. Dicklyon (talk) 05:15, 31 January 2008 (UTC)[reply]

Support - The digital use of "bandwidth" is definitely an error that has crept into common speech. It is also not correct to say that information rate (digital bw) is proportional to bandwidth (analog) and this should be removed from the entry. Shannon's equation gives C=B*log₂(S/N+1), C in bps and B in Hz which appears linear. However for the common case where ultimate noise, N, is thermally rather than interference limited, this is equal to KTB, thus B shows up inside the logged argument and the relationship is no longer linear (proportional). The maximum information capacity situation, the so called Shannon limit, represents this case; where the bandwidth (actual/analog) has been increased greatly and the signal power is very nearly identical to the noise power; S/N approximately unity. N6gn (talk) 18:14, 5 February 2008 (UTC)[reply]