Talk:Gain compression

Page contents not supported in other languages.
From Wikipedia, the free encyclopedia

Interesting how you can make a new page out of a lot of chat on another talk page isnt it?--Light current 07:47, 4 February 2006 (UTC)[reply]

Soft clipping[edit]

Soft clipping and "gain compression" are really the same thing, and I think we should cover them in the same article. I think that would really help clarify the concepts; comparing how they are used. Should the article be here or there? — Omegatron 14:35, 17 March 2006 (UTC)[reply]

Is that why some of the page sounds like lecture notes from someone's audio electronics class? "Thus we have compression of gain." We? Still, I'm glad the page was here since it helped me out. I'll have to come back and cleanup the writing style later. 64.81.175.48 16:57, 5 May 2006 (UTC)[reply]


It would be helpful to see an example of non soft clipping or a plot showing an ideal response for the voltage transfer curve. — Preceding unsigned comment added by Deaf Fish (talk • contribs) 18:54, 1 September 2011 (UTC)[reply]

Split off dynamic range compressors[edit]

THe inclusion of dynamic range comressors into this page seems to be causing some difficulty. Perhaps a completely sperate para ( or different page) is needed to remove any ambiguity. Comments? 8-)--Light current 15:00, 12 May 2006 (UTC)[reply]

I have rearranged para to try to segregate intentional use of compression (agc etc) from the other main maenings on the page. Is this satisfactoily clear now., or does it need more work?--Light current 15:10, 12 May 2006 (UTC)[reply]

Well, the only reason dynamic range compression is on this page is to differentiate them. It already has articles (audio level compression, signal compression, ...). So it should be nearby the discussion of the non-linearity of gain compression. — Omegatron 15:46, 12 May 2006 (UTC)[reply]

OK. Well I hope they are now differentiated satisfactorily. 8-|--Light current 16:22, 12 May 2006 (UTC)[reply]

Confusing[edit]

Quote: "But it's not a change in gain; it's non-linear distortion. The output level stays relatively the same as the input level goes higher." -- What is distorted if not the output? Additionally, I failed to understand the difference between level (dynamic range) compression, gain compression and soft clipping. Is it right that intentions/reasons are different but the effect is the same? Is the soft clipping a technique for DRC? --Javalenok 09:16, 21 February 2007 (UTC)[reply]

Topology of what?[edit]

Not an earth-shattering issue but would someone be able to add a few words to elaborate on this gloss: "due to topology differences", found in the Relevance section? Yes, vacuum tubes and transistors have different "topologies" from each other. There's a lot more glass in a vacuum tube, for instance, and tubes stand taller in a chassis than a typical power FET does. Is this what you mean? :) --Jelsova (talk) 17:16, 16 November 2008 (UTC)[reply]

Who did you write this for?[edit]

You know, this article is completely unintelligible to anyone who doesn't already know what gain compression is.--Atlantictire (talk) 23:24, 24 August 2010 (UTC)[reply]

Caption for lead image[edit]

Since the caption is intended to explain the image, it would really help to say red curve using red font, and green curve using green font so readers can quickly find the explanations for those curves. It also helps to bold face the bottom explanation of what is the definition of the article. Why the objection? Mr. PIM (talk) 22:21, 13 June 2013 (UTC)[reply]

The image itself already has the green and red curves identified. We don't need green or red fonts, and we don't need bolding. Binksternet (talk) 01:17, 14 June 2013 (UTC)[reply]
Yes, the image does identify the curves, but
a) the descriptions within the image are much more terse
b) it's pretty difficult to read the descriptions from within the article
Your argument of "We don't need the colored fonts" is an WP:IDONTLIKEIT type of an argument. Can you provide more policy based arguments against the formats? I provided arguments that the colors improve the readability of the caption and make it easier to quickly get the information. Mr. PIM (talk) 01:54, 14 June 2013 (UTC)[reply]
I'm sorry, I thought the relevant manual of style sections were better known. Here they are:
There it is. Binksternet (talk) 02:27, 14 June 2013 (UTC)[reply]
The Manual of Style does not quite carry the weight of a Wikipedia policy, but regardless, the manual of style vis a vis text color applies to the prose, not captions of figures (you even wrote that in your summary of the manual). Regarding the boldface, I have no problem with making the last sentence of the caption italics rather than bold to conform to the Manual of Style. Mr. PIM (talk) 19:07, 14 June 2013 (UTC)[reply]
I see no local consensus reached here regarding the dismissal of MOS in the image caption. If you want to invoke WP:Ignore all rules then you will need to achieve consensus. My vote is to stay with MOS. Binksternet (talk) 13:07, 20 September 2013 (UTC)[reply]
You have not established that having color font in the caption is a violation of the MOS. Mr. PIM (talk) 02:58, 21 September 2013 (UTC)[reply]
What is unclear about "No boldface text allowed in image captions"? What is unclear about "Prose text should never be manually colored"? The caption is a prose description of the image. Binksternet (talk) 14:35, 21 September 2013 (UTC)[reply]
The prose is the main body of the article. Captions and tables are separate from the main body. Will convert boldface to italics. Mr. PIM (talk) 23:16, 21 September 2013 (UTC)[reply]
Prose appears in the article lead, the article body, and it may appear in pull quotes, in quotation boxes, and in image captions. This image caption has complete sentences composed of prose. You have got to stop violating MOS. Binksternet (talk) 07:29, 22 September 2013 (UTC)[reply]

Add pics for "Differences between clipping and compression" (and maybe limiting, too)[edit]

The section "Differences between clipping and compression" could be helped with two types of pictures:

  1. A picture of an example transfer function for "hard clipping" vs "soft clipping" vs general "compression" vs "limiting".
  2. A picture of a sine wave or general audio and what happens to it when it has "hard clipping" vs "soft clipping" vs general "compression" vs "limiting".

Em3rgent0rdr (talk) 07:11, 8 March 2024 (UTC)[reply]

Have you seen examples of such diagrams in other sources? ~Kvng (talk) 14:09, 11 March 2024 (UTC)[reply]
"Electronics for Guitarists" pg 134 Figure 5.4 and Fig 5.5 and 5.8 give transfer function examples which has the *horizontal* section of the transfer functions specifically labeled as "clipping": https://www.google.com/books/edition/Electronics_for_Guitarists/PPg5_lPQJyMC?hl=en&gbpv=1&pg=PA134&printsec=frontcover&dq=clipper
Pg 145 Fig 5.15 gives diode circuits examples which also have the horizonal sections of the transfer functions labeled as clipping, and specifically 5.15b uses a germanium diode and shows how the transfer function has a more gradual curve and shows how the sine wave clips more softly: https://www.google.com/books/edition/Electronics_for_Guitarists/PPg5_lPQJyMC?hl=en&gbpv=1&pg=PA145&printsec=frontcover&dq=clipper
My understanding of Hard-clip vs soft-clip vs limiter vs general-compression examples
I quickly started to sketch out a diagram to distinguish these terms, by having the transfer function curve on the left and having a sine wave example on the right. The transfer functions at low values here are drawn with a slope of exactly 1:1. I'd have the transfer function of "Hard clip" then abruptly change from that slope into being perfectly horizontal. "Soft clip" would have a smooth knee during the transition before leveling off to be horizontal. And skimming the internet and seems there is a loose rule of thumb that 20:1 ratio or higher is considered "limiter", so I would draw a transfer function with a limiter as having a slope that is almost horizontal...and then I would also have another transfer function that includes a general example of compression as having about a 2:1 ratio. Here is some usage of the ratio range for the word "limiter":
  • https://www.iconcollective.edu/audio-compressor-ratio-explained says anywhere from 20:1 to infinity:1 is "limiting". (I would however specifically say infinity:1 is "clipping")
  • https://www.practical-music-production.com/audio-limiter/ also uses 20:1 to distinguish: "an audio limiter (also known as a sound limiter) is usually set at 20:1 or higher, going up to infinity:1 (∞:1). This is the biggest difference you’ll find between a compressor and a limiter."
  • https://www.fullcompass.com/gearcast/compressors-and-limiters-de-mystified is also using 20:1 to distinguish: "Some people consider a compression ratio over 20:1 as essentially the same as limiting, and will call a compressor that can do 20:1 compression a “compressor/limiter.” This can be a source of endless debate on internet forums, but the reality is that the audio industry has no language police—so there’s no definitive answer on whether a 20:1 compression can be called limiting, or whether the ratio has to be ∞:1."
But those examples for limiter ratios are from the audio DAW mixing world, not analog amplifiers. So I don't know how exactly that term is defined for the world of gain compression. Em3rgent0rdr (talk) 22:13, 11 March 2024 (UTC)[reply]
That's good as far as it goes. The piece missing is that most compressors and limiters have time constants that act to reduce the overall gain so you can have compression or limiting that changes the envelope with little or no waveform distortion. Note of these diagrams apply to these devices. The diagrams that apply to these devices are illustrated as in Dynamic range compression § Types. ~Kvng (talk) 22:49, 11 March 2024 (UTC)[reply]
ok yes.
After reading Dynamic range compression article, I came to the conclusion that all to this discussion is out-of-scope of this Gain compression article, so I did a big edit that cut out a ton from this article, which should is only concerned with unintentional compression from amplifier circuits. The only part of this discussion I cut that would be relevant to this article might be different actual amplifier circuits that experience different degrees of compression, like that Figure 5.4 of an simple opamp with its hard-clipped transfer function, and maybe for comparison give a simple transistor amplifier circuit that experiences a soft clipping, and maybe another amplifier circuit with a more gradual transfer function like the figure currently in the article. Em3rgent0rdr (talk) 05:38, 12 March 2024 (UTC)[reply]
Retrieved from "https://en.wikipedia.org/w/index.php?title=Talk:Gain_compression&oldid=1213301206"