Talk:Adaptive noise cancelling

This seems like this could be covered by Noise cancellation or Adaptive filtering. Pinging WP EE and WP Acoustics for their thoughts. AngusW🐶🐶F (bark • sniff) 18:50, 21 January 2023 (UTC) Reply

The noise cancelling link goes to Active Noise Control. This is probably because noise cancelling headphones are covered under this topic.

Adaptive Noise Cancelling and Active Noise Control are similar sounding terms but are in fact different topics in different scientific disciplines with some incidental overlap. Active Noise Control is a topic in the field of acoustics and Adaptive Noise Cancelling is topic in the field of adaptive systems and signal processing.

Active noise control is an area of acoustics that deals with the suppression of sounds that represent noise in physical spaces by introducing another version of the noise sound. In this context 'noise' means noise in the commonly used sense of an undesirable sound that one can hear.

According to the Active Noise Control article this has apparently been attempted without adaptive techniques, long before the development of adaptive noise cancelling techniques in signal processing emerged. However, once adaptive noise cancelling emerged as a technique in signal processing it was adopted as a technique that could also be used in acoustic active noise control.

As the draft article explains, Adaptive Noise Cancelling is a technique evolved from work in adaptive filters and had nothing to do with work in active noise control in acoustics. In Adaptive Noise Cancelling 'noise' refers to any interference that is corrupting a target signal in communication or control - an artefact in electric or electromagnetic signals. The terms 'noise' and 'interference' in this context are used interchangeably. In fact in the book Adaptive Signal Processing by Widrow and Stern the chapter that deals with adaptive noise cancelling is entitled "Adaptive Interference Cancelling". However, the term "adaptive noise cancelling" is the term now well established and widely used.

So the application of adaptive noise cancelling in active noise control is somewhat incidental and represents an intersection between the two fields that is a minute part of both fields. Active (acoustic) noise cancelling can be done with or without adaptive techniques and only a few noise cancelling headphones use adaptive processing. On the other hand, active noise control is just one of the many applications of adaptive noise cancelling (see the draft article).

So Active Noise Control and Adaptive Noise Cancelling are two separate topics that are appropriately dealt with by separate articles that reference each other. John Kaunitz (talk) 02:51, 29 January 2023 (UTC) Reply

Yes, I agree that Active Noise Control and Adaptive Noise Cancelling are two separate topics. However, I think that Adaptive Noise Cancelling should be covered under Adaptive filtering, to avoid duplication.

Wikipedia is many things. One of these things is that it is disorganized. We have a lot of articles that largely duplicate other articles. That is not inherently a problem, if the duplicated information is correct and consistent. The problem is maintaining the articles. You may think that once you write the article and get it right, that it will be done. Unfortunately, there is constant vandalism. Right now, for every active experienced editor there are 1000 articles. I have about 900 articles on my watch list. If I was watching my fair share, I would have about 2000 because many of the active editors are doing other things. Almost all my time goes into article preservation.

Lots of times vandalism is obvious. It’s a few random words of nonsense or foul language. We have a bot that does a good job of fixing that. But sometimes, it is subtle, like changing a single plus sign to a negative sign. It is not obvious that it is vandalism. It might be a legitimate correction. Someone, like me, must look at the change and determine if it is correct. If it is not correct, we also must look at the editor to see if he or she is showing a pattern of such changes to determine if they are a deliberate vandal or just a good faith editor making mistakes.

So, I groan (I guess I’m getting old and weary) when someone wants to add an article that seems largely redundant (to me) with another article. But it won’t make the whole thing collapse. But it provides more target for the vandals. Constant314 (talk) 03:05, 29 January 2023 (UTC) Reply

I would have thought that the major considerations in presenting a topic on Wikipedia is its notability and how to best serves the requirements of potential Wikipedia users who google the term. As I see it, users are looking for concise easy-to-understand articles that provide answers to questions relating to notable topics, such as: What is it? What does it do? How does it work? How did it arise? Where is it relevant? The Adaptive Noise Cancelling draft was produced with this in mind, given that no such article currently exist.

Adaptive filter s and the LMS algorithm, key to concept, were invented by Widrow and Hoff around 1960 at Stanford. The extensive analysis of adaptive filtering in the context of optimal filtering in stochastic signal processing during the next 15 years was published in numerous papers and reports.

Adaptive noise cancelling was proposed around 1970 also by Widrow as a radically new and different use of adaptive filters to eliminate interference corrupting a signal where a second ‘pure’ version of the interference was available. The early work on adaptive noise cancelling was summarised and presented to the professional community in a 1975 paper by Widrow et al. "Adaptive Noise Cancelling: Principles and Applications" as a major conceptual innovation beyond the adaptive filter itself, which opened the way to numerous `new applications. In fact this turned out to be a seminal paper that to date has been cited by 2878 scientific papers and 382 patents. https://ieeexplore.ieee.org/document/1451965

Adaptive filters and adaptive noise cancelling have thus been considered notable innovations and concepts in their own right both by Widrow and the professional community et large. However, Adaptive Noise Cancelling is actually the term most likely to be googled by the public as it is now used in the context of noise cancelling headphones in widespread use. Adaptive Filters and Adaptive Noise Cancelling are thus both notable topics in their own right that merit separate Wikipedia articles.

Whilst the analysis and terminology of adaptive filters and algorithms is in terms of stochastic signals, adaptive noise cancelling usually involves interference and signals such as machinery noise or ECGs which are more appropriately treated as time varying deterministic signals. The draft article uses this model and terminology.

The draft refers to the existing Adaptive Filters and Least-mean-squares Filter articles wherever possible but the explanation of the difference between the classical and adaptive noise cancelling configurations necessitates some direct discussion for the above reason.

I suggest that the Adaptive Noise Cancelling draft should be considered on its merits, given the above and existing articles on related topics. John Kaunitz (talk) 01:52, 30 January 2023 (UTC) Reply

We have tricks called anchors (Template:Anchor) and redirects. The anchor is a bit of hidden code to which you can point a redirect. If you add your material to Adaptive filter and add an anchor, then you can create a redirect called Adaptive Noise Cancelling that will take the reader directly to the section about noise canceling. It serves the reader's needs and improves maintainability. Constant314 (talk) 02:17, 30 January 2023 (UTC) Reply

Adaptive noise cancelling introduced new terminology(reference) related to this configuration of the adaptive filter which differs from the previous classical treatment. Whilst the previous analysis and terminology of adaptive filters and algorithms is in terms of stochastic signals, adaptive noise cancelling usually involves interference and signals such as machinery noise or ECGs which are more appropriately treated as time varying deterministic signals. The draft article uses this model and terminology. Combining the two terminologies in the one (Adaptive Filter) article would be most awkward and confusing.

The draft refers to the existing Adaptive Filter s and Least-mean-squares Filter articles wherever possible but the explanation of the difference between the classical and adaptive noise cancelling configurations necessitates some direct discussion for the above reason.

In summary, basic adaptive filtering and adaptive noise cancelling are two notable concepts and topics in their own right that justify separate Wikipedia articles that provide a simple, easily understandable explanation of each.

I trust that the Adaptive Noise Cancelling draft will be considered on its merits, given the above and existing Wikipedia articles on relevant topics. John Kaunitz (talk) 02:25, 30 January 2023 (UTC) Reply

Notability is important for all material added to Wikipedia, but not so important when it comes to whether it is added to an existing article, or it becomes a separate article. But I have had my say, so let me end with wishing you good luck and happy editing. Constant314 (talk) 02:30, 30 January 2023 (UTC) Reply

Sure there is overlap with existing articles but this is a valuable contribution. I don't see a good reason to decline it. The organization of coverage in this area could use improvement and adding this article doesn't help with that but this contribution does provide a lot of information not currently available in existing articles. We can improve organization over time. 14:05, 25 March 2023 (UTC)https://en.wikipedia.org/wiki/User_talk:Kvng

kaunitzj (John Kaunitz) is the author or co-author of the following references cited in the article:

"General Purpose Hybrid Adaptive Signal Processor" (1971). Published as a Stanford University Technical Report and Google book. https://books.google.com.au/books/about/General_Purpose_Hybrid_Adaptive_Signal_P.html?id=nDMFAAAAIAAJ&redir_esc=y

"Adaptive Filtering of Broadband Signals as Applied to Noise Cancelling" (1972) As a PhD dissertation this work was approved by a committee of three Stanford professors and meets the requirement of "original contribution to scholarship or scientific knowledge" https://gap.stanford.edu/handbooks/gap-handbook/chapter-4/subchapter-8/page-4-8-1. It has also been published as a Stanford University Technical Report and Google book. https://books.google.com.au/books/about/Adaptive_Filtering_of_Broadband_Signals.html?id=q-oiYAAACAAJ&redir_esc=y https://www.proquest.com/openview/a0708ba4c8f5d960c99737e7628be8af/1?pq-origsite=gscholar&cbl=18750&diss=y

"Noise Subtracting Filter Study" (1973) Published as a Stanford University Technical Report.

All of the above are published by a reputable academic institution and meet Wikipedia's criteria for a recognised reliable source

John Kaunitz is also one of 9 co-authors of the widely referenced paper "Adaptive Noise Cancelling; Principles and Applications" by Widrow et al. Proceedings of the IEEE (Dec 1975). Kaunitz had minimal active involvement in producing this paper but in parts it relies on his previous work. https://ieeexplore.ieee.org/document/1451965 To date this has been cited by 2878 scientific papers and 382 patents.

John Kaunitz has had no subsequent involvement in the field of adaptive systems over the past 50 years and has none at present other than the draft of this article. The draft is the result of a survey of currently available sources on the subject. The article is factual, expresses no opinions and is not known to be in conflict with other publications. The article represents no conflict of interest.

Given this, is there any objection to removing the {{COI }} banner from the article? ~Kvng (talk) 04:18, 29 May 2023 (UTC) Reply

​ ​ The Talk page should have a declared connected user blurb added, but I agree with removing the article tag from this article. This is one time I almost wish there was an external link to a "sound fence" product for canceling typical car and airplane noise outside in yards, while allowing wind chimes sounds, for example. :) -- Yae4 (talk) 16:09, 30 May 2023 (UTC) Reply

I have added {{Connected contributor }} above and removed the {{COI }} template from the article. ~Kvng (talk) 19:07, 30 May 2023 (UTC) Reply

• Comment: Active noise control is a lightweight article entirely about control of acoustic noise and does not go into the theory or the math. Active noise control would be an application of Adaptive noise cancelling, so there is some merit in more discussion of active noise cancelling. However, given the substantial overlap with Adaptive filter, I believe that the material offered in this draft should be added to Adaptive filter. Part of the material should go in the application section and part should be used to improve the theory and history sections. The additional references would also be appreciated. Note: Adaptive noise cancelling is not unorthodox. Constant314 (talk) 23:23, 21 January 2023 (UTC) Reply

This article was prepared in good faith, in line with Wikipedia’s general invitation for contributions, after familiarisation with Wikipedia’s editorial and review policies and practices.

The article covers a notable topic, currently not covered in Wikipedia. It is written in the author’s own words based on reliable and verifiable sources. Hence it is believed to meet Wikipedia’s key editorial criteria.

[].

• The article was submitted for publication in early December 2022.

• The article was declined by TheChunky (talk) on the basis that information was unsourced and of questionable notability.

• After a further extensive literature search, significant changes were made to the article to highlight the sources on which the article was based, increase coverage by referencing additional applications and including citations both in-line and attached to section headings indicating the sources on which those sections are based. This resulted in what is considered an improved article and it was resubmitted in early January 2023.

• The article was again declined by AngusWoof (talk) requesting consideration for including the content in existing articles: Active Noise Control and Adaptive Filter. The reviewer did not imply the draft article needed to be otherwise modified.

• In response, a paragraph was inserted in the article explaining that in spite of similar names, Adaptive Noise Cancelling and Active noise control are different topics in different fields, with some overlap. The suggestion to include the content of the draft in the Adaptive Filter article is discussed above. It is pointed out that this would be confusing, problematic and inappropriate. The two topics are significant in their own right and are best covered by separate articles. I am willing and planning to contribute to improve the Adaptive Filter and Least Mean Squares Filter articles (which actually cover the same topic) but I don't see how these can combined with this draft without an adverse result. The modified article was then resubmitted for publication in early February 2023.

• The article was again declined in 23 February 2023 by Pbritti (talk) for the ‘procedural’ reason that the article had not been sufficiently modified in light of previous comments. In response to a request for clarification the reviewer responded that the article lacked in-line citations, that the meaning of citations attached to section headings was unclear, that the article should be a third of its length (without explaining why) and that the style should be improved.

These reasons appear to be at odds with Wikipedia’s published review guidelines in that 1) it is not clear why the declined article was too long at 4500 words (articles are considered too long when they exceed about 8000 words). (see Article Size) and 2) the reasons relating to in-line citation appear to be inconsistent with Wikipedia’s editorial guidelines. See General standards and invalid reasons for declining submissions:

"1. Avoid declining an article because it correctly uses general references to support some or all of the material. The content and sourcing policies require inline citations for only four specific types of material, most commonly direct quotations and contentious material (whether negative, positive or neutral) about living persons."

In light of the above, the reason for declining of the article on 23 February, appears highly questionable. In the interest of progress I nevertheless shortened the article by removing some sections, replaced the headings attached to citations by repeated inline citations and reworded some parts. This resulted in the current draft resubmitted in early April. I hope the draft will receive fair consideration. John Kaunitz (talk)

This has been in AfC purgatory for long enough. I have accepted the draft. If anyone believes the article should not exist on Wikipedia, you are welcome to nominate it for deletion. ~Kvng (talk) 14:58, 20 April 2023 (UTC) Reply

There is far too much dependence on primary sources for this to be a B-class article. When a technique has been around fifty or so years, there will be high quality reliable secondary sources (WP:HQRS) for anything that is notable. These should be found before promoting the article to B-class. Constant314 (talk) 16:39, 29 May 2026 (UTC) Reply

Nor sure what your concern is. The invention of the adaptive noise cancelling concept and technique was documented in a Stanford Ph.D. dissertation The Adaptive Filtering of Broadband Signals as Applied to Noise Cancelling. The criteria for a Stanford Ph.D. thesis is as follows: "The doctoral dissertation must be an original contribution to scholarship or scientific knowledge and must exemplify the highest standards of the discipline". Doctoral Degree Requirements | Stanford University Bulletin. So, the proof-of-concept demonstrations described in the article describe original research that has not been demonstrated before. This document is cited in the article so, it is not clear what additional citations are required.

The adaptive noise cancelling concept was published more broadly by a widely referenced paper Adaptive Noise Cancelling Concept and Applications that has been cited by 3135 scientific papers and 409 patents.

The Adaptive Noise Cancelling article covers this topic and cites numerous references that deal with it. So it is not clear what else is required for this article to qualify for as a B-class article. John Kaunitz (talk) 13:12, 30 May 2026 (UTC) Reply

Correction: In the previous post, Adaptive Noise Cancelling: Concepts and Applications should read: Adaptive Noise Cancelling: Principles and Applications.

Note that in the Adaptive Noise Cancelling article the section Applications, relies entirely on secondary sources with 10 references quoted. So, the criticism that this article lacks secondary sources appears unfounded. John Kaunitz (talk) 22:54, 30 May 2026 (UTC) Reply

My concerns about the quality of the article are several. I am not contesting the veracity of the content.

• 1. I did not say that there were no secondary sources. I said that there was an over dependence on primary sources.
• 2. The IEEE paper, Adaptive Noise Cancelling Concept and Applications, that has been cited by 3135 scientific papers and 409 patents is a high quality primary source. It is certainly an important paper that got people's attention and got them to thinking. It does not establish priority of the concept. It only established that its content was published on a certain date. It is a reliable source for the state of the art at the time, but it is not a reliable source for contemporary practices. It is a reliable source for the math. In a start class and maybe in a C-class article, primary references are OK, but to move up in quality, regard them as markers to be replaced by reliable secondary sources. Widrow's book is likely a WP:HQRS. Many of the primary reference could be replaced with references to that book or other books. Please include page numbers.
• 3. I am not impugning the content of your thesis. I am criticizing the suitability of Ph.D. theses as references in Wikipedia articles. The criteria that a Stanford Ph.D. thesis must be an original contribution to scholarship or scientific knowledge and must exemplify the highest standards of the discipline is standard language that is the same as 1000s of other universities. Compliance with that standard does not guarantee notability or quality. The threshold for original contribution is so low that it does not guarantee notability. Highest standards means that there is no cause for scholastic discipline (no manufactured data and no plagiarism). The fact that there is some original content in a thesis is not sufficient to establish that a proof of concept has been revealed.
• 4. You are a self-revealed COI editor. Thank you for that. Your right to contribute to this article is not being questioned. However, COI edits are not condoned without a talk page consensus. Inserting references to your own publication should never be done without a strong talk page consensus. These references should be replaced with non-COI references (hopefully WP:RS) to raise the quality of the article.
• 5. There is WP:UNDUE emphasis on your own contribution. The fact that you are one of the authors of Widrow's IEEE paper is not a concern. To avoid the appearance of an autobiography article, you should scrub all mentions of your name in the article.
• 6. There is undue content on genesis, priority, proof-of-concept. The article should concentrate on adaptive noise cancelling, what it is, what it does, how it works, how it is applied. In an article this short, genesis should be a couple of sentences at most. I would not object to saying that Widrow's IEEE paper is widely regarded as a seminal work.

Items 4, 5, and 6 are major problems blocking the article quality. Constant314 (talk) 14:44, 31 May 2026 (UTC) Reply

There is no need to argue over whether an article should be rated B class, and there is no need to scrub names from references. I see your point above but it is rather generic for someone like myself who noticed the post at WT:WikiProject Electronics. What specific problem can be identified? That is, can you quote some text from the article (or a diff) and say what change should occur? Johnuniq (talk) 00:57, 10 June 2026 (UTC) Reply

I have had my say. At this point I would like some other editors to comment and then have an uninvolved, but experienced, editor conduct a new assessment. Constant314 (talk) 03:08, 10 June 2026 (UTC) Reply

You might not have noticed that Johnuniq is an uninvolved, but experienced, editor. You might want to engage further.

I have recently started a review of this article. It will take at least several months to get through.

In the meantime, I have knocked the rating to C since ORTF sees it as a marginal B and you've applied a tag (not allowed on B articles). ~Kvng (talk) 15:00, 16 June 2026 (UTC) Reply

I will wait for the review.

Just for comparison here some other articles in the same area of interest.

• B-Class: Floating-point arithmetic, Fourier Transform, Discrete Fourier Transform
• C-class: digital filter, adaptive filter, impulse response, Least mean squares filter, Kronecker delta, Analogue filter, Fast Fourier transform
• Start-class: Finite impulse response, Infinite impulse response, ADALINE

C-class is a respectable rating for articles in this area of interest. Constant314 (talk) 17:45, 16 June 2026 (UTC) Reply

This article currently carries a tag claiming it relies excessively on references to primary sources and calls for the article to be improved by adding secondary or tertiary sources.

However, this article was produced after a substantial literature search in an attempt to identify such sources and these have already been included as references. So, the remedy recommended to remove the maintenance template has in fact already taken place in the course of creating the article.

The reason for this maintenance template is unclear and it appears unwarranted.

Given the following considerations, it is suggested that this tag and maintenance template should be removed.

1. It is not clear by what measure the article is deemed excessively dependent on primary sources.

The original literature review has resulted in the list of 19 references, which may be categorized as follows:

· The four Stanford Electronics Labs reports (references 1,5,6 and 7) are primary sources. These have been published by Stanford as books with limited print editions, and are catalogued as Google books. These have been deemed to meet Wikipedia's criteria for recognised reliable sources (see COI discussion above).

· Reference 2: Adaptive Noise Cancelling: Principles and Applications by Widrow et al; is both a primary and secondary source. As a secondary source, the article provides a survey of related early works and applications, including all of the above primary sources. As a primary source the article presents some original research results.

· The other 14 references are considered secondary sources that are either books providing a comprehensive coverage of adaptive signal processing or are examples of applications that cite Reference 2 as the general reference on the topic.

It is not clear why this mix of the 19 references provided would represent an excessive reliance on primary sources.

2. As a result of early editorial suggestions, there are numerous in-line citations in the article. In general, in each instance, all relevant references are cited and this includes both the primary and secondary sources. So, even though this results in numerous citations of primary sources, this does not mean reliance on primary sources, as long as secondary sources are also listed alongside. With a few exceptions this is almost always the case.

Wikipedia policies recognizes the relevance of primary sources, where historical precedence is of significance.

3. As I understand, Wikipedia’s preference for secondary sources is to establish notability of the topic. However, this is addressed in the article directly by pointing out the large number of citations of Reference 2 over the past 50 years.

The content of the article encompasses the key principles of adaptive noise cancelling, the mathematical model and applications. The first two of these aspects have not signifiantly changed since first documented in Reference 1, The Adaptive Filtering of Broadband Signals as Applied to Noise Cancelling, and popularized in Reference 2, Adaptive Noise Cancelling: Principles and Applications. This is why Reference 2 remains the standard reference after 50 years. Over the past 3 years, since this Wikipedia article was first published, it has been cited by a further 200 scientific papers and 20 patents.

In light of the above, it is difficult to see that the quality of sourcing of information for this article represents a problem, when compared with the norm for Wikipedia articles, to the extent that it warrants being singled out with a tag. John Kaunitz (talk) 06:29, 27 June 2026 (UTC) Reply

I expect that the review by ~Kvng will confirm my assessment.

However,

• 3, 4, 8, and 9 are secondary.
• 2 is primary but highly regarded. No complaint with that one.
• The rest are primary. Being collated into a book does not make them secondary.

There are 88 citations.

• Twelve are secondary.
• Twelve to #2.
• There are 64 citations to other primary papers.
• Notably there 41 to Kaunitz, the COI editor, none of which have been vetted by a talk page discussion.

Constant314 (talk) 12:28, 27 June 2026 (UTC) Reply

You fail to point out that out of the 41 cases where citations to my paper occur, there is only one instance where there are no other references cited alongside. In all other cases a number of references are quoted to support the point involved.

Wikipedia recognises that the classification of sources into primary and secondary depends on the context and is not clear cut. In my view, sources that deal with applications of adaptive noise cancelling are secondary because these articles don't deal with the topic of adaptive noise cancelling as such but deal with the topic of the application subject area. Therefore, their relevance here it to establish notability of adaptive noise cancelling. So, what should be counted as a primary or secondary source for the purpose of your count is something we are unlikely to agree on and is something I am not interested in pursuing.

I hope that we are not going to lose sight of the big picture issue here: is this article deficient with respect to sources it relies on to the extent it deserves to be tagged? In any case, the article reflects a reasonable best effort on my part in this respect and, unless you care to point out specific omissions, it is not clear how further improvement is to be brought about. John Kaunitz (talk) 07:41, 28 June 2026 (UTC) Reply

I am an uninvolved administrator who noticed this at a noticeboard. Wikipedia is full of people with opinions and it is not possible to accommodate all of them, all of the time. Has anyone identified specific text with a specific source and stated clearly what the problem is? How many editors think the tag is warranted? How many think not? I suspect the taggers might be in the minority, and if there are no identified and agreed problems, please remove the tag and leave it up to me to respond if it is replaced. I might need to be reminded if I miss something. Johnuniq (talk) 08:48, 28 June 2026 (UTC) Reply

There are 41 instances of Kaunitz citing his own master's thesis. That is reason enough to tag the article. Constant314 (talk) 10:38, 28 June 2026 (UTC) Reply

I know you are an editor in good standing who has been here forever. However, sanctions will follow if you continue to disrupt this page with high-level hand waving. Either find another page to work on, or provide specific details of text from the article with the source, and state why it is a problem. Johnuniq (talk) 11:04, 28 June 2026 (UTC) Reply

I have placed a {{Self-citation}} template, which is indisputable. With that, I will stand down. Constant314 (talk) 11:23, 28 June 2026 (UTC) Reply