Talk:Offset binary

Could we edit the table to represent the different non-8 k excesses that exist?

This table just represent what the excess is when k is 8, (for a 16 bit representation), however it is possible to find models where the excess is not the half of the available bits. In this case, a table is useful to show that. Sistemx (talk) 10:11, 22 January 2020 (UTC)

Related codes

As it stands, the entire "Related codes" section is useless due to a lack of information and context for the tables. —MIGUELbM 09:01, 7 January 2022 (UTC)

Need source for prevalence

Is 2^n-1 for an n-bit number really the most common case, or is it excess-3, i.e., 4 bits with an offset of 3, used to simplify decimal adders? Either way, the article should cite an RS. -- Shmuel (Seymour J.) Metz Username:Chatul (talk) 16:02, 23 June 2025 (UTC)

I find this citation needed disingenuous; However, thank you very much for actually providing a comment in the talk section. Very few do this. For a common case of a DAC (digital to analog converter) (8,16,18,24,etc) the offset binary is the high bit. Where exactly are you finding prevalence in dispute? There are hundreds of DAC chips with this offset-K encoding. It may not be the case in some mathematics discipline or in some comp.sci/digital logic spaces. Ie, an academic common place. If the 'prevalence' was clarified, would it help you feel at ease? Ie, lots of lay people looking at Wikipedia for information on offset-k will probably be trying to understand the mathematics and conversion of a DAC chip to some common twos-complement notation.

The other codes mentioned in the article, I have never seen in my lifetime. I can see that some time ago IBM mainframes and other compute devices may have used alternate encodings in the article. For people studying ancient compute platforms, this encoding might not be 'prevalent'. As with 1s-complement and 2s-complement the high bit has a sign meaning. offset-binary, the title of the article, implies that this is the case. `offset + value`, where offset is $2^{n-1}$ and value is the signed value. ~2025-33113-31 (talk) 22:13, 12 November 2025 (UTC)

To try and be more constructive.

There is no standard for offset binary, but most often with numerical representations the K for an n-bit binary word is K = 2n−1 (for example, the offset for a four-digit binary number would be 23=8). For character set encodings, this may not apply.

I believe it was clear from context. However, I think some page were merged and from a person of 'character encoding' focus, the statement is not true, so you may rightfully object to it. However, for cases where the encoding is mathematical or numeric in nature, it is often the $2^{n-1}$ bit that is 'k' for the offset.

The term often is not definitive and I don't think it needs a reference. It is only to clue the reader that this could likely be their case. ~2025-33113-31 (talk) 23:52, 12 November 2025 (UTC)