User talk:TallKite: Difference between revisions

Line 251:

:: The observation of M2 being sometimes smaller than m2 or A1 makes me wonder if the naming couldn't be somewhat confusing, see [https://dev.xen.wiki/w/User:Xenwolf/SandBox#Test_Module:edo_tools table with M2, m2, A1 of all EDOs between 1 and 100 (inclusively)] using the "patent fifth" <code>round(edo*log2(3/2))</code> in all cases. --[[User:Xenwolf|Xenwolf]] ([[User talk:Xenwolf|talk]]) 14:21, 9 December 2020 (UTC)

::: I see your point, but I can't see any better option. I think calling them anything else would be even more confusing. I suppose we could omit the data for that handful of edos with a fifth > 3\5 or < 4\7. But to really understand these edos, you have to eventually come to terms with the fact that m2 < 0 or m2 > M2. So I think it's important information that should be included in the template. And for the vast majority of edos, the M2/m2/A1 numbers give an instant "lay of the land". Perhaps the solution is to have the template's phrases "Major 2nd", "Minor 2nd" and "Aug 1sn" link to an explanatory article? I could write something up. I've given supersharp and superflat edos quite a bit of thought over the years.

::: On a side note, I feel that there is a tendency for microtonalists to use too many decimal places. We sometimes forget that it's quite hard to hear a 1 cent difference, and a tenth of a cent is nearly impossible to hear. And yet people routinely write out thousandths of a cent. Of course, if an interval is stacked, those fractions of a cent can add up. Thus precision is good when talking about a rank-2 generator. But in any situation where an interval doesn't get stacked, I'm in favor of fewer decimal places. Because it's extra mental work to read the extra numbers, and filter out the unnecessary information. For example, I see no reason for an edo's table of intervals to have hundredths of a cent. Whole cents are fine. When would you ever need to know that e.g. 7\22 isn't 382¢ but actually 381.818¢? And if you needed the exact cents for some techy reason, you wouldn't get them from a xenwiki page. You would just type 7*1200/22 into a spreadsheet or a calculator. In fact, 22edo has interval tables in both formats. Take a look and ask yourself which is more readable? Now it would be a ton of work to edit all those interval tables, not suggesting that. Instead I bring this up because I saw on the linked page your tables for the edomappings. I see that you have the percentage error with a decimal place. I feel whole numbers is plenty accurate for the percentages. And tenths of a cent are fine for the cents errors.

::: Also, I notice what I call the edomapping you call the degree. But in conventional music theory, degree means something else. https://en.wikipedia.org/wiki/Degree_(music) The degree of 7\12 is not 7, but 5 (for 5th). So I don't think degree is the right term. I see you have both the reduced and unreduced edomappings. That's a good idea, but might it be better to have them on two separate rows of the table?

::: Finally, I see that you omit the prime 2 column. I know the information in that column is pretty redundant, but IMO it logically belongs there, and it would make the table more understandable to newbies. --[[User:TallKite|TallKite]] ([[User talk:TallKite|talk]]) 08:38, 10 December 2020 (UTC)

@@ Line 251: / Line 251: @@
 :: The observation of M2 being sometimes smaller than m2 or A1 makes me wonder if the naming couldn't be somewhat confusing, see [https://dev.xen.wiki/w/User:Xenwolf/SandBox#Test_Module:edo_tools table with M2, m2, A1 of all EDOs between 1 and 100 (inclusively)] using the "patent fifth" <code>round(edo*log2(3/2))</code> in all cases. --[[User:Xenwolf|Xenwolf]] ([[User talk:Xenwolf|talk]]) 14:21, 9 December 2020 (UTC)
+::: I see your point, but I can't see any better option. I think calling them anything else would be even more confusing. I suppose we could omit the data for that handful of edos with a fifth > 3\5 or < 4\7. But to really understand these edos, you have to eventually come to terms with the fact that m2 < 0 or m2 > M2. So I think it's important information that should be included in the template. And for the vast majority of edos, the M2/m2/A1 numbers give an instant "lay of the land". Perhaps the solution is to have the template's phrases "Major 2nd", "Minor 2nd" and "Aug 1sn" link to an explanatory article? I could write something up. I've given supersharp and superflat edos quite a bit of thought over the years.
+::: On a side note, I feel that there is a tendency for microtonalists to use too many decimal places. We sometimes forget that it's quite hard to hear a 1 cent difference, and a tenth of a cent is nearly impossible to hear. And yet people routinely write out thousandths of a cent. Of course, if an interval is stacked, those fractions of a cent can add up. Thus precision is good when talking about a rank-2 generator. But in any situation where an interval doesn't get stacked, I'm in favor of fewer decimal places. Because it's extra mental work to read the extra numbers, and filter out the unnecessary information. For example, I see no reason for an edo's table of intervals to have hundredths of a cent. Whole cents are fine. When would you ever need to know that e.g. 7\22 isn't 382¢ but actually 381.818¢? And if you needed the exact cents for some techy reason, you wouldn't get them from a xenwiki page. You would just type 7*1200/22 into a spreadsheet or a calculator. In fact, 22edo has interval tables in both formats. Take a look and ask yourself which is more readable? Now it would be a ton of work to edit all those interval tables, not suggesting that. Instead I bring this up because I saw on the linked page your tables for the edomappings. I see that you have the percentage error with a decimal place. I feel whole numbers is plenty accurate for the percentages. And tenths of a cent are fine for the cents errors.
+::: Also, I notice what I call the edomapping you call the degree. But in conventional music theory, degree means something else. https://en.wikipedia.org/wiki/Degree_(music) The degree of 7\12 is not 7, but 5 (for 5th). So I don't think degree is the right term. I see you have both the reduced and unreduced edomappings. That's a good idea, but might it be better to have them on two separate rows of the table?
+::: Finally, I see that you omit the prime 2 column. I know the information in that column is pretty redundant, but IMO it logically belongs there, and it would make the table more understandable to newbies. --[[User:TallKite|TallKite]] ([[User talk:TallKite|talk]]) 08:38, 10 December 2020 (UTC)