tables with a size which is not a power of two?

[tjingboem]

https://csound.com/manual/SiggenTableacc.html

Is this Note no longer true for 7.0?
[Note] Note
Although tables with a size which is not a power of two can be created using a negative size (see f score statement), some opcodes will not accept them.

[vlazzarini]

no longer true. Also there is no need to use negative size since 6.0.

[tjingboem]

good. i will remove this Note.
Could you review https://csound.com/manual/f.html and see if this information is still correct in this respect?

[joachimheintz]

yes but i suggest that on this manual page we should mention the issues we discussed about the extended guard point and its effect on the reported table size. perhaps victor can say whether this is correct and can go into the manual: Note that setting a *positive* table size as a power of two plus one is interpreted by Csound as a power of two size with extended guard point for interpolation. E.g. positive table size 8 and 9 both will report 8 as actual table size: ftlen(ftgen(0,0,8,2,0)) -> 8 ftlen(ftgen(0,0,9,2,0)) -> 8 A *negative* table size however will set the real size of the table, without distiguishing between power-of-two-plus-one and other sizes: ftlen(ftgen(0,0,-8,2,0)) -> 8 ftlen(ftgen(0,0,-9,2,0)) -> 9

…

On 30/04/2024 10:02, vlazzarini wrote: no longer true. Also there is no need to use negative size since 6.0. — Reply to this email directly, view it on GitHub <#706 (comment)>, or unsubscribe <https://github.com/notifications/unsubscribe-auth/AAQYHKS63L4YK2AARYV24V3Y75F25AVCNFSM6AAAAABG7V52BSVHI2DSMVQWIX3LMV43OSLTON2WKQ3PNVWWK3TUHMZDAOBUGY2TEMZUGQ>. You are receiving this because you are subscribed to this thread.Message ID: ***@***.***>

[csounder]

important distinction and clarification *- Dr.B* *Dr. Richard Boulanger* Professor Electronic Production and Design *Berklee College of Music* Professional Writing & Technology Division On Tue, Apr 30, 2024 at 12:26 PM joachimheintz ***@***.***> wrote:

…

yes but i suggest that on this manual page we should mention the issues we discussed about the extended guard point and its effect on the reported table size. perhaps victor can say whether this is correct and can go into the manual: Note that setting a *positive* table size as a power of two plus one is interpreted by Csound as a power of two size with extended guard point for interpolation. E.g. positive table size 8 and 9 both will report 8 as actual table size: ftlen(ftgen(0,0,8,2,0)) -> 8 ftlen(ftgen(0,0,9,2,0)) -> 8 A *negative* table size however will set the real size of the table, without distiguishing between power-of-two-plus-one and other sizes: ftlen(ftgen(0,0,-8,2,0)) -> 8 ftlen(ftgen(0,0,-9,2,0)) -> 9 On 30/04/2024 10:02, vlazzarini wrote: > no longer true. Also there is no need to use negative size since 6.0. > > — > Reply to this email directly, view it on GitHub > <#706 (comment)>, <https://us-west-2.protection.sophos.com?d=github.com&u=aHR0cHM6Ly9naXRodWIuY29tL2Nzb3VuZC9tYW51YWwvaXNzdWVzLzcwNiNpc3N1ZWNvbW1lbnQtMjA4NDY1MjM0NA==&i=NWYxNzBkMDNiNTVmZGEwZmIyNjczYmRm&t=TGxqVXV3TXVWbVAxczRUYjU2eTUwYk5pL0NZMEZwcmhSbmg1TXVNcFpBQT0=&h=b40a6e4dd5994ec9a3d729e64aced802&s=AVNPUEhUT0NFTkNSWVBUSVZ-CzEl059Ym1RlqwKl3ZhNfC-dR2N5IDL9CUAdMA_6tA%3E,> > or unsubscribe > <https://github.com/notifications/unsubscribe-auth/AAQYHKS63L4YK2AARYV24V3Y75F25AVCNFSM6AAAAABG7V52BSVHI2DSMVQWIX3LMV43OSLTON2WKQ3PNVWWK3TUHMZDAOBUGY2TEMZUGQ>. <https://us-west-2.protection.sophos.com?d=github.com&u=aHR0cHM6Ly9naXRodWIuY29tL25vdGlmaWNhdGlvbnMvdW5zdWJzY3JpYmUtYXV0aC9BQVFZSEtTNjNMNFlLMkFBUllWMjRWM1k3NUYyNUFWQ05GU002QUFBQUFCRzdWNTJCU1ZISTJEU01WUVdJWDNMTVY0M09TTFRPTjJXS1EzUE5WV1dLM1RVSE1aREFPQlVHWTJURU1aVUdR&i=NWYxNzBkMDNiNTVmZGEwZmIyNjczYmRm&t=blBlMXVwclI3RDl4Zkdad2tKSUZJTjRXbml6WFJtS2oycUdJenZQampuYz0=&h=b40a6e4dd5994ec9a3d729e64aced802&s=AVNPUEhUT0NFTkNSWVBUSVZ-CzEl059Ym1RlqwKl3ZhNfC-dR2N5IDL9CUAdMA_6tA%3E.> > You are receiving this because you are subscribed to this thread.Message > ID: ***@***.***> > — Reply to this email directly, view it on GitHub <https://us-west-2.protection.sophos.com?d=github.com&u=aHR0cHM6Ly9naXRodWIuY29tL2Nzb3VuZC9tYW51YWwvaXNzdWVzLzcwNiNpc3N1ZWNvbW1lbnQtMjA4NTg3MDcyNQ==&i=NWYxNzBkMDNiNTVmZGEwZmIyNjczYmRm&t=clhSenZVYnFFb0NuYmFzVW9pM1MzaDN1cFFPQ1lNa0VyMjhLc3hHR2o2RT0=&h=b40a6e4dd5994ec9a3d729e64aced802&s=AVNPUEhUT0NFTkNSWVBUSVZ-CzEl059Ym1RlqwKl3ZhNfC-dR2N5IDL9CUAdMA_6tA>, or unsubscribe <https://us-west-2.protection.sophos.com?d=github.com&u=aHR0cHM6Ly9naXRodWIuY29tL25vdGlmaWNhdGlvbnMvdW5zdWJzY3JpYmUtYXV0aC9BQUxXWUZXVllGR0xJQ1JZTk1HR1JEVFk3N0EyWEFWQ05GU002QUFBQUFCRzdWNTJCU1ZISTJEU01WUVdJWDNMTVY0M09TTFRPTjJXS1EzUE5WV1dLM1RVSE1aREFPQlZIQTNUQU5aU0dV&i=NWYxNzBkMDNiNTVmZGEwZmIyNjczYmRm&t=dytkNEFJUDhBdlZYV1lMNFp5d3N3VDV3VU9FcW94Tms1aTFlQytQWHNVRT0=&h=b40a6e4dd5994ec9a3d729e64aced802&s=AVNPUEhUT0NFTkNSWVBUSVZ-CzEl059Ym1RlqwKl3ZhNfC-dR2N5IDL9CUAdMA_6tA> . You are receiving this because you are subscribed to this thread.Message ID: ***@***.***>

[tjingboem]

GEN20 still Must be a power of 2 ( + 1).