0% found this document useful (0 votes)
679 views5 pages

Allen Digital Organ Stop Cards Procedures VI

method for programming IBM cards to create new stops for Allen digital organs, models 1100, 2100, 3100

Uploaded by

kpsteve
Copyright
© Attribution Non-Commercial (BY-NC)
We take content rights seriously. If you suspect this is your content, claim it here.
Available Formats
Download as DOC, PDF, TXT or read online on Scribd
0% found this document useful (0 votes)
679 views5 pages

Allen Digital Organ Stop Cards Procedures VI

method for programming IBM cards to create new stops for Allen digital organs, models 1100, 2100, 3100

Uploaded by

kpsteve
Copyright
© Attribution Non-Commercial (BY-NC)
We take content rights seriously. If you suspect this is your content, claim it here.
Available Formats
Download as DOC, PDF, TXT or read online on Scribd
You are on page 1/ 5

Allen Digital Organ Stop Cards Procedures

1. Introduction.
The Allen Digital Organs utilize a simple IBM 80-column card reader to input data for additional voices (stops !e"ond those provided as standard #ith the instrument$ These cards esta!lish values for %& of '( points specif"ing the electronic #aveform to !e replicated at the appropriate fre)uencies *e"ed to esta!lish the pitch and tim!re of the voices$ Although each c"cle of the #aveform is divided into '( e)ual points+ onl" the first %& of those values are input from the ,top -ard.the remaining %& are provided !" internal circuitr" to !e inverted and in reverse order+ completing a #aveform #ith point s"mmetr"$ Man" additional voices are availa!le from the Allen Organ -ompan"+ of course+ !ut sometimes an organist desires or needs a stop that is not e/actl" one of those #hich can !e !ought$ This document addresses the need of those organists to produce digital ,top -ards of their o#n design$ The procedures #ill !e discussed in three main areas0 1e-scaling2 3eneration2 and Ad4ustment$ These procedures reference the attached 5/cel spreadsheet$

2. Re-scaling.
,ometimes the onl" complaint a!out an e/isting ,top -ard voice is that it is either too loud or too soft to #or* properl" #ith the other stops to !e invo*ed #ith it$ If the voice tim!re is satisfactor" !ut the loudness is unsatisfactor"+ all that need !e done is to re-scale the #aveform to have more or less amplitude #ithout altering its harmonic structure$ 6oo* at the ,top -ard$ ,everal things #ill !e apparent0 a$ Along the !ottom of the card #ill !e a pattern of punches in the 8- and 7-ro#s.%& sets of these punch pairs$ These are the 8cloc*ing9 punches and are used !" the input mechanism to discern #hich columns of the card contain data for entr" into the voice memor"$ All cards must have this pattern of 8 and 7 punches$ The 8 punches go in columns :+ %0+ %'+ %&+ %7+ ((+ (;+ (8+ '%+ '<+ ':+ <0+ <'+ <&+ <7+ and ;(2 the 7 punches go in columns &+ 7+ %(+ %;+ %8+ (%+ (<+ (:+ '0+ ''+ '&+ '7+ <(+ <;+ <8+ and ;%$ !$ A!ove the cloc*ing punches is an area containing punches that provide the data for the voice #aveform (%=( c"cle $ This area includes the ro#s 0+ %+ (+ '+ <+ ;+ and & #ith punches in columns &+ 7+ %(+ %;+ %8+ (%+ (<+ (:+ '0+ ''+ '&+ '7+ <(+ <;+ <8+ and ;% (%& specific columns $ The punches in this area are #hat #e are interested in$ >o# reference ,heet ( of the 5/cel spreadsheet 8-onverting Bet#een -olumn Binar" ?ormat and Decimal9$ In one of the data columns+ note the ro#s #ith punches (8%9 values in order from ro# & to ro# 0 (zero $ ?or each of the data columns (even if it has no punches perform the follo#ing prodecure0 a$ Match the pattern of punches to those in the ,heet ( chart and o!tain the corresponding decimal value from the matching cell$ >ote that punch patterns #ith a punch in ro# & (the sign !it #ill have negative values and #ill !e found in the right half of the chart$ !$ 5nter the found decimal value (#ith sign into the "ello# cell in ,heet % ro# '& under the proper card column (la!eled in ro# (( $ Alternativel"+ use the lo#er field on ,heet ' to decode the !inar" values !" using %s and 0s (ones and zeroes to represent the punches and non-punches on the card$ The cells that #ill accept input are colored "ello#$ 5nter the appropriate value (% or 0 into each of the cells !" card ro# and column2 note that zero (809 entries displa" as !lan*s in this field$ The decoded decimal values are sho#n directl" !elo# each card data column+ and are graphed on -hart ; in green$ These values are #hat are to !e transferred to ,heet % ro# '& (as in !$ immediatel" a!ove $ I@m sorr"+ !ut these values have to !e transferred !" hand !ecause those ro# '& cells are not used onl" for this purpose$

Ma*e sure that all the values in the ,heet % column D "ello# cells (ro#s ; through (0 85nter ,trength ?actor9 are 0 (zero + then chec* -hart ;$ The red curve is the graph of the voice #aveform (full c"cle ta*en from the card$ On ,heet %+ place a 8%9 in the salmon-colored cell D'&2 this #ill cause the data entered in the "ello# cells 5'& through T'& (#hich "ou t"ped in to !e used as the input to the scaling formulas instead of the harmonic s"nthesis factors (#hich "ou zeroed out $ -hart ; #ill no# also displa" in !lue the #aveform curve ad4usted to !e ma/imum possi!le amplitude (largest value is &' $ The t#o curves should !e the same shape !ut pro!a!l" #ill not sho# the same amplitude$ >o# enter into the "ello# cell at D(7 a value to !e used to scale do#n the ma/imum amplitude (volume to #hat is desired for the ne# ,top -ard$ This #ill cause the !lue curve on -hart ; to ta*e on a ne# level either a!ove or !elo# the red curve (original values $ Ahen this ad4ustment is appro/imatel" #hat is #anted+ use the red values in ,heet % columns 5(7 through T(7 to punch a ne# ,top -ard !" the procedure descri!ed later in this document$ (The upper field of ,heet ' no# depicts #here the re)uired punches go to achieve this design$

3. Generation.
Another #a" of defining a ne# voice ,top -ard is to specif" the harmonic content desired in the tim!re$ The nature of the card input procedure onl" permits values for the first %& harmonics (fundamental plus %; overtones to !e involved+ !ecause %& data points for a half-c"cle can onl" specif" unam!iguousl" a fre)uenc" of %& times the fundamental or less$ Aaveforms #ith higher harmonics are+ of course+ possi!le+ !ut not !" using the harmonic s"nthesis approach$ Barmonic s"nthesis is also the method used !" the Bammond Organ dra#!ars C and the ,heet % spreadsheet is annotated to indicate ho# those dra#!ars relate to the harmonics2 the standard organ pitch register notations (8@ for the fundamental+ up to %=(@ for the highest harmonic are also included so that additional voices at other than unison pitch can also !e specified$ ?irst insure that the value in the salmon-colored cell at ,heet % D'& is e)ual to 0 (zero 2 this #ill prevent an" values in that ro# from !eing erroneousl" added into the computations to follo#$ >o# onl" the "ello# cells D; through D(0 (to specif" the relative strengths of the different harmonics and D(7 (to ma*e the ad4ustment for loudness #ill have an effect$ 5nter values for the relative strengths of the various harmonics f through %&f into the "ello# cells D; through D(0$ There are t#o #a"s that these values can !e estimated0 a$ B" using Bammond dra#!ar values$ If "ou have a Bammond voice specification that "ou #ant to transcri!e to the Allen Digital Organ+ the dra#!ar settings can !e directl" used to do this using the Barmonic ,"nthesis methodolog"$ On ,heet % of the 5/cel spreadsheet+ note that column - in ro#s ; through %( sho#s B%+ B(+ A%+ B'+ A(+ A'+ (none + and B<.these correspond to the !lac* (8B9 and #hite (8A9 dra#!ars+ from left to right$ The !ro#n dra#!ars cannot !e simulated !" this procedure+ as the" represent su!-harmonics #hich cannot !e encoded$ A dra#!ar all the #a" in represents a factor of 0 (zero 2 pulling the dra#!ar out produces factors of 0$%(;+ 0$(;+ 0$':;+ 0$;+ 0$&(;+ 0$:;+ 0$8:;+ and %$0 (all the #a" out $ Dust enter the corresponding factor values in each dra#!ar ro#$ Once all the dra#!ar position factors are in place+ the resultant #aveform #ill !e displa"ed in !lue on -hart '2 ad4ust the loudness !" entering a value in "ello# cell D(7 on ,heet % (a value of %$0 #ill produce the ma/imum loudness possi!le $ The red values in !lue cells 5(7 through T(7 can then !e used to punch a ne# ,top -ard using the procedure descri!ed later in this document$ !$ Eno#ledge+ 3uessing+ or Ahatever$ ,ome organists ma" have a good understanding of the true harmonic content of various organ stops+ #hereas others ma" onl" have a passing *no#ledge of this topic$ There are a fe# principles #hich can !e applied to ma*ing guesses a!out #hat harmonic factors ma" !e appropriate0 (% Diapasons (Frincipals nearl" all have a ver" strong fundamental in #hatever pitch register is !eing generated+ plus a strong second harmonic (t#ice the fundamental pitch $ 8Flump9 diapasons (5nglish #ill have little additional harmonic structure+ !ut 8string"9 diapasons #ill have more$ In pro!a!l" no case

should an" of those higher harmonics !e factored !" more than a!out 0$( or the diapason nature of the stop #ill !e lost$ A fe# diapasons #ill have a strong 'rd harmonic (' times the fundamental pitch + !ut again this factor #ill not !e ver" large$ (( ?lutes have a strong fundamental in their pitch register #ith ver" small factors for all the higher harmonics$ ,topped flutes #ill have zero (or nearl" zero factors for the 8even9 harmonics ((f+ <f+ &f+ 8f+ etc$ for an 8@ stop2 <f+ 8f+ %(f and %&f for a <@ stop2 and &f and %(f for a ( (='@ stop 2 open flutes #ill have non-zero factors for all harmonics+ !ut the" #ill all !e small$ B" comparison+ a Theater Organ Ti!ia stop #ould have mostl" the fundamental fre)uenc" represented.all other harmonic factors #ould !e near zero$ ,uch a specification #ill produce a nearl" perfect sine #ave for the #aveform and a ver" dull tim!re (the dullest possi!le #ould !e a pure sine #ave.all harmonics G zero e/cept the fundamental $ (' 1eeds #ill have strong contri!utions !" the first several harmonics+ some of #hich ma" even !e more present than the fundamental$ ,pecification of a reed stop is highl" speculative and #ill re)uire some e/perimentation$ Modification of an e/isting reed voice !" the procedures of the ne/t section ma" !e more useful$ (< ,trings cannot !e #ell simulated !" the harmonic s"nthesis process !ecause the" need harmonics #ell a!ove those availa!le in this procedure$ Once a set of harmonic factors have !een entered and a loudness factor decided upon+ the red values in !lue cells 5(7 through T(7 can !e used to punch a ne# ,top -ard using the procedures detailed later in this document$

4. Adjust

ent.

The process of Ad4ustment can !e used to alter an e/isting ,top -ard specification to ma*e minor changes in its tim!re$ This process might !e useful to put a more 8)uint"9 sound to an e/isting flute+ for e/ample$ ?ollo# the procedure for doing a re-scaling of the e/isting stop+ !ut then add non-zero factors for certain of the harmonics using the "ello# cells in D; through D(0$ These factors #ill onl" !e )uite small+ and can !e either positive or negative (add harmonics to ma*e the stop !righter+ su!tract them to ma*e it duller $ The original stop !eing ad4usted #ill !e depicted in -hart ; as the red line2 the ad4usted version #ill !e sho#n !" the !lue line$ This process can !e repeated until the #aveform graph indicates #hat alterations in the tone are desired0 a$ A 8fattening9 of the ma4or lo!e #ill usuall" indicate a 8plumper9 sound2 a 8thinning9 #ill ma*e the sound 8stringier9$ !$ Displacement of the ma4or lo!e #ill indicate that the voice #ill !e more 8open9 (purel" 8stopped9 voices are perfectl" s"mmetrical left to right $ c$ -hanges in the num!er of lo!es (humps indicate more or less of the 8odd9 harmonics (more or less 8)uint"9 $ It is also possi!le to !egin #ith a 8stra# man9 voice specification instead of an actual ,top -ard as the starting point for ma*ing ad4ustments0 a$ The most 8!uzz"9 sound possi!le is a pure sa#tooth #aveform and #ould !e o!tained !" entering the values &(+ ;8+ ;<+ ;0+ <&+ <(+ '8+ '<+ '0+ (&+ ((+ %8+ %<+ %0+ &+ and ( into the -omparison ,top -ard "ello# cells in 5'& through T'& (for an 8@ reed and putting 8%9 into salmon-colored !o/ D'&$ ?or a <@ reed+ enter the values &0+ ;(+ <<+ '&+ (8+ (0+ %(+ <+ -<+ -%(+ -(0+ -(8+ -'&+ -<<+ -;(+ and -&0$ >ote that this set of constants #ill produce a red curve that has ( complete c"cles of sa#tooth+ #hich is #hat #ould !e e/pected of a <@ stop (t#ice the fre)uenc" of an 8@ stop $ Other constant sets can !e found to produce a sa#tooth #ave at higher pitch registers$

!$ If the desire is to produce #ood#ind tones+ a s"mmetrical #aveform such as a s)uare #ave is the proper starting point$ ?or an 8@ s)uare #ave+ enter the values &' into all %& columns2 for a <@ s)uare #ave+ enter &' into the first 8 columns and -&' into the last 82 for a (@ s)uare #ave+ enter &' into the first < columns+ -&' into the ne/t <+ !ac* to &' for < more+ and finish #ith -&' in the last <$ A s)uare #ave unmodified #ill sound li*e a 8zizz"9 o!oe$ Enoc*ing the s)uare corners off !" using smaller values at the transitions #ill ma*e the tone less screech"$ c$ Because %& is not divisi!le !" '+ ;+ or :evenl" it is not possi!le to produce sa#tooth or s)uare #aves at the mutation pitches ( (='@+ % '=;@+ % %='@+ or % %=:@$ Once the chart sho#s #aveform changes that appear desira!le+ use the loudness factor in D(7 to ad4ust the loudness and the procedures later in this document to punch the ne# ,top -ard$

!. Producing t"e ne# Stop Card.


Once a pattern of red values in ,heet % cells 5(7 through T(7 has !een produced+ it is time to transfer these values to the ne# ,top -ard$ This is a relativel" simple process$ 1efer to ,heet ( of the 5/cel spreadsheet$ ?or each of the %& card columns+ perform the follo#ing steps0 a$ 6oo* up the red value for the column in the 8-onverting Bet#een -olumn Binar" ?ormat and Decimal9 ta!les matching it to the 8Decimal Halue9 column of the ta!le$ 1emem!er that negative values are in the right half of the ta!le and positive values are in the left half$ !$ >ote the pattern of punches associated #ith that decimal value$ The 8I9 characters mar* the ro#s to !e punched in that particular column$ It is possi!le that all : ro#s (0 through & could !e punched (a value of -% or none might !e (a value of zero $ All other com!inations of punches and !lan*s are possi!le$ c$ Be sure that the pattern of 8- and 7-ro# punches is also present on the card$ If the" are not there (!" using a pre-punched s*eleton card + "ou #ill also have to punch those$ The" must !e punched e/actl" as found on a standard ,top -ard$ Do not s*ip an" 4ust !ecause that column might not have an" other punches (the cloc*ing punches are #hat allo# that zero value to !e entered $ Alternativel"+ !eginning #ith version IH of this spreadsheet+ there is a ne# ,heet ' that visuall" depicts the IBM card la"out and indicates #hich columns and ro#s are to !e punched !" !lac* mar*s in them$ The upper field of this ,heet represents an IBM card and specificall" depicts the ro# and column #here ever" punch must !e made (#here the mar*s are $ The ,heet ' data is created automaticall" from the values in cells 5(7 through T(7$ There are man" different devices for punching IBM 80-column cards and the" all operate some#hat differentl"$ This document does not attempt to cover all the different #a"s that multiple punches in a single column might !e done.it #ill !e necessar" for the organist to learn ho# to do this #ith the machine availa!le$ There #ill usuall" !e a !utton called a 8multiple punch9 !utton+ and this must !e depressed and *ept depressed #hile all punches in that column are made+ then released (at #hich time the card #ill !e advanced one position $ ,ome practice in producing correctl"-punched cards #ill pro!a!l" !e re)uired$

$. A%ter#ord.
I have tried to ma*e this process as simple and straightfor#ard as possi!le$ The original document I produced "ears ago #as ver" arithmetic-intensive and re)uired a lot of manual calculator activit"$ In this more modern version I have incorporated the capa!ilities of the 5/cel spreadsheet facilit" to significantl" reduce the amount of arithmetic re)uired and to produce graphical displa"s of the resultant #aveforms$ I hope that the procedures descri!ed a!ove are sufficientl" detailed and uncomplicated to !e useful$ If "ou have an" )uestions+ please direct them to me$ I do recommend a lot of 8pla"ing around9 #ith the various features

of the spreadsheet #ith chec*ing of the -hart ; displa" each time to get a feel for ho# the various control entries interact$ Bere are three overall considerations0 a$ ?or a simple re-scaling of an e/isting stop+ set all the ,trength ?actor "ello# cells (D; through D(0 on ,heet % to 0 (zero + set the salmon-colored !o/ D'& to %$0+ and enter the data from the e/isting card into "ello# cells 5'& through T'&$ The values to !e entered into these cells can !e o!tained either !" using ,heet ( to manuall" convert the !inar" codes or !" using the ne# field of ,heet '$ -hart ; #ill sho# the original ,top -ard curve in red and the loudness-ad4usted (!" value in D(7 curve in !lue$ !$ Ahen using the ,"nthesis process+ either using Bammond dra#!ar C settings or !" *no#ledge of the desired harmonic content+ use the ,trength ?actor "ello# cells (D; through D(0 for input and set the salmon-colored !o/ D'& to 0 (zero $ Although this prevents the entries in "ello# cells 5'& through T'& from having an effect on the resultant #aveform+ values in those columns #ill produce a red curve in -hart ;+ and this might !e confusing$ I recommend that those "ello# cells !e set to 0 (zero also$ c$ Jsing the ,"nthesis process to modif" an e/isting (or 8stra# man9 #aveform+ all the "ello# cells ma" have values in them and the salmon-colored !o/ #ill !e set to %$0$ The use of the distinct processes together simplifies ma*ing minor changes in tim!re to e/isting ,top -ard voices$

C"ange &og.
Hersion III corrected an error in the harmonic-s"nthesis procedure that erroneousl" e/cluded the upper half of the availa!le harmonics and mis-la!eled the K@ line as %@$ -harts < and ; #ere also added as improvements of -harts % and '+ respectivel" (end-c"cle zero-crossings handled $ The references in this document have !een changed to reflect the added ro#s re)uired for the fi/ to the harmonics$ This document should not !e used for earlier versions$ Hersion IH incorporated the ,heet ' capa!ilit" of displa"ing #hat the punched IBM card #ould loo* li*e$ It also added cell protection to prevent inadvertent destruction of necessar" values !" over#riting$ Hersion H 4ust made some minor cosmetic changes to Hersion IH$ Hersion HI added a ne# card-image field to ,heet ' !elo# the original one$ This ne# field allo#s entr" of zeroes or ones into the "ello# cells to replicate #here holes are punched in ro#s 0 through & on the Tone -ard$ Ahen this has !een done+ the larger cells !elo# each data column sho# the computed decimal values of the !inar" data that had !een encoded on the card$ These values are also plotted on -hart ; as a green line$ (If this is confusing+ t"pe zeroes into all the "ello# cells on ,heet ' and the green line #ill then lie on the /-a/is and !e out of the #a"$

You might also like

pFad - Phonifier reborn

Pfad - The Proxy pFad of © 2024 Garber Painting. All rights reserved.

Note: This service is not intended for secure transactions such as banking, social media, email, or purchasing. Use at your own risk. We assume no liability whatsoever for broken pages.


Alternative Proxies:

Alternative Proxy

pFad Proxy

pFad v3 Proxy

pFad v4 Proxy