Replaced the period table by byte-deltas, saved 36 bytes and compression is even better!

README.md

@@ -1,7 +1,7 @@
 # Raspberry Casket
 A fast and small open source Pretracker replayer
 
-## Raspberry Casket Player V2.x (26-Jul-2023)
+## Raspberry Casket Player V2.x (10-Aug-2023)
 
 Provided by Chris 'platon42' Hodges <chrisly@platon42.de>
 
@@ -90,7 +90,7 @@ The original code compressed with *Blueberry's* Shrinkler goes from
 18052 bytes down to 9023 bytes.
 
 Raspberry Casket, depending on the features compiled in, is about
-5992 bytes and goes down to ~4238 bytes (in isolation).
+5942 bytes and goes down to ~4202 bytes (in isolation).
 
 So this means that the optimization is not just "on the outside".
 
@@ -144,8 +144,9 @@ solve this problem.
 - Split wave generation out of main file, reorganised content into header files.
 - Optimized some more code paths for Raspberry Casket replayer.
 - In the wave generator optimized away a table (32 words), replacement code is even smaller!
+- Replaced the period table by byte-deltas, saved 36 bytes and compression is even better!
 - Added Presto player draft.
-- Drop-in replacement code size: 5992 bytes.
+- Drop-in replacement code size: 5942 bytes.
 
 ### V1.x (unreleased)
 - Fixed a bug regarding the copper output mode with looping waves having a loop-offset.

src/raspberry_casket.asm

@@ -1,5 +1,5 @@
 ;--------------------------------------------------------------------
-; Raspberry Casket Player V2.x (26-Jul-2023)
+; Raspberry Casket Player V2.x (10-Aug-2023)
 ; ~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~
 ;
 ; Provided by Chris 'platon42' Hodges <chrisly@platon42.de>
@@ -87,7 +87,7 @@
 ; 18052 bytes down to 9023 bytes.
 ;
 ; Raspberry Casket, depending on the features compiled in, is about
-; 5992 bytes and goes down to ~4238 bytes (in isolation).
+; 5942 bytes and goes down to ~4202 bytes (in isolation).
 ;
 ; So this means that the optimization is not just "on the outside".
 ;
@@ -162,7 +162,11 @@ PRETRACKER_PARANOIA_MODE = 0
 ; I guess this might add some phasing to the sound (which would be
 ; only audible in a mono mix, I would think), but it comes with an
 ; extra multiplication per channel.
-; If you really want that, enable this switch.
+; Moreover, I traced down that this may reduce the Amiga period
+; value lower than the minimum allowed value of 124. This may cause
+; distortions (like playing a note in Protracker with pitch B-3)
+; on the delayed track.
+; If you really want (to risk) that, enable this switch.
         IFND    PRETRACKER_DUBIOUS_PITCH_SHIFT_FOR_DELAYED_TRACK
 PRETRACKER_DUBIOUS_PITCH_SHIFT_FOR_DELAYED_TRACK = 0
         ENDC
@@ -590,28 +594,28 @@ pre_PlayerInit:
 
 ; ----------------------------------------
 
-        lea     pre_period_table(pc),a0
+        lea     pre_delta_period_table(pc),a0
         lea     pv_period_table(a4),a1
 
 .calcperiodtable
         ; fill the missing entries in the period table by interpolating
+        move.w  (a0)+,d2
         moveq.l #3*NOTES_IN_OCTAVE-1,d7
 .periodtableloop
-        move.w  (a0)+,d0
+        moveq.l #0,d0
-        move.w  (a0),d1
+        moveq.l #0,d1
-        sub.w   d0,d1
+        move.b  (a0)+,d1
-        swap    d0
+        move.w  d2,d0
-        clr.w   d0
+        sub.w   d1,d2
         swap    d1
-        clr.w   d1
+        lsr.l   #4,d1
-        asr.l   #4,d1
 
         moveq.l #16-1,d6
 .perfineipolloop
-        swap    d0
         move.w  d0,(a1)+
         swap    d0
-        add.l   d1,d0
+        sub.l   d1,d0
+        swap    d0
         dbra    d6,.perfineipolloop
         dbra    d7,.periodtableloop
 
@@ -791,7 +795,7 @@ pre_PlayerTick:
         move.l  d1,(a1)+    ; pcd_pat_vol_ramp_speed_b, pcd_pat_2nd_inst_num4_b, pcd_pat_2nd_inst_delay_b, pcd_wave_offset_b
         move.l  d1,(a1)+    ; pcd_inst_pitch_slide_w and pcd_inst_sel_arp_note_w
         move.w  d1,(a1)+    ; pcd_inst_note_pitch_w
-        addq.w  #2,a1
+        addq.l  #2,a1
 
         move.l  d1,(a1)+    ; pcd_inst_line_ticks_b, pcd_inst_pitch_pinned_b, pcd_inst_vol_slide_b, pcd_inst_step_pos_b
 
@@ -2024,14 +2028,14 @@ pre_PlayerTick:
         bra.s   .loop_handling_done
 
 .wave_has_no_subloop
+        moveq.l #0,d1
         move.b  pcd_wave_offset_b(a5),d1
         beq.s   .loop_handling_done
         tst.b   wi_allow_9xx_b(a3)
         beq.s   .loop_handling_done
 
         ; apply offset from pattern for sample without subloop
-        lsl.w   #8,d1   ; clear upper byte
+        lsl.w   #7,d1
-        lsr.w   #1,d1
 
         clr.b   pcd_wave_offset_b(a5)
 
@@ -2491,7 +2495,7 @@ pre_PlayerTick:
         rts
 
 ;--------------------------------------------------------------------
-; table data currently about 594 bytes
+; table data currently about 496 bytes
         ; Tables used by WaveGen
 pre_roll_off_table: ; used by WaveGen
         dc.w    $400,$200,$180,$140,$100,$C0,$A0,$80,$78,$74,$6E
@@ -2510,7 +2514,7 @@ pre_log12_table:
 
 pre_modulator_ramp_8: ; used by WaveGen
         ;dc.w     77,293,539,1079,1337,1877,2431,3031 ; the 1079 value is strange (938 better?)
-        dc.w     $4D,$125,$21B,$437,$539,$755,$96D,$BD7
+        dc.w    $4D,$125,$21B,$437,$539,$755,$96D,$BD7
 
 ; linear then steep quadratic slope
 pre_vib_speed_table:
@@ -2566,11 +2570,11 @@ pre_octave_select_table:
         IFNE     PRETRACKER_DUBIOUS_PITCH_SHIFT_FOR_DELAYED_TRACK
         ; -4,-3,-1,1,2,3,4,0
 pre_minus4plus4_table:
-        dc.b     $FC,$FB,$FF,1,2,3,4,0
+        dc.b    $FC,$FB,$FF,1,2,3,4,0
         ENDC
 
-pre_period_table:
+pre_delta_period_table:
-        dc.w     $350,$320,$2F2,$2C8,$2A0,$279,$256,$236,$216,$1F8,$1DC,$1C0
+        dc.w    $350
-        dc.w     $1A8,$190,$179,$164,$151,$13E,$12C,$11B,$10B,$FC,$EE,$E0
+        dc.b    $30,$2e,$2a,$28,$27,$23,$20,$20,$1e,$1c,$1c,$18,$18,$17,$15,$13
-        dc.w     $D4,$C8,$BD,$B2,$A8,$9F,$96,$8D,$86,$7E,$78,$71
+        dc.b    $13,$12,$11,$10,$0f,$0e,$0e,$0c,$0c,$0b,$0b,$0a,$09,$09,$09,$07
-        dc.w     $71
+        dc.b    $08,$06,$07,$00

src/raspberry_casket.i

@@ -61,7 +61,6 @@
 ;   - 6 Bits : Pitch ($01 is C-0)
 ;   - 4 Bits : unused?
 ;   - 12 Bits: Effect
-; - Patterns x steps x 3 bytes
 ;
 ; Wave definitions (WAVE):
 ; - 24 (!) x Null terminated string (or 23 chars max)