video_modes.txt
41.3 KB
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
61
62
63
64
65
66
67
68
69
70
71
72
73
74
75
76
77
78
79
80
81
82
83
84
85
86
87
88
89
90
91
92
93
94
95
96
97
98
99
100
101
102
103
104
105
106
107
108
109
110
111
112
113
114
115
116
117
118
119
120
121
122
123
124
125
126
127
128
129
130
131
132
133
134
135
136
137
138
139
140
141
142
143
144
145
146
147
148
149
150
151
152
153
154
155
156
157
158
159
160
161
162
163
164
165
166
167
168
169
170
171
172
173
174
175
176
177
178
179
180
181
182
183
184
185
186
187
188
189
190
191
192
193
194
195
196
197
198
199
200
201
202
203
204
205
206
207
208
209
210
211
212
213
214
215
216
217
218
219
220
221
222
223
224
225
226
227
228
229
230
231
232
233
234
235
236
237
238
239
240
241
242
243
244
245
246
247
248
249
250
251
252
253
254
255
256
257
258
259
260
261
262
263
264
265
266
267
268
269
270
271
272
273
274
275
276
277
278
279
280
281
282
283
284
285
286
287
288
289
290
291
292
293
294
295
296
297
298
299
300
301
302
303
304
305
306
307
308
309
310
311
312
313
314
315
316
317
318
319
320
321
322
323
324
325
326
327
328
329
330
331
332
333
334
335
336
337
338
339
340
341
342
343
344
345
346
347
348
349
350
351
352
353
354
355
356
357
358
359
360
361
362
363
364
365
366
367
368
369
370
371
372
373
374
375
376
377
378
379
380
381
382
383
384
385
386
387
388
389
390
391
392
393
394
395
396
397
398
399
400
401
402
403
404
405
406
407
408
409
410
411
412
413
414
415
416
417
418
419
420
421
422
423
424
425
426
427
428
429
430
431
432
433
434
435
436
437
438
439
440
441
442
443
444
445
446
447
448
449
450
451
452
453
454
455
456
457
458
459
460
461
462
463
464
465
466
467
468
469
470
471
472
473
474
475
476
477
478
479
480
481
482
483
484
485
486
487
488
489
490
491
492
493
494
495
496
497
498
499
500
501
502
503
504
505
506
507
508
509
510
511
512
513
514
515
516
517
518
519
520
521
522
523
524
525
526
527
528
529
530
531
532
533
534
535
536
537
538
539
540
541
542
543
544
545
546
547
548
549
550
551
552
553
554
555
556
557
558
559
560
561
562
563
564
565
566
567
568
569
570
571
572
573
574
575
576
577
578
579
580
581
582
583
584
585
586
587
588
589
590
591
592
593
594
595
596
597
598
599
600
601
602
603
604
605
606
607
608
609
610
611
612
613
614
615
616
617
618
619
620
621
622
623
624
625
626
627
628
629
630
631
632
633
634
635
636
637
638
639
640
641
642
643
644
645
646
647
648
649
650
651
652
653
654
655
656
657
658
659
660
661
662
663
664
665
666
667
668
669
670
671
672
673
674
675
676
677
678
679
680
681
682
683
684
685
686
687
688
689
690
691
692
693
694
695
696
697
698
699
700
701
702
703
704
705
706
707
708
709
710
711
712
713
714
715
716
717
718
719
720
721
722
723
724
725
726
727
728
729
730
731
732
733
734
735
736
737
738
739
740
741
742
743
744
745
746
747
748
749
750
751
752
753
754
755
756
757
758
759
760
761
762
763
764
765
766
767
768
769
770
771
772
773
774
775
776
777
778
779
780
781
782
783
784
785
786
787
788
789
790
791
792
793
794
795
796
797
798
799
800
801
802
803
804
805
806
807
808
809
810
811
812
813
814
815
816
817
818
819
820
821
822
823
824
825
826
827
828
829
830
831
832
833
834
835
836
837
838
839
840
841
842
843
844
845
846
847
848
849
850
851
852
853
854
855
856
857
858
859
860
861
862
863
864
865
866
867
868
869
870
871
872
873
874
875
876
877
878
879
880
881
882
883
884
885
886
887
888
889
890
891
892
893
894
895
896
897
898
899
900
901
902
903
904
905
906
907
908
909
910
911
912
913
914
915
916
917
918
919
920
921
922
923
924
925
926
927
928
929
930
931
932
933
934
935
936
937
938
939
940
941
942
943
944
945
946
947
948
949
950
951
952
953
954
955
956
957
958
959
960
961
962
963
964
965
966
967
968
969
970
971
972
973
974
975
976
977
978
979
980
981
982
983
984
985
986
987
988
989
990
991
992
993
994
995
996
997
998
999
1000
1001
1002
1003
1004
1005
1006
1007
1008
1009
1010
1011
1012
1013
1014
1015
1016
1017
1018
1019
1020
1021
1022
1023
1024
1025
1026
1027
1028
1029
1030
1031
1032
1033
1034
1035
1036
1037
1038
1039
1040
1041
1042
1043
1044
1045
1046
1047
1048
1049
1050
1051
1052
1053
1054
1055
1056
1057
1058
1059
1060
1061
1062
1063
1064
1065
1066
1067
1068
1069
video_modes.txt
- Phil Gossett
6/27/95
Overview
-----------
The video interface performs the second pass of the antialiased z-buffer
algorithm used in the rdp. It also generates the video sync pulses for
NTSC or PAL in both interlaced and non-interlaced modes, for both 16 and
32 bit color pixels, at either high or low resolution, with or without
video filtering for deflickering interlaced video. It can optionally also
rescale the image in order to sacrifice resolution for rendering speed,
making it possible to load balance the per-pixel rendering time.
This document first describes in some detail the functionality of the
antialiasing parts of the video interface. It then describes the
video interface control registers, and their use for the 28 supported
video display modes.
Note that this document only describes the "normal" modes, supported in
the gbi.h macros. For NTSC and PAL, these should suffice for almost any
imaginable circumstance. However, the sync generation circuitry is more
general than it needs to be, and could be programmed for other video
formats (provided the hardware has sufficient bandwidth).
Antialiasing
---------------
The bulk of the video interface is devoted to the second pass of
the antialiased z-buffer algorithm. In the point sampling modes,
this circuitry is effectively bypassed.
The main goal of this algorithm is to produce an antialiased rendering
of polygonal surfaces without the need for sorting. The key to
achieving this goal is to split the antialiasing problem up into
several pieces, each of which is readily implemented.
There are basically three different kinds of antialiasing. The first
is the antialiasing of textures within polygons. This is accomplished
by the texture hardware, using the industry standard mipmapping
technique. This uses tri-linear interpolation to produce a correctly
sampled texture lookup. (The details of this are beyond the scope of
this document.)
The second kind of antialiasing is the blending of poly fragments
within the pixels they share. The classic example of this is the
pinwheel, where alternating black and white triangles meet at a center
vertex. The pixel within which this vertex lies should be the average
of the colors of all the triangles which share this vertex, weighted
by the area of the pixel at the vertex covered by each of the
triangles. (The details of this are also beyond the scope of this
document.)
The third kind of antialiasing is the blending of the silhouette of a
foreground object against the background. This is traditionally done
at rendering time in the blend unit. Unfortunately, doing it at this
time has bad consequences for z-buffering.
Consider an internal edge of a surface (i.e., an edge shared by two
visible polys not at the silhouette). A priori, when the first of the
two polys is rendered, the blender does not yet know whether it is a
silhouette edge (and hence needs to be blended with the background),
or an internal edge (and hance should not be blended with the
background). Note that if an internal edge does blend with the
background, there will be a line along the edge left when the second
poly blends with the first. Once the blending is done, there is no way
to undo it. Also, note that the background may not even have been
rendered yet, unless the rendering of polys is done in depth-sorted
order, which defeats the purpose of z-buffering.
The only way to deal with this is to postpone the blending of
silhouette edges until after the whole scene is rendered. In fact, the
final blending of the silhouette edges is done at display time by the
video interface. To do this blend on video output, there needs to be
a coverage value left behind in the frame buffer, with which to
interpolate between the foreground (the color in the frame buffer) and
the background (which is assumed to be in one or more of the
neighboring pixels in the frame buffer). This interpolation is as
follows:
output color = cvg*foreground + (1-cvg)*background.
The foreground is always the color stored in the frame buffer for that
pixel. The problem is to find a representive of the background (which
was overwritten for this pixel during the rendering process by the
foreground pixel) amoung the neighbors of the pixel. (For now, let's
ignore the precise definition of which pixels are the neighbors. We'll
get back to this later.) If we looked at the z-buffer (assuming the
image was rendered with z-buffering turned on), we might be able to
use this to distinguish between foreground and background pixels,
perhaps using the neighboring pixel with the largest z (most distant)
as the representative of the background. However, if we did this, the
video interface would have to read the z-buffer, which would add to
the memory bandwidth required. Even worse, this would require double
buffering of z, since if it's being used it display time, it can't be
allowed to be stepped by the rendering of the next frame. Since we
have so little frame buffer memory, this would be prohibitive. So
again, the problem is to find a representive for the background
without using depth information, using only the color and coverage
values for each pixel and its neighbors.
Surprisingly enough, this can be done to a reasonable approximation,
assuming relative constancy of color within neighboring pixels of the
foreground and relative constancy of color within neighboring pixels
of the background. Note that this must be assumed for the background,
since the true background for a foreground pixel has already been
overwritten. Errors in this assumption tend to not be noticable, since
they are consistent with "edge effects" (diffraction, etc.). For a
well-known example of the human visual system's tolerance for this
kind of artifact, note that in the movie "2001: A Space Odyssey", due
to the difficulties of the matting process (mostly done by hand), the
stars in the background of a spaceship would disappear somewhat before
they should actually have been occluded. (It would have been much
worse for them to have disappeared after they should have been
occluded). Despite many very picky people (myself, the son of a
professional cinematographer, included) looking for errors in the
special effects (and there were a few that were noticable), no one
noticed this error. (It was written up by some of the people involved
several years after the release of the film, which is why I know about
it.) Note also that this assumption of color constancy tends to be
true for foreground objects, even for textured objects, since by
mipmapping, the pixels near the silhouette of an object (which are the
only pixels altered by the display time antialiasing in the video
interface) tend to be free of high spatial frequencies tangentially as
well as radially (since mipmapping gives no directionality to the
level of detail). But I digress...
What we are trying to find is some estimate for the color of the lost
(overwritten) background pixel. First, we look at the neighbors of the
central pixel currently being displayed. Once again, we'll defer
discussion of the details of the neighborhood (including how many
neighbors there are) until later. In any case, however many there are,
we sort them (independently by RGB color component), reject the
maximum and minimum values in each component, and take the remaining
(penultimate) maximum and minimum values for each component.
(Actually, before the sort of the neighbors we reject any partially
covered pixels, since they are bound to be part of the foreground, for
which we already have a good representative.) We then take the maximum
of the central pixel (known to be the foreground) and the penultimate
maximum of the neighbors, as well as the minimum of the central pixel
and the penultimate minimum of neighbors. Note that at this point, one
of the maximum and minimum will be representative of the foreground,
and the other will be representative of the background (independently,
color component by color component). But we do not yet know which is
which.
We then average these component maxima and minima. This gives us a
color half way between the foreground and background. We then take the
difference of this color with the central pixel's (foreground)
color. This is half the magnitude of the difference between the
foreground and background colors. So we double this difference and add
it to the central (foreground) pixel. This is our estimate of the
background pixel color, which was overwritten by the foreground pixel
color. We then interpolate between the foreground and background
colors by the coverage value stored in the frame buffer along with the
(foreground) color. This interpolated result is our antialiased pixel.
The careful observer may have noticed that in the 16 bit case, there
aren't enough bits for RGB and coverage. In this pixel size, there are
5 bits each of red, green and blue. We also need at least 3 bits of
coverage. This adds up to 18 bits, not 16. (32 bit mode, of course, is
not a problem, having 8 bits free for coverage.) Since the DRAMs we
are using are really 9 bits, and we have no use for parity (where do
you dump core?) we can use the ninth bit to get 18 bits in 2 bytes.
(We also use these "hidden" bits to pack 18 bits into the z buffer.)
Note that all the operations done here (maximum, minimum, averaging,
etc.) are continuous and monotonic (although nonlinear). These
properties insure that there isn't excessive "popping" of whatever
artifacts there may be as objects move across the pixel grid. Note
that the maximum function can be viewed as one member of a family of
two variable functions, of which the sum and the Pythagorean distance
are the first two members. If such functions are written as:
z^n = x^n + y^n,
then in the limit as n approaches infinity:
z = max(x, y).
Let's return to the issue of neighborhoods. This algorithm proved to
be remarkably sensitive to the choice of a neighborhood. In it's
development, I tried many different neighborhoods, from 2 by 2 with 1
to 3 neighbors, to 3 by 3 with 4 or 8 neighbors, to some pretty
obscure ones. None of the obvious choices worked out very well. There
were two reasons for this. In the small neighborhoods (4 neighbors or
less), there are not enough samples to be able to insure that there
will always be a background pixel represented within the neighborhood.
On the other hand, in the more highly connected neighborhoods (such as
3 by 3 with 8 neighbors), the fact that some (orthogonal) neighbors
are significantly closer than the other (diagonal) neighbors
introduces a bias, creating unacceptable artifacts.
The ideal case would have been a triangular grid, with each pixel
having 6 neighbors (a nice, healthy number), all equidistant (so no
bias). But we are rendering on a square (or at least rectangular)
grid. (I don't want to think about rendering to a triangular grid...)
So what can we do?
A triangular grid can be approximated by "checkerboarding" the pixels.
If you only consider the neighboring squares on a checkerboard that
are of the same color, you can get a slightly deformed triangular
grid. (This deformation turns out to not be problematic, since
anything that correctly antialiases something on one rectangular grid
should also work on another rectangular grid with a different aspect
ratio.) So the neighborhood we use is as follows:
0 X 0 X 0
X 0 * 0 X
0 X 0 X 0,
where * is the central pixel, the X's are the neighbors, and the 0's
are unused. Note that in the case of interlace, since we are only
rendering one field at a time, the rows of neighbors above and below
are actually from 2 lines away. Note that this simply deforms the
triangular grid the other way, and so is not really any different.
Before final processing, one more operation is done (optionally) to
the antialiased pixels, to fix up some slight artifacts of the
algorithm. For any pixels on or adjacent to a silhouette edge, the
divot cicuit takes the median of three adjacent pixels as the color to
be displayed in place of the center pixel. For any single silhouette,
this will always leave the pixels essentially undisturbed, while
filtering out the "divots" (one pixel notches) due to multiple
overlapping silhouetts. While not perfect, this simple circuit
effectively eliminates most of the divot artifacts. Note that this
median filter is not used in regions of full coverage, so high spatial
frequency textures within a surface will not be disturbed by the divot
circuit.
The processed pixels are then rescaled (optionally), gamma corrected
(optionally), and dithered (optionally) to 7 bits per RGB component,
which together with the sync pulses are output to the video DACs.
Display Modes
----------------
The 28 modes supported in the gbi.h macros could have been represented
by 5 binary switches (hi/lo res, 16/32 bit, antialiased/point-sampled,
filtered/not, NTSC/PAL), but 4 of the 32 switch combinations are illegal
due to lack of memory bandwidth. The disallowed modes are all the hi res,
32 bit, antialiased modes. Note that the hi res, 32 bit, point-sampled
modes are allowed.
The gbi.h macros have override flags to inactivate the following features:
gamma_dither_enable - for intentional Mach banding ("special effects")
gamma_enable - for MPEG (which is already gamma corrected)
divot_enable - for decal lines (which interract poorly with divot)
In addition, in all modes, there is an x scale factor to scale the image
up horizontally to the required display format from a reduced number of
rendered pixels per line. In the low-res, non-interlaced modes, there can
also be a y scale factor to scale the image up vertically to the required
display format from a reduced number of rendered lines per frame. The
intended use for this is load balancing in cases where reducing the frame
rate would be unaccetable. The most likely case is running into an object
(eg., a wall) where the time of impact is important to the "feel" more
than maintaining the full resolution. Note that this rescaling is linear,
so there will be some artifacts from the interpolation, such as alternating
bands of blurring where the interpolation factor is near 50% versus bands of
sharpness where the interpolation factor is near 0% or 100%. In particular,
this feature should probably not be used to rescale between NTSC and PAL
pixel aspect ratios. (There are approximately the same numbers of pixels
across in both NTSC and PAL, but PAL has more lines per frame. However, the
frame aspect ratios between NTSC and PAL are the same. So PAL pixels are
"shorter" than NTSC pixels. Usually, as in the SNES, this difference is
ignored, since the ratio is 480 to 525, which is off by about 9%, well
within the adjustment range of most TV monitors.)
Video Interface Control Registers
------------------------------------
There are 16 control registers for the video interface which control all
its functions including sync generation, video rescaling, and antialiasing.
These video interface registers are as follows:
0 : ctrl[16:0] - see below
1 : origin[23:0] - frame buffer origin in bytes
2 : width[11:0] - frame buffer line width in pixels
3 : v_int[9:0] - interrupt when current half-line = v_int
4 : v_current[9:0] - current half line, sampled once per line
(the lsb of v_current is constant within
a field, and in interlaced modes gives the
field number - this is constant for non-
interlaced modes)
5 : burst_start[29:20] - beginning of color burst in pixels from hsync
vsync_width[19:16] - width if vsync in half-lines
burst_width[15:8] - width of color burst in pixels
hsync_width[7:0] - width of hsync in pixels
6 : v_sync_period[9:0] - number of half-lines per field
7 : leap_pattern[20:16] - a 5 bit pattern used for PAL only
h_sync_period[11:0] - the total duration of a line in 1/4 pixels
8 : hsync_leap_a[27:16] - identical to h_sync_period (except for PAL)
hsync_leap_b[11:0] - identical to h_sync_period (except for PAL)
9 : h_video_start[25:16] - start of active video in screen pixels
h_video_end[9:0] - end of active video in pixels from hsync
10 : v_video_start[25:16] - start of active video in screen half-lines
v_video_end[9:0] - end of active video in half-lines from vsync
11 : v_burst_start[25:16] - start of color burst enable in half-lines
v_burst_end[9:0] - end of color burst enable in half-lines
12 : x_offset[27:16] - horizontal subpixel offset (2.10 format)
x_scale[11:0] - 1/horizontal scale up factor (2.10 format)
13 : y_offset[27:16] - vertical subpixel offset (2.10 format)
y_scale[11:0] - 1/vertical scale up factor (2.10 format)
14 : test_addr[6:0] - for diagnostics, not used in normal operation
15 : staged_data[31:0] - for diagnostics, not used in normal operation
Note that all the numerical parameters count from zero. So a count of zero
means one unit is to be counted. In general, the actual count is one plus
the count for any of the control register parameters.
Most of the video registers should not be changed at any time except during
vertical blanking. The registers which are exceptions to this are h_video_start,
h_video_end, v_int and v_current
The power on default settings for the video registers are:
ctrl = 0; // blank video (incl syncs)
origin = 0;
width = 0;
v_int = 0x3FF;
burst_start = 0;
vsync_width = 0;
burst_width = 0;
hsync_width = 0xd1; // for refresh
v_sync_period = 0;
leap_pattern = 0;
h_sync_period = 0xd2047; // for refresh
hsync_leap_b = 0;
hsync_leap_a = 0;
h_video_end = 0;
h_video_start = 0;
v_video_end = 0;
v_video_start = 0;
v_burst_end = 0;
v_burst_start = 0;
x_scale = 0;
y_scale = 0;
x_offset = 0;
y_offset = 0;
test_addr = 0;
staged_data = 0;
The following are the definitions (positive true) of the mode bits in the
video interface control register:
ctrl[1:0] = type[1:0] (pixel size) == 0: blank (no data, no sync)
1: reserved
2: 5/5/5/3 ("16" bit - really 18 bit)
3: 8/8/8/8 (32 bit)
ctrl[2] = gamma_dither_enable (normally on, unless "special effect")
ctrl[3] = gamma_enable (normally on, unless MPEG/JPEG)
ctrl[4] = divot_enable (normally on if antialiased, unless decal lines)
ctrl[5] = vbus_clock_enable (never on - hardware damage may result)
ctrl[6] = serrate (always on if interlaced, off if not)
ctrl[7] = test_mode (for diagnostics, not used in normal operation)
ctrl[9:8] = aa_mode[1:0] == 0: aa & resamp (always fetch extra lines)
1: aa & resamp (fetch extra lines if needed)
2: resamp only (treat as all fully covered)
3: neither (replicate pixels, no interpolate)
ctrl[11] = kill_we (for diagnostics, not used in normal operation)
ctrl[15:12] = pixel_advance (always 3 for optimal operation)
ctrl[16] = dither_filter_enable (normally on for 16 bit, off for 32 bit)
Type is just the pixel size used by the video interface, with a code
for shutting off the video altogether (including the syncs. TV
mointors will display snow or a blank frame.) Gamma_dither_enable adds
in some random noise to the lsbs of the video out before the final
quantization to 7 bits. This is to eliminate Mach banding artifacts.
In general, this should always be on, unless deliberate artifacts are
desired as a "special effect". Gamma_enable turns on the correction
for the gamma non-linearity of TV monitors. This amounts to taking a
square root of the linear color space used within the rendering
pipeline. The TV monitor (according to spec) effectively raises these
color values to the 2.2 to 2.4 power, leaving a residual gamma of 1.1
to 1.2. This error is actually desirable, since a slightly higher than
unity effective gamma produces a more color accurate image in the
typical somewhat darkened TV viewing environment. Note that for MPEG
and JPEG images, this gamma correction is already in the image data,
so the hardware gamma correction should be turned off.
Divot_enable is a part of the antialiasing alogrithm which fixes up
some slight artifacts of the algorithm. For any pixels on or adjacent
to a silhouette edge, the divot cicuit takes the median of three
adjacent pixels as the color to be displayed in place of the center
pixel. This eliminates the "divots" (one pixel notches) seen in the
horizon of some fractal terrains, due to the antialiasing algorithm's
problems with multiple fragments (silhouette edges) per pixel. Note
that this median filter is not used in regions of full coverage (the
surfaces of objects), so any texture detail will not be disturbed by
the divot circuit. Note also that this feature interacts poorly with
one of the more obscure rendering modes (decal line), since this
rendering mode can produce details (lines at the silhouette edge)
which the divot circuit could affect. So this feature can be turned
off (or the effect can simply be tolerated in cases where the divot
defects are more visually annoying than the decal line defects).
Dither_filter_enable turns on a filter which largely undoes the effect
of dithering, restoring a smooth image. This would normally be used
with 16 bit pixels (with their 5 bits per component) with the assumption
that these would be dithered in the rdp. For 32 bit mode, this would
normally be turned off
Vbus_clock_enable must never be turned on. This is a hardware feature
for cases where the clock distribution is other than it is in the
actual Ultra 64. Turning this mode bit on in the current hardware will
result in two output drivers both being enabled on the same wire at
the same time, which can result in permanent damage to the device.
Note that SetImage rdp command must be consistent with the video
display modes for correct operation. In particular, the pixel size
must be the same and the image origin must be consistent (although
with an offset in some modes). The SetScissor rdp command should also
be consistent with the video display mode for optimal rendering
efficiency. In particular, in the hi-res, normal interlace modes,
scissoring the even or odd field out can be used to double buffer
without the need for additional frame buffer memory. One field can be
rendered while the other field is being displayed. The rendering
hardware can skip over the unneeded lines of the scissored out field
to prevent the currently displaying field from being stepped on by the
currently rendering field.
The 28 Supported Video Modes
-------------------------------
video format
| resolution
| | pixel size
| | | antialiasing(aa)/point-sampling(ps)
| | | | interlace
mode | | | | |
--------- - - - - -
vi_ntsc_lpn1 ntsc lo 16 ps non-interlace
vi_ntsc_lpf1 ntsc lo 16 ps interlace
vi_ntsc_lan1 ntsc lo 16 aa non-interlace
vi_ntsc_laf1 ntsc lo 16 aa interlace
vi_ntsc_lpn2 ntsc lo 32 ps non-interlace
vi_ntsc_lpf2 ntsc lo 32 ps interlace
vi_ntsc_lan2 ntsc lo 32 aa non-interlace
vi_ntsc_laf2 ntsc lo 32 aa interlace
vi_ntsc_hpn1 ntsc hi 16 ps normal interlace
vi_ntsc_hpf1 ntsc hi 16 ps deflickered interlace
vi_ntsc_han1 ntsc hi 16 aa normal interlace
vi_ntsc_haf1 ntsc hi 16 aa deflickered interlace
vi_ntsc_hpn2 ntsc hi 32 ps normal interlace
vi_ntsc_hpf2 ntsc hi 32 ps deflickered interlace
vi_pal_lpn1 pal lo 16 ps non-interlace
vi_pal_lpf1 pal lo 16 ps interlace
vi_pal_lan1 pal lo 16 aa non-interlace
vi_pal_laf1 pal lo 16 aa interlace
vi_pal_lpn2 pal lo 32 ps non-interlace
vi_pal_lpf2 pal lo 32 ps interlace
vi_pal_lan2 pal lo 32 aa non-interlace
vi_pal_laf2 pal lo 32 aa interlace
vi_pal_hpn1 pal hi 16 ps normal interlace
vi_pal_hpf1 pal hi 16 ps deflickered interlace
vi_pal_han1 pal hi 16 aa normal interlace
vi_pal_haf1 pal hi 16 aa deflickered interlace
vi_pal_hpn2 pal hi 32 ps normal interlace
vi_pal_hpf2 pal hi 32 ps deflickered interlace
In low-res (320 pixels by 240 lines), there is a choice between
non-interlace and interlace. Non-interlace repeats the same frame each
field. Interlace interpolates between adjacent lines, weighted 75% of
the line above plus 25% of the line below in the first field, then
weighted 25% of the line above plus 75% of the line below in the
second field. Note that there will be no flicker, since there are no
high spatial frequencies. Interlace requires slightly more memory
bandwidth due to the interpolation (but no additional frame buffer
memory). However, it will look better.
In hi-res (640 pixels by 480 lines), there is a choice between normal
interlace and deflickerd interlace. Normal interlace uses just the
data rendered in one field to display that field. This mode can use
one hi-res frame buffer without additional double buffering, since one
field can be displayed while the next is being rendered without
stepping on each other. However, any single pixel high detail will
flicker, since it will only be displayed in one field. Deflickered
interlace averages adjacent lines to filter out the high frequency
vertical detail, but at the cost of requiring double buffering of the
entire hi-res frame, since both rendered fields are used to display
each field.
There are, in general, two sets of settings for each of the two fields
in a frame, which must be switched each vertical blanking interval. In
non-interlace modes, the two settings are the same in each field.
vi_ntsc_lpn1
---------------
register address field 1 value field 2 value
--------- -------- -------- --------
VI_ORIGIN 04400004 00000280 00000280
VI_WIDTH 04400008 00000140 00000140
VI_V_INT 0440000c 00000208 00000208
VI_BURST 04400014 03e52239 03e52239
VI_V_SYNC 04400018 0000020d 0000020d
VI_H_SNYC 0440001c 00000c15 00000c15
VI_LEAP 04400020 0c150c15 0c150c15
VI_H_START 04400024 006c02ec 006c02ec
VI_V_START 04400028 002501ff 002501ff
VI_V_BURST 0440002c 000e0204 000e0204
VI_X_SCALE 04400030 00000200 00000200
VI_Y_SCALE 04400034 00000400 00000400
VI_CTRL 04400000 0001320e 0001320e
vi_ntsc_lpf1
---------------
register address field 1 value field 2 value
--------- -------- -------- --------
VI_ORIGIN 04400004 00000280 00000280
VI_WIDTH 04400008 00000140 00000140
VI_V_INT 0440000c 00000208 00000208
VI_BURST 04400014 03e52239 03e52239
VI_V_SYNC 04400018 0000020c 0000020c
VI_H_SYNC 0440001c 00000c15 00000c15
VI_LEAP 04400020 0c150c15 0c150c15
VI_H_START 04400024 006c02ec 006c02ec
VI_V_START 04400028 002301fd 002501ff
VI_V_BURST 0440002c 000e0204 000e0204
VI_X_SCALE 04400030 00000200 00000200
VI_Y_SCALE 04400034 01000400 03000400
VI_CTRL 04400000 0001324e 0001324e
vi_ntsc_lan1
---------------
register address field 1 value field 2 value
--------- -------- -------- --------
VI_ORIGIN 04400004 00000280 00000280
VI_WIDTH 04400008 00000140 00000140
VI_V_INT 0440000c 00000208 00000208
VI_BURST 04400014 03e52239 03e52239
VI_V_SYNC 04400018 0000020d 0000020d
VI_H_SYNC 0440001c 00000c15 00000c15
VI_LEAP 04400020 0c150c15 0c150c15
VI_H_START 04400024 006c02ec 006c02ec
VI_V_START 04400028 002501ff 002501ff
VI_V_BURST 0440002c 000e0204 000e0204
VI_X_SCALE 04400030 00000200 00000200
VI_Y_SCALE 04400034 00000400 00000400
VI_CTRL 04400000 0001311e 0001311e
vi_ntsc_laf1
---------------
register address field 1 value field 2 value
--------- -------- -------- --------
VI_ORIGIN 04400004 00000280 00000280
VI_WIDTH 04400008 00000140 00000140
VI_V_INT 0440000c 00000208 00000208
VI_BURST 04400014 03e52239 03e52239
VI_V_SYNC 04400018 0000020c 0000020c
VI_H_SYNC 0440001c 00000c15 00000c15
VI_LEAP 04400020 0c150c15 0c150c15
VI_H_START 04400024 006c02ec 006c02ec
VI_V_START 04400028 002301fd 002501ff
VI_V_BURST 0440002c 000e0204 000e0204
VI_X_SCALE 04400030 00000200 00000200
VI_Y_SCALE 04400034 01000400 03000400
VI_CTRL 04400000 0001305e 0001305e
vi_ntsc_lpn2
---------------
register address field 1 value field 2 value
--------- -------- -------- --------
VI_ORIGIN 04400004 00000500 00000500
VI_WIDTH 04400008 00000140 00000140
VI_V_INT 0440000c 00000208 00000208
VI_BURST 04400014 03e52239 03e52239
VI_V_SYNC 04400018 0000020d 0000020d
VI_H_SYNC 0440001c 00000c15 00000c15
VI_LEAP 04400020 0c150c15 0c150c15
VI_H_START 04400024 006c02ec 006c02ec
VI_V_START 04400028 002501ff 002501ff
VI_V_BURST 0440002c 000e0204 000e0204
VI_X_SCALE 04400030 00000200 00000200
VI_Y_SCALE 04400034 00000400 00000400
VI_CTRL 04400000 0000330f 0000330f
vi_ntsc_lpf2
---------------
register address field 1 value field 2 value
--------- -------- -------- --------
VI_ORIGIN 04400004 00000500 00000500
VI_WIDTH 04400008 00000140 00000140
VI_V_INT 0440000c 00000208 00000208
VI_BURST 04400014 03e52239 03e52239
VI_V_SYNC 04400018 0000020c 0000020c
VI_H_SYNC 0440001c 00000c15 00000c15
VI_LEAP 04400020 0c150c15 0c150c15
VI_H_START 04400024 006c02ec 006c02ec
VI_V_START 04400028 002301fd 002501ff
VI_V_BURST 0440002c 000e0204 000e0204
VI_X_SCALE 04400030 00000200 00000200
VI_Y_SCALE 04400034 01000400 03000400
VI_CTRL 04400000 0000324f 0000324f
vi_ntsc_lan2
---------------
register address field 1 value field 2 value
--------- -------- -------- --------
VI_ORIGIN 04400004 00000500 00000500
VI_WIDTH 04400008 00000140 00000140
VI_V_INT 0440000c 00000208 00000208
VI_BURST 04400014 03e52239 03e52239
VI_V_SYNC 04400018 0000020d 0000020d
VI_H_SYNC 0440001c 00000c15 00000c15
VI_LEAP 04400020 0c150c15 0c150c15
VI_H_START 04400024 006c02ec 006c02ec
VI_V_START 04400028 002501ff 002501ff
VI_V_BURST 0440002c 000e0204 000e0204
VI_X_SCALE 04400030 00000200 00000200
VI_Y_SCALE 04400034 00000400 00000400
VI_CTRL 04400000 0000301f 0000301f
vi_ntsc_laf2
---------------
register address field 1 value field 2 value
--------- -------- -------- --------
VI_ORIGIN 04400004 00000500 00000500
VI_WIDTH 04400008 00000140 00000140
VI_V_INT 0440000c 00000208 00000208
VI_BURST 04400014 03e52239 03e52239
VI_V_SYNC 04400018 0000020c 0000020c
VI_H_SYNC 0440001c 00000c15 00000c15
VI_LEAP 04400020 0c150c15 0c150c15
VI_H_START 04400024 006c02ec 006c02ec
VI_V_START 04400028 002301fd 002501ff
VI_V_BURST 0440002c 000e0204 000e0204
VI_X_SCALE 04400030 00000200 00000200
VI_Y_SCALE 04400034 01000400 03000400
VI_CTRL 04400000 0000305f 0000305f
vi_ntsc_hpn1
---------------
register address field 1 value field 2 value
--------- -------- -------- --------
VI_ORIGIN 04400004 00000500 00000a00
VI_WIDTH 04400008 00000500 00000500
VI_V_INT 0440000c 00000208 00000208
VI_BURST 04400014 03e52239 03e52239
VI_V_SYNC 04400018 0000020c 0000020c
VI_H_SYNC 0440001c 00000c15 00000c15
VI_LEAP 04400020 0c150c15 0c150c15
VI_H_START 04400024 006c02ec 006c02ec
VI_V_START 04400028 002301fd 002501ff
VI_V_BURST 0440002c 000e0204 000e0204
VI_X_SCALE 04400030 00000400 00000400
VI_Y_SCALE 04400034 00000400 00000400
VI_CTRL 04400000 0001334e 0001334e
vi_ntsc_hpf1
---------------
register address field 1 value field 2 value
--------- -------- -------- --------
VI_ORIGIN 04400004 00000500 00000a00
VI_WIDTH 04400008 00000280 00000280
VI_V_INT 0440000c 00000208 00000208
VI_BURST 04400014 03e52239 03e52239
VI_V_SYNC 04400018 0000020c 0000020c
VI_H_SYNC 0440001c 00000c15 00000c15
VI_LEAP 04400020 0c150c15 0c150c15
VI_H_START 04400024 006c02ec 006c02ec
VI_V_START 04400028 002301fd 002501ff
VI_V_BURST 0440002c 000e0204 000e0204
VI_X_SCALE 04400030 00000400 00000400
VI_Y_SCALE 04400034 02000800 02000800
VI_CTRL 04400000 0001324e 0001324e
vi_ntsc_han1
---------------
register address field 1 value field 2 value
--------- -------- -------- --------
VI_ORIGIN 04400004 00000500 00000a00
VI_WIDTH 04400008 00000500 00000500
VI_V_INT 0440000c 00000208 00000208
VI_BURST 04400014 03e52239 03e52239
VI_V_SYNC 04400018 0000020c 0000020c
VI_H_SYNC 0440001c 00000c15 00000c15
VI_LEAP 04400020 0c150c15 0c150c15
VI_H_START 04400024 006c02ec 006c02ec
VI_V_START 04400028 002301fd 002501ff
VI_V_BURST 0440002c 000e0204 000e0204
VI_X_SCALE 04400030 00000400 00000400
VI_Y_SCALE 04400034 00000400 00000400
VI_CTRL 04400000 0001305e 0001305e
vi_ntsc_haf1
---------------
register address field 1 value field 2 value
--------- -------- -------- --------
VI_ORIGIN 04400004 00000500 00000a00
VI_WIDTH 04400008 00000280 00000280
VI_V_INT 0440000c 00000208 00000208
VI_BURST 04400014 03e52239 03e52239
VI_V_SYNC 04400018 0000020c 0000020c
VI_H_SYNC 0440001c 00000c15 00000c15
VI_LEAP 04400020 0c150c15 0c150c15
VI_H_START 04400024 006c02ec 006c02ec
VI_V_START 04400028 002301fd 002501ff
VI_V_BURST 0440002c 000e0204 000e0204
VI_X_SCALE 04400030 00000400 00000400
VI_Y_SCALE 04400034 02000800 02000800
VI_CTRL 04400000 0001305e 0001305e
vi_ntsc_hpn2
---------------
register address field 1 value field 2 value
--------- -------- -------- --------
VI_ORIGIN 04400004 00000a00 00001400
VI_WIDTH 04400008 00000500 00000500
VI_V_INT 0440000c 00000208 00000208
VI_BURST 04400014 03e52239 03e52239
VI_V_SYNC 04400018 0000020c 0000020c
VI_H_SYNC 0440001c 00000c15 00000c15
VI_LEAP 04400020 0c150c15 0c150c15
VI_H_START 04400024 006c02ec 006c02ec
VI_V_START 04400028 002301fd 002501ff
VI_V_BURST 0440002c 000e0204 000e0204
VI_X_SCALE 04400030 00000400 00000400
VI_Y_SCALE 04400034 00000400 00000400
VI_CTRL 04400000 0000334f 0000334f
vi_ntsc_hpf2
---------------
register address field 1 value field 2 value
--------- -------- -------- --------
VI_ORIGIN 04400004 00000a00 00001400
VI_WIDTH 04400008 00000280 00000280
VI_V_INT 0440000c 00000208 00000208
VI_BURST 04400014 03e52239 03e52239
VI_V_SYNC 04400018 0000020c 0000020c
VI_H_SYNC 0440001c 00000c15 00000c15
VI_LEAP 04400020 0c150c15 0c150c15
VI_H_START 04400024 006c02ec 006c02ec
VI_V_START 04400028 002301fd 002501ff
VI_V_BURST 0440002c 000e0204 000e0204
VI_X_SCALE 04400030 00000400 00000400
VI_Y_SCALE 04400034 02000800 02000800
VI_CTRL 04400000 0000324f 0000324f
vi_pal_lpn1
---------------
register address field 1 value field 2 value
--------- -------- -------- --------
VI_ORIGIN 04400004 00000280 00000280
VI_WIDTH 04400008 00000140 00000140
VI_V_INT 0440000c 0000023a 0000023a
VI_BURST 04400014 0404233a 0404233a
VI_V_SYNC 04400018 00000271 00000271
VI_H_SYNC 0440001c 00150c69 00150c69
VI_LEAP 04400020 0c6f0c6e 0c6f0c6e
VI_H_START 04400024 00800300 00800300
VI_V_START 04400028 005f0239 005f0239
VI_V_BURST 0440002c 00110267 00110267
VI_X_SCALE 04400030 00000200 00000200
VI_Y_SCALE 04400034 00000400 00000400
VI_CTRL 04400000 0001320e 0001320e
vi_pal_lpf1
---------------
register address field 1 value field 2 value
--------- -------- -------- --------
VI_ORIGIN 04400004 00000280 00000280
VI_WIDTH 04400008 00000140 00000140
VI_V_INT 0440000c 0000023a 0000023a
VI_BURST 04400014 0404233a 0404233a
VI_V_SYNC 04400018 00000270 00000270
VI_H_SYNC 0440001c 00150c69 00150c69
VI_LEAP 04400020 0c6f0c6e 0c6f0c6e
VI_H_START 04400024 00800300 00800300
VI_V_START 04400028 005f0239 005d0237
VI_V_BURST 0440002c 00110267 00110267
VI_X_SCALE 04400030 00000200 00000200
VI_Y_SCALE 04400034 01000400 03000400
VI_CTRL 04400000 0001324e 0001324e
vi_pal_lan1
---------------
register address field 1 value field 2 value
--------- -------- -------- --------
VI_ORIGIN 04400004 00000280 00000280
VI_WIDTH 04400008 00000140 00000140
VI_V_INT 0440000c 0000023a 0000023a
VI_BURST 04400014 0404233a 0404233a
VI_V_SYNC 04400018 00000271 00000271
VI_H_SYNC 0440001c 00150c69 00150c69
VI_LEAP 04400020 0c6f0c6e 0c6f0c6e
VI_H_START 04400024 00800300 00800300
VI_V_START 04400028 005f0239 005f0239
VI_V_BURST 0440002c 00110267 00110267
VI_X_SCALE 04400030 00000200 00000200
VI_Y_SCALE 04400034 00000400 00000400
VI_CTRL 04400000 0001311e 0001311e
vi_pal_laf1
---------------
register address field 1 value field 2 value
--------- -------- -------- --------
VI_ORIGIN 04400004 00000280 00000280
VI_WIDTH 04400008 00000140 00000140
VI_V_INT 0440000c 0000023a 0000023a
VI_BURST 04400014 0404233a 0404233a
VI_V_SYNC 04400018 00000270 00000270
VI_H_SYNC 0440001c 00150c69 00150c69
VI_LEAP 04400020 0c6f0c6e 0c6f0c6e
VI_H_START 04400024 00800300 00800300
VI_V_START 04400028 005f0239 005d0237
VI_V_BURST 0440002c 00110267 00110267
VI_X_SCALE 04400030 00000200 00000200
VI_Y_SCALE 04400034 01000400 03000400
VI_CTRL 04400000 0001305e 0001305e
vi_pal_lpn2
---------------
register address field 1 value field 2 value
--------- -------- -------- --------
VI_ORIGIN 04400004 00000500 00000500
VI_WIDTH 04400008 00000140 00000140
VI_V_INT 0440000c 0000023a 0000023a
VI_BURST 04400014 0404233a 0404233a
VI_V_SYNC 04400018 00000271 00000271
VI_H_SYNC 0440001c 00150c69 00150c69
VI_LEAP 04400020 0c6f0c6e 0c6f0c6e
VI_H_START 04400024 00800300 00800300
VI_V_START 04400028 005f0239 005f0239
VI_V_BURST 0440002c 00110267 00110267
VI_X_SCALE 04400030 00000200 00000200
VI_Y_SCALE 04400034 00000400 00000400
VI_CTRL 04400000 0000330f 0000330f
vi_pal_lpf2
---------------
register address field 1 value field 2 value
--------- -------- -------- --------
VI_ORIGIN 04400004 00000500 00000500
VI_WIDTH 04400008 00000140 00000140
VI_V_INT 0440000c 0000023a 0000023a
VI_BURST 04400014 0404233a 0404233a
VI_V_SYNC 04400018 00000270 00000270
VI_H_SYNC 0440001c 00150c69 00150c69
VI_LEAP 04400020 0c6f0c6e 0c6f0c6e
VI_H_START 04400024 00800300 00800300
VI_V_START 04400028 005f0239 005d0237
VI_V_BURST 0440002c 00110267 00110267
VI_X_SCALE 04400030 00000200 00000200
VI_Y_SCALE 04400034 01000400 03000400
VI_CTRL 04400000 0000324f 0000324f
vi_pal_lan2
---------------
register address field 1 value field 2 value
--------- -------- -------- --------
VI_ORIGIN 04400004 00000500 00000500
VI_WIDTH 04400008 00000140 00000140
VI_V_INT 0440000c 0000023a 0000023a
VI_BURST 04400014 0404233a 0404233a
VI_V_SYNC 04400018 00000271 00000271
VI_H_SYNC 0440001c 00150c69 00150c69
VI_LEAP 04400020 0c6f0c6e 0c6f0c6e
VI_H_START 04400024 00800300 00800300
VI_V_START 04400028 005f0239 005f0239
VI_V_BURST 0440002c 00110267 00110267
VI_X_SCALE 04400030 00000200 00000200
VI_Y_SCALE 04400034 00000400 00000400
VI_CTRL 04400000 0000301f 0000301f
vi_pal_laf2
---------------
register address field 1 value field 2 value
--------- -------- -------- --------
VI_ORIGIN 04400004 00000500 00000500
VI_WIDTH 04400008 00000140 00000140
VI_V_INT 0440000c 0000023a 0000023a
VI_BURST 04400014 0404233a 0404233a
VI_V_SYNC 04400018 00000270 00000270
VI_H_SYNC 0440001c 00150c69 00150c69
VI_LEAP 04400020 0c6f0c6e 0c6f0c6e
VI_H_START 04400024 00800300 00800300
VI_V_START 04400028 005f0239 005d0237
VI_V_BURST 0440002c 00110267 00110267
VI_X_SCALE 04400030 00000200 00000200
VI_Y_SCALE 04400034 01000400 03000400
VI_CTRL 04400000 0000305f 0000305f
vi_pal_hpn1
---------------
register address field 1 value field 2 value
--------- -------- -------- --------
VI_ORIGIN 04400004 00000500 00000a00
VI_WIDTH 04400008 00000500 00000500
VI_V_INT 0440000c 0000023a 0000023a
VI_BURST 04400014 0404233a 0404233a
VI_V_SYNC 04400018 00000270 00000270
VI_H_SYNC 0440001c 00150c69 00150c69
VI_LEAP 04400020 0c6f0c6e 0c6f0c6e
VI_H_START 04400024 00800300 00800300
VI_V_START 04400028 005f0239 005d0237
VI_V_BURST 0440002c 00110267 00110267
VI_X_SCALE 04400030 00000400 00000400
VI_Y_SCALE 04400034 00000400 00000400
VI_CTRL 04400000 0001334e 0001334e
vi_pal_hpf1
---------------
register address field 1 value field 2 value
--------- -------- -------- --------
VI_ORIGIN 04400004 00000500 00000a00
VI_WIDTH 04400008 00000280 00000280
VI_V_INT 0440000c 0000023a 0000023a
VI_BURST 04400014 0404233a 0404233a
VI_V_SYNC 04400018 00000270 00000270
VI_H_SYNC 0440001c 00150c69 00150c69
VI_LEAP 04400020 0c6f0c6e 0c6f0c6e
VI_H_START 04400024 00800300 00800300
VI_V_START 04400028 005f0239 005d0237
VI_V_BURST 0440002c 00110267 00110267
VI_X_SCALE 04400030 00000400 00000400
VI_Y_SCALE 04400034 02000800 02000800
VI_CTRL 04400000 0001324e 0001324e
vi_pal_han1
---------------
register address field 1 value field 2 value
--------- -------- -------- --------
VI_ORIGIN 04400004 00000500 00000a00
VI_WIDTH 04400008 00000500 00000500
VI_V_INT 0440000c 0000023a 0000023a
VI_BURST 04400014 0404233a 0404233a
VI_V_SYNC 04400018 00000270 00000270
VI_H_SYNC 0440001c 00150c69 00150c69
VI_LEAP 04400020 0c6f0c6e 0c6f0c6e
VI_H_START 04400024 00800300 00800300
VI_V_START 04400028 005f0239 005d0237
VI_V_BURST 0440002c 00110267 00110267
VI_X_SCALE 04400030 00000400 00000400
VI_Y_SCALE 04400034 00000400 00000400
VI_CTRL 04400000 0001305e 0001305e
vi_pal_haf1
---------------
register address field 1 value field 2 value
--------- -------- -------- --------
VI_ORIGIN 04400004 00000500 00000a00
VI_WIDTH 04400008 00000280 00000280
VI_V_INT 0440000c 0000023a 0000023a
VI_BURST 04400014 0404233a 0404233a
VI_V_SYNC 04400018 00000270 00000270
VI_H_SYNC 0440001c 00150c69 00150c69
VI_LEAP 04400020 0c6f0c6e 0c6f0c6e
VI_H_START 04400024 00800300 00800300
VI_V_START 04400028 005f0239 005d0237
VI_V_BURST 0440002c 00110267 00110267
VI_X_SCALE 04400030 00000400 00000400
VI_Y_SCALE 04400034 02000800 02000800
VI_CTRL 04400000 0001305e 0001305e
vi_pal_hpn2
---------------
register address field 1 value field 2 value
--------- -------- -------- --------
VI_ORIGIN 04400004 00000a00 00001400
VI_WIDTH 04400008 00000500 00000500
VI_V_INT 0440000c 0000023a 0000023a
VI_BURST 04400014 0404233a 0404233a
VI_V_SYNC 04400018 00000270 00000270
VI_H_SYNC 0440001c 00150c69 00150c69
VI_LEAP 04400020 0c6f0c6e 0c6f0c6e
VI_H_START 04400024 00800300 00800300
VI_V_START 04400028 005f0239 005d0237
VI_V_BURST 0440002c 00110267 00110267
VI_X_SCALE 04400030 00000400 00000400
VI_Y_SCALE 04400034 00000400 00000400
VI_CTRL 04400000 0000334f 0000334f
vi_pal_hpf2
---------------
register address field 1 value field 2 value
--------- -------- -------- --------
VI_ORIGIN 04400004 00000a00 00001400
VI_WIDTH 04400008 00000280 00000280
VI_V_INT 0440000c 0000023a 0000023a
VI_BURST 04400014 0404233a 0404233a
VI_V_SYNC 04400018 00000270 00000270
VI_H_SYNC 0440001c 00150c69 00150c69
VI_LEAP 04400020 0c6f0c6e 0c6f0c6e
VI_H_START 04400024 00800300 00800300
VI_V_START 04400028 005f0239 005d0237
VI_V_BURST 0440002c 00110267 00110267
VI_X_SCALE 04400030 00000400 00000400
VI_Y_SCALE 04400034 02000800 02000800
VI_CTRL 04400000 0000324f 0000324f