Etiquette, expectations, entitlement...@autoexec_bin | #magiclantern | Discord | Reddit | Server issues
to raw.c or to chdk-dng.c ?Edgar
i guess it is not your fault, but mine you can keep the -m32 removed maybe.can you check if compression still works?
edgar@linux-lyl7:~/test> ./mlv_dump -c -l 9 -o compress.mlv 1.MLV MLV Dumper v1.0-----------------[i] Mode of operation: - Input MLV file: '1.MLV' - Rewrite MLV - Compress frame data - Output into 'compress.mlv'[i] File 1.MLV opened[i] Processing...[i] Unknown Block: NULL, skipping saving: 790272 -> 740650 (93.72%)[i] Unknown Block: NULL, skipping saving: 790272 -> 748475 (94.71%)
...Block: IDNT Offset: 0x000001ac Size: 84 Time: 28.726000 ms Camera Name: 'ERR:1 md:0x 0 ml:0' Camera Serial: '' Camera Model: 0x00000000...
you wanted to try it again, you said a few weeks ago the module is not camera-dependent, you can use it for any model.it should work with any camera on which the original raw_rec works too.but it may require some symbols in the ML/MODULES/<model>.SYM, so you may get an error.which error do you get? which model?
can you hg pull and recompile magic lantern again?...you then should see your camera information in the recorded movie.
ML ASSERT:(int)size == hSuite->sizeat ../../src/exmem.c:122 (shoot_malloc_suite), task raw_rec_tasklv:1 mode:3
6D freezes when stopping 2nd recordML ASSERT:(int)size == hSuite->sizeat ../../src/exmem.c:122 (shoot_malloc_suite), task raw_rec_tasklv:1 mode:3
6D freezes when stopping 2nd record
edgar@linux-lyl7:~/MagicLantern/magic-lantern-mlv/modules/raw_rec> diff raw_rec_original.c raw_rec.c68,69c68,69< #include <math.h< #include <cropmarks.h>>---> #include <math.h>> #include <cropmarks.h>342c342< static unsigned int raw_rec_should_preview(unsigned int ctx);---> static uint32_t raw_rec_should_preview(uint32_t ctx);
can you open raw.h in src/ directory and replace the raw_info struct with this one?Code: [Select]/* raw image info (geometry, calibration levels, color, DR etc); parts of this were copied from CHDK */struct raw_info { int32_t api_version; // increase this when changing the structure uint32_t buffer; // points to image data int32_t height, width, pitch; int32_t frame_size; int32_t bits_per_pixel; // 14 int32_t black_level; // autodetected int32_t white_level; // somewhere around 13000 - 16000, varies with camera, settings etc // would be best to autodetect it, but we can't do this reliably yet union // DNG JPEG info { struct { int32_t x, y; // DNG JPEG top left corner int32_t width, height; // DNG JPEG size } jpeg; struct { int32_t origin[2]; int32_t size[2]; } crop; }; union // DNG active sensor area (Y1, X1, Y2, X2) { struct { int32_t y1, x1, y2, x2; } active_area; int32_t dng_active_area[4]; }; int32_t exposure_bias[2]; // DNG Exposure Bias (idk what's that) int32_t cfa_pattern; // stick to 0x02010100 (RGBG) if you can int32_t calibration_illuminant1; int32_t color_matrix1[18]; // DNG Color Matrix int32_t dynamic_range; // EV x100, from analyzing black level and noise (very close to DxO)};
/* raw image info (geometry, calibration levels, color, DR etc); parts of this were copied from CHDK */struct raw_info { int32_t api_version; // increase this when changing the structure uint32_t buffer; // points to image data int32_t height, width, pitch; int32_t frame_size; int32_t bits_per_pixel; // 14 int32_t black_level; // autodetected int32_t white_level; // somewhere around 13000 - 16000, varies with camera, settings etc // would be best to autodetect it, but we can't do this reliably yet union // DNG JPEG info { struct { int32_t x, y; // DNG JPEG top left corner int32_t width, height; // DNG JPEG size } jpeg; struct { int32_t origin[2]; int32_t size[2]; } crop; }; union // DNG active sensor area (Y1, X1, Y2, X2) { struct { int32_t y1, x1, y2, x2; } active_area; int32_t dng_active_area[4]; }; int32_t exposure_bias[2]; // DNG Exposure Bias (idk what's that) int32_t cfa_pattern; // stick to 0x02010100 (RGBG) if you can int32_t calibration_illuminant1; int32_t color_matrix1[18]; // DNG Color Matrix int32_t dynamic_range; // EV x100, from analyzing black level and noise (very close to DxO)};
you can keep the -m32 removed maybe.
add a#include <stdint.h>to the c file as first line
