Messages

If nobody says anything in maybe a day, I'll submit a pull request for review so it can at least be tested.   

But I would rather get it working on mine first. 

~Phil

I was looking over the code again and found a few errors in my code, fixed those, tested, and still nothing shows up if I hit rec as soon as I have screen before the liveview ML bars come over the top except the histogram and the rec red light.  Not sure what I'm missing.  If anyone wants to take a peek and give me hints:

https://bitbucket.org/frenchiefilms/phil/src/6c947a0e0965796e040177305aa083f3e17cf43d/src/bitrate-5d3.c?at=unified

~Phil

It happens fast, but it seems like when I press the play button to view the file it just is a black screen with white text, I just got it again:

Invalid Raw:
A:/pathtocard/file

but then I just tested and if I exit and reload then I can view the sequence just fine.  Odd.

Oh I just got the screen to stay:

Invalid RAW:
A:/DCIM/100EOS5D/M11-1226.RAW

I have 'no image' showing up too, so maybe there's some other corrupt file on the card I need to remove and it is unrelated.  I'm guessing it is safe to ignore for now.

~Phil

I tested this out, and not getting pink frames so far, but I do seem to get every other recording to say "Invalid RAW" quickly so I can't play them back.  This is with regular RAW, for some reason the mlv_rec module is failing to build with:

Code Select Expand


[ GCC      ]   mlv_dump
mlv_dump.c: In function 'lua_call_va':
mlv_dump.c:64: warning: unused variable 'verbose'
mlv_dump.c:63: warning: unused variable 'nres'
mlv_dump.c:63: warning: unused variable 'narg'
In file included from ../lv_rec/lv_rec.h:24,
                 from mlv_dump.c:305:
../../src/raw.h: At top level:
../../src/raw.h:185: warning: declaration does not declare anything
../../src/raw.h:193: warning: declaration does not declare anything
mlv_dump.c: In function 'main':
mlv_dump.c:2027: error: 'struct raw_info' has no member named 'active_area'
mlv_dump.c:2028: error: 'struct raw_info' has no member named 'active_area'
mlv_dump.c:2029: error: 'struct raw_info' has no member named 'jpeg'
mlv_dump.c:2030: error: 'struct raw_info' has no member named 'jpeg'
mlv_dump.c:2036: error: 'struct raw_info' has no member named 'active_area'
mlv_dump.c:2037: error: 'struct raw_info' has no member named 'active_area'
mlv_dump.c:2038: error: 'struct raw_info' has no member named 'jpeg'
mlv_dump.c:2039: error: 'struct raw_info' has no member named 'jpeg'
mlv_dump.c:2557: error: 'struct raw_info' has no member named 'active_area'
mlv_dump.c:2558: error: 'struct raw_info' has no member named 'active_area'
mlv_dump.c:2559: error: 'struct raw_info' has no member named 'active_area'
mlv_dump.c:2560: error: 'struct raw_info' has no member named 'active_area'
make: *** [mlv_dump] Error 1


Nope, no luck, I keep thinking I see somthing flicker but not seeing anything.  Are the default overlays covering it up or something?  Is there a way to disable the liveview bars in record mode for testing?  Is that what covers it up?

~Phil

Oh know what, I tweaked a few things as guesses and rebuilt, if you think that is okay, I can try to fix it back and rebuild give me a few.

~Phil

Nope, sadly it didn't, it still just shows the top with temp, framerate, and disk remaining.   I did try but it almost takes 3 seconds or more to show the screen.

~Phil

okay so the compile runs fine, but doesn't work as expected.  If someone can look at my src and tell me what seems off here I'd appreciate the pointers.  I'm sure its my
limited C skills hitting me.

https://bitbucket.org/frenchiefilms/phil/src/e4ae07b43c4879a71774a6524da8cec68d693994/src/bitrate-5d3.c?at=unified

~Phil

man noob move, I forgot you cd into platform/5D3.123/ to do make zip.  That worked.

~Phil

I'm building it by bypassing that, I read the Makefile, looks like the target for zip: after the docs is to cd into platform/all and make zip.  I'll report back with progress, I'd still like to know why that python script is missing?

Okay so I think I have a pass at this that is compiling, but all of a sudden, when I do "make zip" it errors out:

Code Select Expand


cd /home/pdavis/src/123/magic-lantern/platform/all; make docs
make[1]: Entering directory `/home/pdavis/src/123/magic-lantern/platform/all'
cd ../../doc && RST2LATEX=rst2latex.py python2 mkdoc.py
[[Focus distance]]
[[Cropmarks]]
[[External Recorders]]
[[Unbricking]]
[[install]]
[[HDMI recorder]]
sh: rst2latex.py: command not found
make[1]: *** [docs] Error 1
make[1]: Leaving directory `/home/pdavis/src/123/magic-lantern/platform/all'
make: *** [docs] Error 2



I don't know where the python latex script would be provided, but I'm going to search the wiki etc, it seems my build is failing due to docs, but I thought previously that the docs weren't forced on a make zip?

~Phil

Good to know I'm not the only one.  a1ex, are you using the 1.2.3 build on 5D3 or is this something global?

~Phil

So question about that then, would I replace this code with something like this:

Code Select Expand



case 1: // elapsed
            LVINVO_UPDATE_FUNC(elapsed)
            {
                     //code here to get same data avialable?
             }
            );


and then define the elapsed in the lv info section as well?

Code Select Expand



static struct lvinfo_item info_items[] = {




Also I don't know where the old UI used to show these values?  Bottom, top or wherever it fits?

~Phil

I'm looking into this, would this be due to the changes to the LVINFO_UPDATE_FUNC and this code is not in that new format?
from src/bitrate-5d3.c:

Code Select Expand


switch(rec_indicator)
    {
        case 0:
            return;
        case 1: // elapsed
            bmp_printf(
                FONT(FONT_MED, COLOR_WHITE, COLOR_BLACK),
                time_indic_x + 160 - 6 * font_med.width,
                time_indic_y,
                "%3d:%02d",
                elapsed_time / 60,
                elapsed_time % 60
            );
            return;


so if I replaced the code there with the new LVINFO_UPDATE_FUNC information would that do the trick?

Nobody has any input on if my pics could be a problem with code or are expected?   Or should I submit this issue to a new thread about the build specifically?

~Phil

Nevermind, I'll just dropbox them. 

So here's the scoop.  First off I was using some of the burn in and test features and had a few crashes.  Here are the logs and BMP screenshots. The two crashes were actually with the build linked here, the odd video behavior is with my build done afterwards. 

first crash
https://www.dropbox.com/s/yo7thtsigfcc64f/LOG000.LOG
https://www.dropbox.com/s/9ovokwkiy33axrv/VRAM0.BMP

and here is the second crash

https://www.dropbox.com/s/4o1bt95flngtfa7/LOG001.LOG
https://www.dropbox.com/s/bx3v2luqtmlmw0w/VRAM1.BMP

Also I've got three images, they'll be in order a DNG file saved as png from three video recordings I did with MLV.  1 is 1080p and had horrid colors all messed up.  2 is just fine, right after the first failure, and 3 is at a lower resolution so that I could record to my slower card.  It also was all messed up.  Is this a known problem?  Do I just have something set wrong or? 
1080p bad https://www.dropbox.com/s/sekurd8e9mi7h8z/bad1080p.png
1080p good https://www.dropbox.com/s/forj472o66is0qp/good1080p.png
864p bad https://www.dropbox.com/s/ck9rl0lxgypjz0d/bad864.png

This is what I'm seeing as off so far.

otherwise focus peaking, zebras on over exposure, audio levels seem fine.  The only other thing that seems  odd is sometimes things seem to be fighting for liveview and blink on/off a lot.  My focus points in the canon firmware seems to overwrite things if they're blinking and it alternates between the two, maybe normal behavior but I"m new to the ML in a lot of ways.


~Phil

I've been playing around a bit, and have the logs and bmp from crash screen,  where is the preferred landing spot for files?  I'll try to see about doing a public dropbox folder in the meantime.

I've had some really odd behavior with MLV raw recording.  This is from my own build, it all works, but sometimes the videos are extremely high noise and pinkish.  Not sure how to explain it better than that, but I'll get a few screencaps from it as well.

~Phil

Okay I've played with it a bit more, and gotten my build system working.  I've not yet tried my last working build, I'm using the one Chris linked earlier.  As was asked earlier, what should be tested?  I've tried several of the debug menu stuff like burn in tests and benchmarks and they're all seeming to crash.  I can get the logs and bmp if they're needed, but they may also not be the root of what needs testing.

Figured it out. Centos needed the libgcc.i686 rpm.

g3ggo got me on the right track, centos has old gcc. Getting that resolved.

which would be good except that the other modules don't build when it fails there.  I guess this becomes more related to general devel instead of this 123 build though.  I'll do as you say and work on the other issues in the general devel forum area.

Thanks,

~Phil

I got it figured out.  The libgcc.i686 version was missing on my system.  Now a ton more build just fine, but I get this error on mvl_dump which means that I can't use the mlv_rec for the new Magic Lantern video format.  Any ideas?  This def seems more like a C error:

[ GCC      ]   mlv_dump
[ GCC      ]   mlv_dump
[ GCC      ]   mlv_dump
mlv_dump.c: In function 'lua_call_va':
mlv_dump.c:64: warning: unused variable 'verbose'
mlv_dump.c:63: warning: unused variable 'nres'
mlv_dump.c:63: warning: unused variable 'narg'
In file included from ../lv_rec/lv_rec.h:24,
                 from mlv_dump.c:305:
../../src/raw.h: At top level:
../../src/raw.h:185: warning: declaration does not declare anything
../../src/raw.h:193: warning: declaration does not declare anything
mlv_dump.c: In function 'main':
mlv_dump.c:2025: error: 'struct raw_info' has no member named 'active_area'
mlv_dump.c:2026: error: 'struct raw_info' has no member named 'active_area'
mlv_dump.c:2027: error: 'struct raw_info' has no member named 'jpeg'
mlv_dump.c:2028: error: 'struct raw_info' has no member named 'jpeg'
mlv_dump.c:2034: error: 'struct raw_info' has no member named 'active_area'
mlv_dump.c:2035: error: 'struct raw_info' has no member named 'active_area'
mlv_dump.c:2036: error: 'struct raw_info' has no member named 'jpeg'
mlv_dump.c:2037: error: 'struct raw_info' has no member named 'jpeg'
mlv_dump.c:2555: error: 'struct raw_info' has no member named 'active_area'
mlv_dump.c:2556: error: 'struct raw_info' has no member named 'active_area'
mlv_dump.c:2557: error: 'struct raw_info' has no member named 'active_area'
mlv_dump.c:2558: error: 'struct raw_info' has no member named 'active_area'
make[3]: *** [mlv_dump] Error 1

Any thoughts on this error? 

~Phil

Did my fix not work for this?

Sorry I found that you didn't put this in as a pull request, so I just put your fix into my tree and submitted the pull request,  I gave credit to you for the find though.  a1ex if you want to wait and reject mine to give 1% credit by all means do so

~Phil

This is a really bad idea and you could potentially even damage your camera by doing this. Modules compiled for 1.1.3 are not compatible with 1.2.3. Once the 1.2.3 support is merged into ML outdated module code will be prevented from loading at all, but that safeguard is not yet in place:

https://bitbucket.org/hudson/magic-lantern/pull-request/354/ported-ml-over-to-5d3-firmware-123/activity#comment-1154370

Okay then anyone have any ideas why my raw2dng fails to build with the errors I'm seeing?  I'll try the build just linked, but I'd love for the modules to build on my dev system.

~Phil

I ended up just coying the modules from the last nightly build.

After more raw playing, I note every time I record raw I get:

Hack error at ff4acd4:
expected e3a0032, got 6e654c2c

Is this a known bug or something related to the new 123 build?

Its still recording raw, but with my slow card I only get a few seconds. 

~Phil