Messages

i'll install ubuntu and try to figure out how to fix this.

Wow, thanks! Let me know if I can provide any other info.

I should also say that it says "Run-time error '445': Object doesn't support this action" when I try to set the folder.

And, darn: I just ran it again via wine, and played around, but never tried to set the output folder (so there was no crash), and the exact same three lines came on stdout, so I think those log lines are not in fact related to the folder setting crash (directly, anyway)... let me know if there's a debug mode or something I can run it with.

One way to trigger the three log lines is to open the settings dialog; I have definitely seen it without opening the settings, though, but I can't find a way now besides the settings menu...

I also saw another wine error line apparently indicating that I need to install the winbind package on my system, so I did that as well. (Only mentioning for any others who are running via wine in linux.)

You can scroll down all tabs. Or resize window.

Sorry, this was just because the output file type did not involve dithering, etc. Please ignore.

I had some trouble with hot pixel removal, but I'll keep reading up on the forum here and see if I can figure it out

What do you open? MLV or RAW?

It was an MLV. I zoomed in on a couple hot pixels with the "+" activated, and the box would turn green, and I would click, but nothing would happen. Then I rendered and the hot pixels were still present. There is a good chance that I'm doing something wrong, because I'm not even sure how it's supposed to work, which is why I think I should probably read a little more first... :-)

Actually all tiffs that presented in exports list are 16 bit.

Ah, great, thanks.

Damn, every time I come across a new thread on this forum there is another incredible piece of Magic Lantern software! You folks are out of control.

I am offering a humble donation. I hope everyone is donating to this amazing project!

A few things:

- the download links on the first post are reversed re: INTEL vs. AMD

- I tried running MLVP in a windows XP VM (I'm on Linux usually) -- the file dialog that opens MLV files can access network drives, but the file dialog that sets the output folder can not. (In this case the 'Network Drive' is actually the mapped share drive for my VM).

- when running in the VM, it would hang when I tried to render the output. I'm sure this is low on the list of bugs, but just wanted to mention it in case it was useful. (I can also use my normal Win 8.1 install and that works fine.)

- I tried running it in Wine in Linux; first I did "winetricks vb6run", then I did:
  wine "z:\path\to\mlvproducer\MLVProducer.exe"
  ...but when i try to set the output directory, it crashes. Also the "include audio' and related radio buttons aren't present in that window. (Just reporting all this in case it's interesting.) This is the error on the crash:

Code Select Expand

err:ole:apartment_getclassobject DllGetClassObject returned error 0x80004002
err:ole:create_server class {172bddf8-ceea-11d1-8b05-00600806d9b6} not registered
err:ole:CoGetClassObject no class object {172bddf8-ceea-11d1-8b05-00600806d9b6} could be created for context 0x15


- I had some trouble with hot pixel removal, but I'll keep reading up on the forum here and see if I can figure it out

- 16bit TIFF export would be cool?

Thanks for the great software!

[UPDATE]

MLVFS is amazing, as are most ML-related things -- but I'm having a couple of issues.

I swear I tried to google this and came up empty :-)

I'm using a 100D (Rebel SL1). Latest ML version for it.

- when I use MLVFS the extracted frames are always exactly half the reported resolution -- is this expected due to demosaicing or something? All the tutorials and examples I see do not seem to halve the resolution, so it seems odd. E.g. the resolution set in the camera and shown in MLVFS is 1728x972, but the resulting .dng's are 864x486. Normal?

UPDATE: In MLVProducer, the frames from the same .MLV file are created full-sized. So this seems related to MLVFS somehow?

- I'm on linux; changes in the MLVFS interface are rarely reflected in the files, despite constant reloading of the webpage interface and refreshing in the file browser or exiting/entering via the command line. The normal-vs-davinci filenames change reflects itself immediately (without even a refresh of the page), everything else seems to not take effect no matter what I do. Once I was able to get some of the options to work, but usually I can not. The settings of the form seem to only take effect the first time the .MLV file is entered in the web GUI -- if I restart it, the settings take effect. Perhaps I need to disable some caching or something?

I know I can just set those options on the command line, so it's not a crisis, but it'd be handy for comparisons if the web interface worked.

Just mentioning this in case it might be relevant: I'm on linux -- it compiled fine but I wasn't sure how to set up the execution environment. I ended up running the executable from inside the build directory where it was created, but I had to copy the contents of data/ to the same directory so it could find the html/js files, etc.

Thanks for any ideas!

Mazahaka009 -- I just posted a couple posts back about that here -- was raw video recording working for you before upgrading to the new firmware?

I'm impressed that you got anything from it -- i was only able to get the hack error...

( If anyone's curious like I was, the Canon update seems a minor lens compatibility thing only: http://support-au.canon.com.au/contents/AU/EN/0400270502.html )

I AM ETERNALLY GRATEFUL FOR THIS PORT AND ML IN GENERAL. THANKS!

Question re: RAW video -- like a poster above, I am unable to get it to work on my 100D. I'm using a UHS-1/Class 10 card, so I'm not expecting high-res recording. I have it set for 960x540, 16:9, 128MB warm-up, extra hacks on, buffer fill method 4, reserve card space on. When I record it does not skip any frames, but it gives the "hack error" the other poster described, and no file is created. I can figure out a way to record the hack error if that would be useful.

Anyone having success with this? Anything I can do to help debug?

Many thanks,
-s