Menu

Show posts

This section allows you to view all posts made by this member. Note that you can only see posts made in areas you currently have access to.

Show posts Menu

Topics - Levas

#1
To give a more simpler option to new people, I've uploaded the build I'm using myself.
It has all the modules needed and activated for raw recording.
It also has the settings I'm using (this is the actual build I'm using on my SD card, I removed the rom.bin files in the data section, which will be rewritten with your own camera roms at first startup)

This is just a ML build, including all the needed modules and set up in the way I'm using it, so for use, you camera must have firmware 1.1.6 and already be able to run ML (bootflag enabled).
If you haven't already enabled the bootflag, download the nightly build for the 6D on the downloads page and do a firmware update with the 'ML-SETUP.FIR' file which is in the nightly build.

Now here is this all in one, already set up lazy build:
https://drive.google.com/file/d/1_mZQUQrossmi8KgjOFg-HkcuLU7_qz8n/view?usp=sharing

When using the above build, you will see at that the following modules are enabled:


Out of the box, raw video recording in 1824x1026 - 16:9 aspect ratio in 14 bit lossless MLV format with audio is set.
Now in order to make use of the higher resolutions in crop mode, select a crop mode in the movie tab:


After selecting a crop mode preset, be sure to check the resolution displayed behind RAW VIDEO in the movie tab menu screen.
You can change/set this resolution by entering RAW VIDEO menu (highlight RAW VIDEO on the movie tab screen and press Q button on camera) now you see this menu:

Although instead of N/A there are resolution and aspect ratio displayed.
You can change the resolution by highlighting the option and use the scroll wheel on top of the camera to adjust it by increments of 16 pixels wide.
Now be sure to be in 5x zoom when recording and setting up the resolution.

Once crop_mode and higher resolution is set, you can record in 5x zoom these higher resolutions.
When in 5x zoom, you will get Canon 5x zoom preview, but this is not the actual framing of your recording.
To see the actual framing, you can do half-shutter press on the shutter button (lightly press the shutter button), this enables a preview(although frame rate is rather slow) of your actual recording/framing.
You can use the half shutter button before recording and during recording. Since the actual preview/framing is rather slow in frame rate, I choose to have Canon preview by default and occasionally check framing with half shutter press.

Now with these high resolutions, you will see that you won't get long recording times, since the camera can't write fast enough.
To get some longer recording times, you can enable 12 or 10 bit options in crop_mode submenu.
Go to the ML menu and MOVIE tab, highlight your crop_mode and press Q button:


There is a variety of options here, but most don't work (these options are used on EOS-M), but the most important option is working here, bitdepth.
You can select 8, 9, 10 or 12 bit here. If OFF is selected, you're using plain 14 bit raw option.

10 and 12 bit are really useful, 8 and 9 bit gives posterised shadows, but you can check for yourself  :P .

For viewing, extracting, post process and convert the MLV files I recommend to use MLV app:
https://www.magiclantern.fm/forum/index.php?topic=20025.0

ENJOY  :D

EDIT, this build has the SD_UHS speed hack enabled by default. This works good for SD cards that can handle this speed hack, for example the sanddisk extreme pro 95MB/s series or the newer Sandisk Extreme pro 170MB/s series.
Some cards, which can't handle this hack, can become slower with this hack enabled. Or maybe even won't write at all when the hack is enabled.
When experiencing problems, disable the SD_UHS module in the modules tab (Or buy a better SD card  ;D):



#2
Post-processing Workflow / Mac automator and kandaoRaw+
October 31, 2018, 08:36:40 PM
I'v found this program, KandaoRaw+, which can blend/average multiple raw files into a single new raw file (and remove ghosting and such).
What I like is the fact that it outputs a new raw file  :D and not just a jpg file or something  :P
And it does this really fast.

The program is free and available for Mac and windows:
https://www.kandaovr.com/en/software/#KDRAW

The program can blend/average up to 16 raw files into a single new raw file.

Now I would like to use this on a magic lantern raw video dng sequence.
The idea is to use kandaoraw+ to blend/average the frames, by starting with frame 1 and using the 15 following frames to create a new frame 1 file.
Next thing to do is blend/average frame 2 and the following 15 frames to create a new blended/average frame 2.
etc.

The problem is, I have no idea how to automate this and if it is even possible to automate, my first thought was using automator on Mac.
But I can't find out how to use automator with kandaoraw+ and let it blend each frame with their 15 following frames...
Anybody here an idea ?

#3
I'm having problems with dark frame subtraction and the lossless dng file format.
I wonder if any of you have the same experience or that it is the software(RawTherapee) I use ?

I find out that with adobe DNG converter, you can convert normal DNG files to Lossless DNG files.
So I'm testing with converting old DNG's from raw recordings I did years ago to lossless DNG format.
But for Iso 6400 and sometimes ISO 3200 I made dark frames. which work fine with the none lossless DNG files.
But once converted to lossless, the dark frame subtraction doesn't work (picture gets really dark and the remaining highlights look posterized).
I've also tried converting the dark frames to lossless DNG file format, so both files have the same format, but this didn't gave other results.

The lossless file format from adobe DNG converter should be the same LJ92 compression ML uses.
Did maybe any of you use full res silent picture with lossless option and tried dark frame subtraction with it ?
#4
Hello pixelpeepers,

I was wondering if people already seen/tested the difference in video quality of 10bit magic lantern raw vs 14 bit magic lantern raw.
I've used 10 bit raw today in the real world. And I'm used to color aliasing, and in most cases I have no problem in removing the color aliasing.
But the footage I shot today got very hard to remove, or almost impossible to remove, color aliasing.
Could it be explained in the difference between 10bit vs 14 bit, could that give more color aliasing ?
Or was it just a coincidence, since I shot a lot of bricks today, while visiting an old castle :P

#5
Hi there,

I want to mess around with ML and the 1000d.

I don't no much(or anything at all) about compiling magic lantern and the c language.

Long ago some work has been done by Coutts
http://bitbucket.org/coutts/1000d_dev/src

I used the instructions in this post to be able to compile
http://www.magiclantern.fm/forum/index.php?topic=16012.0

I'm now able to compile the standard platforms, like for example the newest 6d build. I get the fir, autoexec.bin and the zipfile.
But I can't get it to work with the stuff Coutts has made.

I get this error message in terminal:

rm *.o; rm AUTOEXEC.arm.elf
rm: *.o: No such file or directory
rm: AUTOEXEC.arm.elf: No such file or directory
make: *** [clean] Error 1

So there are some linking problems, but I don't know where to link to these files, I know they are available in the 'gcc-arm-none-eabi-4_8-2013q4' folder on my home directory

Anyone here willing to help/ suggest ideas
#6
Saw this new photo editing program Affinity.
http://affinity.serif.com/en-gb/

supported cameras:
http://affinity.serif.com/forum/index.php?/topic/4630-supported-develop-raw-cameras/

I'm curious how it handles magic lantern video dng's, anyone tried it yet ?
I can find nothing about the debayering it uses, so it probably uses one technique...is it any good for line skipping video files ?
#7
A1ex is talking about a temporal antialiasing filter. See quote at the bottom of this post.

I like the idea and was wondering if it is possible to control the aperture in video mode during one single frame ?
Let's say you're filming in PAL (25fps).
Is it possible to use a shutter of 1/25th of a second and for example use:
f22 for 1/100th of the time
f4.0 for 1/50th of the time
f22 for 1/100th of the time.
Which results in one single frame which took 1/25th  of a second?

Or, does anyone know if it's possible to use this in full res silent picture mode.
Let's say, time-lapse with 10 seconds interval. Can I get a 10 seconds single exposure, where the diafragma is f22 in the first and last few seconds of exposure and in the middle a seconds of f8 exposure ?

------------------------------------------------------------------------------------------------------------
Quote from A1ex in topic: www.magiclantern.fm/forum/index.php?topic=11787.150;topicseen

Another neat trick that can be done with a high-fps sensor, even if the final output has a low fps: temporal antialiasing filter.

Other companies did it in a different way, but the end result would be pretty similar:
http://tessive.com/time-control-faq/
http://www.red.com/learn/red-101/cinema-temporal-aliasing

and it was discussed here as well:
http://www.magiclantern.fm/forum/index.php?topic=5074
http://www.magiclantern.fm/forum/index.php?topic=8555.msg82898#msg82898

I'm going to implement this in ML for timelapse, with these functions. It will probably drain the battery pretty fast (because the sensor will be capturing light all the time), but it should give a very nice motion blur. Didn't try, but the theory makes sense.

This is also possible in post, but not practical (you would have record at say 300 fps and transcode to 24).
#8
General Chat / Open platform community - Olympus
September 22, 2014, 10:27:50 AM
Olympus starts something called the Open platform community

http://opc.olympus-imaging.com/en/index.html

Sounds like an open source m4/3 camera, with this approach, m4/3 could get interesting  :D
#9
 ;D

stupid idea but:

Would it be possible to activate the canon lens cleaning function(shaking the first low pas filter continuous) while recording raw video ?
And would it be possible that the lens cleaning function would act as a anti aliasing filter for video ?


#10
 ::)

Is it possible for magic lantern to use the Canon in lens image stabilization to create multiple photo files with little shift in perspective so a super resolution image could be created from these multiple files ?
Just as talked about in this article:
http://appleinsider.com/articles/14/05/08/future-iphones-may-use-optical-image-stabilization-to-create-super-resolution-images

I believe this technique was also used in the past in camcorders, with a shifting sensor instead of optical image stabilization, but the results are the same, higher resolution...

I have no idea if this is possible for magic lantern, nor if there's interest from the users.
Would be nice to create some stunning high resolution landscape images...
Or use it for example for video, once we reach 50 fps on the 5d Mark III wit MLV video, I say, let's use image stabilization and create some stunning 25fps super resolution video...
#11
Hello,

Lately I'm shooting live performances of musicbands indoors.(Loving the magic zoom and focus peaking for my manual lens!)
The light on stage varies a lot, so I'm shooting at fixed aperture and fixed iso and let the camera do the math for shutter value.
Most of the times I end up with hundreds of pictures, of which I select the best 10.

I was wondering if it is possible (in any way, maybe there are better alternatives then ML) to find out what the average shutter value is on all the photo's on my memory card.
If I have this information, I can consider for example a lower iso value (and know most of the pictures are still good and don't have a shutter time that's too long...).

So what I'm looking for is an option in ML that analysis the exif info of all the photo's on a memory card and return some numbers, for example:
The average, highest and lowest values of Aperture, Shuttertime and ISO.

If someone know's an existing solution for this, let me know, I'm on a mac(I know that spotlight can read the exif info) and use lightroom (is there any option in lightroom for this?).





#12
Hi there,

Does anyone know if I should once in a while re-fromat my card, make it bootable again and put ML on it, for better performance/stability ?
Cause I've found the same build of ML to behave different on different cards(read/write speed isn't the issue)

Could it be that the card get messed up with files by heavy recording and deleting (and not formatting) ?

#13
I'm curious what is the max FPS people achieved with the FPS override function in ML in standard 1080p (in combination with rawrec or MLV-rec modules) ?

I'v seen and used ML builds for 6d where 35 FPS is possible in 1080p mode.
Did anyone got higher than 35fps with trial and error settings in the FPS override menu in ML ?
#14
Hi fellow raw shooters,

Does anyone know how to edit multiple images in raw-therapee ?
I have a folder which contains all the DNG's from one sequence, I select one of them to edit (white balance, demosaicing etc), then I can copy the things I've edit to the clipboard, but then I'm stuck.
I can click on another picture and paste the edit on it, but I can't find a way or a button to select ALL the files in the folder at once and the paste the edit to ALL the images ???

Working on the latest Mac version of raw-therapee.

Used to work with Lightroom, but lightroom gives rather soft images compared to DaVinci Resolve.
Worked a while in DaVinci Resolve to find out that the demosaicing gives me more aliasing than lightroom. Especially with the Samyang 14mm lens (probably too sharp too handle)
Than I heard about Raw-therapee and the multiple options of demosaicing, tried it, WOW!!! Probably the best way to get rid of aliasing stuff.
But how do I batch process these images......