Author Topic: Danne's crop_rec_4k experiments for EOS M  (Read 155425 times)

Tullen

  • Freshman
  • **
  • Posts: 73
Re: Danne's crop_rec_4k experiments for EOS M
« Reply #2250 on: April 13, 2020, 12:03:54 PM »
Hi again. I apologize for the many posts and that I might ask for feedback a bit early in the development, making it harder for you to come with useful feedback. Never the less, here is another question.

Desired outcome concept
I am planing to structure the guide around the concept of "desired outcome" where "The camera settings, lens and lens settings used, sd card, other accessories and post processing decides how useful a tool the total system is to achieve a desired outcome." So basically, what kind of film clip you want to end up with.

Examples
Here a desired outcome could be: A clean noise free "cinematic looking" (shallow depth of field) shot of a person in front of a monument at night. Or: A sharp, high dynamic range shot of a landscape with a wide viewing angel. Or: A slow motion clip.

Aspects related to desired outcomes
So far I have identified the following aspects that relates to desired outcomes (and to the settings and tools that we can use with the Eos M camera):
Depth of field (cinematic look?/ lens and lens settings used)
Slow motion (high fps/post processing)
Framing (crop factor and lens/speed booster)
Image quality (noise, moire, clippings (highlight detail), crushing (shadow detail), sharpness, jello effect, dynamic range)
Image stabilization (gimbal, post process stabilization or other)

Looking for feedback
What do you think about this approach?
Have I missed any relevant aspects?

Danne

  • Developer
  • Hero Member
  • *****
  • Posts: 7240
Re: Danne's crop_rec_4k experiments for EOS M
« Reply #2251 on: April 13, 2020, 01:32:34 PM »
Binning modes seems relevant. 1x1 being more or less moiré and aliasing free.
I can't say if your approach will work effectively for most users but go ahead and try.

Wlad81

  • Member
  • ***
  • Posts: 131
Re: Danne's crop_rec_4k experiments for EOS M
« Reply #2252 on: April 13, 2020, 08:17:20 PM »
So, could you please be so kind to tell me what is better for shooting Full HD or around that in 14 bit loseless? 100D or EOS M?
Canon EOS 5D Mk III + Canon 24-105 F/4 L IS USM + SanDisk Exreme Pro 64 GB (SD, ML Nightly.2021Feb07.5D3113) + SanDisk Extreme Pro 128 GB (CF).

Danne

  • Developer
  • Hero Member
  • *****
  • Posts: 7240
Re: Danne's crop_rec_4k experiments for EOS M
« Reply #2253 on: April 13, 2020, 08:31:39 PM »
They are more or less the same, 100D being a buggy mf.

Daniel_hungary

  • New to the forum
  • *
  • Posts: 2
Re: Danne's crop_rec_4k experiments for EOS M
« Reply #2254 on: April 13, 2020, 10:22:01 PM »
So hey again. If you remember me? A post about flickering RAW video a few posts back.

Someone asked for the unprocessed footage. No point to do that, it is the same before processing. I suspect the raw levels. This "bug" drives me nuts. Dual Iso would be a nice feature, but this makes it unusable. Almost every time it comes out like this. Am I missing something in the menu? Is it something in the factory Canon menu? Dunno

:(

Walter Schulz

  • Contributor
  • Hero Member
  • *****
  • Posts: 8259
Re: Danne's crop_rec_4k experiments for EOS M
« Reply #2255 on: April 13, 2020, 11:10:12 PM »
If you don't want to get your problem analysed: It's your party.
If you want someone to look into it: Unprocessed footage needed.

optical-work

  • New to the forum
  • *
  • Posts: 19
Re: Danne's crop_rec_4k experiments for EOS M
« Reply #2256 on: April 14, 2020, 02:39:16 AM »
About dual-iso, I engaged in getting a result without flickering before. I worked with mlvapp and mlvfs without success. It's not obvious for me but it's either how dual-iso captured or how it's processed. So I think dual-iso isn't ideal for video. A plus is shorter post-processing times.

Tullen

  • Freshman
  • **
  • Posts: 73
Re: Danne's crop_rec_4k experiments for EOS M
« Reply #2257 on: April 14, 2020, 11:42:51 AM »
Hi again.

Compared to the features of the nightly builds at https://builds.magiclantern.fm/features.html which features are added or removed/not working in the crop_rec_4k_mlv_snd_raw_only_2020Jan18.EOSM202?

Hope answering wont be to hard, but if I am to go through the different features in the guide (with the help of the wiki) it would be a great starting point to know the above.

Cheers/ Tullen

Danne

  • Developer
  • Hero Member
  • *****
  • Posts: 7240
Re: Danne's crop_rec_4k experiments for EOS M
« Reply #2258 on: April 14, 2020, 11:54:52 AM »
Can you find out yourself? Personally I have no time to spend around this approach but it should be possible to dig out info by testing.

Tullen

  • Freshman
  • **
  • Posts: 73
Re: Danne's crop_rec_4k experiments for EOS M
« Reply #2259 on: April 14, 2020, 12:23:06 PM »
Can you find out yourself? Personally I have no time to spend around this approach but it should be possible to dig out info by testing.

Yes, of course. Would been faster if someone already knew on the top of their head or at least knew more than me, but I can do it. How would you recommend I check? Just go through the menus and see what's there, what's missing and what's added?

tweak

  • New to the forum
  • *
  • Posts: 40
Re: Danne's crop_rec_4k experiments for EOS M
« Reply #2260 on: April 14, 2020, 01:17:52 PM »
Im using f1.4 tamron 4-12mm and i ripped plastic cover to expose aperture mechanism and control it.Im planning to get something from 12 to 80mm maybe more tele as well.
Crop mode on M is exact fit for 1/2" lenses, 2/3" and super 16mm film camera lenses
--
Adjusted 2.5k again, i had thin lines probably because skipping top and bottom was not accurate.
Code: [Select]
case CROP_PRESET_2K_EOSM:
            /* set ratio preset */
            if (ratios == 0x1)
            {
                skip_left       = 36;
                skip_right      = 36;
                skip_top        = 8;
                skip_bottom     = 38;
            }
            if (ratios == 0x2)
            {
                skip_left       = 36;
                skip_right      = 36;
                skip_top        = 0;
                skip_bottom     = 28;
            }
            if (ratios == 0x3)
            {
                skip_left       = 197;
                skip_right      = 197;
                skip_top        = 8;
                skip_bottom     = 204;
            }
            break;


Code: [Select]
            case CROP_PRESET_2K_EOSM:
  cmos_new[5] = 0x2c0;             /* vertical (first|last) */
              cmos_new[7] = 0x2a8;    /* pink highlights without this */
                break;
               
Well , looks like lines are there from pressing left arrow on wheel and coming back to framing preview.They disappear slowly and are gone after about 10 seconds.

Can someone please help me with where I need to put this code? I also would like to use these same lenses and have trouble because of uneven vignette.  :)

cmh

  • Contributor
  • Member
  • *****
  • Posts: 138
Re: Danne's crop_rec_4k experiments for EOS M
« Reply #2261 on: April 15, 2020, 09:07:55 AM »
This code is part of a module named crop_rec.c
Assuming that you use an EOS M, you'd need to compile Danne's build.
If you are on Windows:
Download Cygwin at https://www.cygwin.com/setup-x86_64.exe in "%UserProfile%"\Downloads"

Open a command prompt (press Windows+R, type cmd) and get Cygwin installed with the required packages:
Code: [Select]
cd %UserProfile%"\Downloads
setup-x86_64.exe -s ftp://cygwin.uib.no/pub/cygwin/ -q -P mingw64-i686-gcc-core,gcc-core,make,mercurial,python37-docutils,perl,wget,unzip,zip,diffutils

Open the Cygwin Terminal from the start menu.
Download the GNU Arm Embedded Toolchain and extract it in the right folder:
Code: [Select]
cd && wget https://developer.arm.com/-/media/Files/downloads/gnu-rm/9-2019q4/RC2.1/gcc-arm-none-eabi-9-2019-q4-major-win32.zip.bz2
unzip -q ~/gcc-arm-none-eabi-9-2019-q4-major-win32.zip.bz2 -d gcc-arm-none-eabi-9-2019-q4-major-win32
rm ~/gcc-arm-none-eabi-9-2019-q4-major-win32.zip.bz2

Clone the repository and the right branch (hg identify -b should show crop_rec_4k_mlv_snd_raw_only):
Code: [Select]
cd && hg clone https://bitbucket.org/Dannephoto/magic-lantern_jip-hop/src/unified/
mv ~/unified ~/magic-lantern_dannephoto
cd ~/magic-lantern_dannephoto && hg up crop_rec_4k_mlv_snd_raw_only
hg identify -b

Modify what you want to in C:\cygwin\home\<username>\magic-lantern_dannephoto.
In your case crop_rec.c (replace <username> by yours).

Get to the right platform (EOS M here), cleanup, compile, move zip in C:\cygwin\home\<username>, cleanup again:
Code: [Select]
cd ~/magic-lantern_dannephoto/platform/EOSM.202
make clean
make -j4 zip
mv magiclantern-Nightly*.zip ~/
make clean
cd && ls

Copy the content of C:\cygwin\home\<username>\magiclantern-Nightly.YMD.EOSM202.zip to your SD card and you're set (a reinstallation might be necessary depending on your modifications).


WilliamEwart

  • New to the forum
  • *
  • Posts: 2
Re: Danne's crop_rec_4k experiments for EOS M
« Reply #2262 on: April 23, 2020, 02:41:35 AM »
Hello, I was wondering if the EOS M could shoot continuous at 1080p at a 4x3(1436 x 1080) aspect ratio? I am planning on shooting a short film this summer and want to get better quality video than what my Fuji xt1 can get. I have a Kowa 2x anamorphic that I plan on using so I'd like to shoot at 4x3 so that I get a 2.4 aspect ratio after the stretch. My main concern is that the EOS M has a limited record time at anything deeper than 10 bit at that ratio but I can not find specific information about it and I would like to know for curtain before pulling the trigger on buying one. Thank you.

McFly

  • New to the forum
  • *
  • Posts: 10
Re: Danne's crop_rec_4k experiments for EOS M
« Reply #2263 on: April 23, 2020, 05:05:43 PM »
Hello, I was wondering if the EOS M could shoot continuous at 1080p at a 4x3(1436 x 1080) aspect ratio? I am planning on shooting a short film this summer and want to get better quality video than what my Fuji xt1 can get. I have a Kowa 2x anamorphic that I plan on using so I'd like to shoot at 4x3 so that I get a 2.4 aspect ratio after the stretch. My main concern is that the EOS M has a limited record time at anything deeper than 10 bit at that ratio but I can not find specific information about it and I would like to know for curtain before pulling the trigger on buying one. Thank you.
I have the same equipments. I'm recently shooting continuous at 1392x1158. With 2x stretch, the resolution changes 2784x1158 (2.40:1 aspect ratio). Of course, you can also shoot continuous 1536x1158 if you want (1:33 a/r to 2.66:1 a/r).

WilliamEwart

  • New to the forum
  • *
  • Posts: 2
Re: Danne's crop_rec_4k experiments for EOS M
« Reply #2264 on: April 23, 2020, 05:16:17 PM »
I have the same equipments. I'm recently shooting continuous at 1392x1158. With 2x stretch, the resolution changes 2784x1158 (2.40:1 aspect ratio). Of course, you can also shoot continuous 1536x1158 if you want (1:33 a/r to 2.66:1 a/r).
Thank you. What is the bit depth that you can get at that resolution?

cmh

  • Contributor
  • Member
  • *****
  • Posts: 138
Re: Danne's crop_rec_4k experiments for EOS M
« Reply #2265 on: April 23, 2020, 06:19:12 PM »
I just tested 1440x1080 with the mv1080 profile on a bright scene with the sky in frame.
It goes past 52 MB/s at times when overexposed at 14 bits so it's not continuous
I get 45 MB/s overexposed at 12 bits and exposed for the sky at 14 bits, around 36 MB/s exposed for the sky at 12 bits.
12 bits should be safe but I've only tested recording a minute.

I can test 2,5k real quick if you don't mind the crop.
edit: 2,5 k, same resolution and bit depth so I get more or less the same results .

tweak

  • New to the forum
  • *
  • Posts: 40
Re: Danne's crop_rec_4k experiments for EOS M
« Reply #2266 on: April 24, 2020, 02:54:48 AM »
I just tested 1440x1080 with the mv1080 profile on a bright scene with the sky in frame.
It goes past 52 MB/s at times when overexposed at 14 bits so it's not continuous
I get 45 MB/s overexposed at 12 bits and exposed for the sky at 14 bits, around 36 MB/s exposed for the sky at 12 bits.
12 bits should be safe but I've only tested recording a minute.

I can test 2,5k real quick if you don't mind the crop.
edit: 2,5 k, same resolution and bit depth so I get more or less the same results .

You can just back the resolution down a bit if you want continuous, I doubt you could notice the difference, especially using anamorphic lenses (I can't). If you want to get it back up to 1080 just upscale on export from MLVapp.

tweak

  • New to the forum
  • *
  • Posts: 40
Re: Danne's crop_rec_4k experiments for EOS M
« Reply #2267 on: April 24, 2020, 02:57:38 AM »
This code is part of a module named crop_rec.c
Assuming that you use an EOS M, you'd need to compile Danne's build.

I'm on a Mac... I gave it all a go but couldn't get it to work  :( .

@2blackbar - would you be willing to share your build with me? Would definitely be a huge help, I've got the same lens as you.

Cheers.

canoneer

  • New to the forum
  • *
  • Posts: 39
Re: Danne's crop_rec_4k experiments for EOS M
« Reply #2268 on: April 24, 2020, 11:21:55 AM »
Hello - a little help as I have looked into using the focus stack tool on my M - but in my absence from using ML actively for a while I am not sure the build I have support the focus tools at all.

I have build Nightly.2017May28.EOSM202.

And I can not find any of the focus tools in the menus.

I know I probably am clueless - but....help?

canoneer

  • New to the forum
  • *
  • Posts: 39
Re: Danne's crop_rec_4k experiments for EOS M
« Reply #2269 on: April 24, 2020, 11:31:14 AM »
I now see that I approached this issue earlier as well without ny resolution - sorry if this repeating myself, but the issues with using the M lenses are they still there? And what if I use an "ef" lens will I then have access to the tools on my M? If so which build should I look for?

Thankful and happy ML user (apart from my lack of m-focus-tool).

(The focus move (and and trigger without focusing) work with Canons Camera connect app on M3 with same lens - so I thought the lens´ability was there generically....)

Walter Schulz

  • Contributor
  • Hero Member
  • *****
  • Posts: 8259
Re: Danne's crop_rec_4k experiments for EOS M
« Reply #2270 on: April 24, 2020, 11:57:36 AM »
Top of page -> Downloads -> Download nightly builds -> EOS M -> Troubleshooting
Focus stacking not available in Nightly Builds for EOS M: Top of same page -> Features -> Focus

Try Danne's build for EOS M:
https://www.magiclantern.fm/forum/index.php?topic=23041.0

2blackbar

  • Senior
  • ****
  • Posts: 495
Re: Danne's crop_rec_4k experiments for EOS M
« Reply #2271 on: April 24, 2020, 12:12:13 PM »
I'm on a Mac... I gave it all a go but couldn't get it to work  :( .

@2blackbar - would you be willing to share your build with me? Would definitely be a huge help, I've got the same lens as you.

Cheers.
https://drive.google.com/open?id=1q7WE8TF_IjIRuYA-LWoRFHWoMb4fFZ2z
Im using mostly 2.5k mode with this lens and set it to  1920x1080p by dropping down resolution from 2520xsomethingdontermember.
You can finetune framing by toching on those triangles on sides, your framing will moev a bit.
Sometimes after you finish recording , framing is moved , you can reset it to be fine again by refreshing ml view, by pressing info button a couple of times to come back to liveview.

tweak

  • New to the forum
  • *
  • Posts: 40
Re: Danne's crop_rec_4k experiments for EOS M
« Reply #2272 on: April 25, 2020, 03:37:22 AM »
Cheers! @2blackbar, got it working now.

raptor

  • New to the forum
  • *
  • Posts: 3
Re: Danne's crop_rec_4k experiments for EOS M
« Reply #2273 on: April 26, 2020, 09:06:24 PM »
Hi, is anyone here who uses FRSP (full res silent picture) feature on a regular basis or at least played a bit with it?

1, What is the range of the usable shutter speeds when it comes to FRSP on the EOS M?
2, Can you use FRSP with exposure bracketing? Or focus stacking?
3, Can you successfully integrate FRSP into Lua scripts? For example: 1st command: take FRSP; 2nd command: change exposure; 3rd command: take FRSP and so on.

I am thinking about buying a used EOS M10 (found one for a pretty good price) but I want to use primarly the pure electronic shutter (FRSP) because I will be taking huge amount of photos (reference images and photogrammetry).

The second option is to switch to a micro 4/3 system that officially supports pure electronic shutter (Panasonic, Olympus, Fuji, ...). But that would be more expensive as I already own some Canon gear and accessories. Also some native lenses (e.g. Fuji) are quite pricey and hard to get as used. Also the size of the sensor is smaller.
But the unique feature of the EOS M + Magic Lantern combo would definitively be the ability to run custom scripts (providing the FRSP works in scripts). Hardly any other script-able cameras on the market.

Just to be clear I am not completely new to FRSP, I know some of its limitations, I own EOS 550D and experiment with it but unfortunately the 550D has extremely narrow range of shutter speeds you can use for FRSP (0.5 to 0.8 sec). I need a camera with bigger range. I don't mind using ND or polarising filter for bright scenes.


Thank you!

Danne

  • Developer
  • Hero Member
  • *****
  • Posts: 7240
Re: Danne's crop_rec_4k experiments for EOS M
« Reply #2274 on: April 27, 2020, 04:25:45 AM »
Lua can be used to modify shutter, iso, other settings. If frsp enabled shouldn't make no difference.
Longer shutter times than with 550d. Don't think so but not sure here. I tend to use lua scripts and run 5k anamorphic flv preset instead for easier workflows straight in mlv app.
If I want long shutter times I would set cam to cr2 photo mode and run some lua or ml bracketing feature with or without a customized lua script. I never heard of a worn out eos m shutter ever.
The lua function really is a winner. Not any cam I know of outside ml has this feature available or anything close to it.