Author Topic: Bilal's crop_rec_4k experiments for 650D / T4i  (Read 4145 times)

theBilalFakhouri

  • Developer
  • Hero Member
  • *****
  • Posts: 657
Bilal's crop_rec_4k experiments for 650D / T4i
« on: September 21, 2020, 09:56:22 AM »
Edit 6/6/2021: this thread is outdated, please move to the main thread for both 650D / 700D (T4i / T5i):
https://www.magiclantern.fm/forum/index.php?topic=25784.msg231049#msg231049


New Build for 650D
Download link: crop_rec_4k.2020Sep30.650D104.zip

New-Preset" border="0
Build-modules" border="0
New-Build" border="0

It's the same build from 700D, follow these posts for How to use it and What contains:
https://www.magiclantern.fm/forum/index.php?topic=5951.msg231049#msg231049
https://www.magiclantern.fm/forum/index.php?topic=5951.msg231134#msg231134

-Confirmed working by @LeandroFreitas

Please feedback, after you have some fun :P
700D 1.1.5 | no more ISOless LV err 8 / SDR104 @ 240 MHz - Constant! | Fixed Scrambled LiveView in Higher resolution | Real-Time correct framing in the Way

Walter Schulz

  • Contributor
  • Hero Member
  • *****
  • Posts: 7991
Re: Bilal's crop_rec_4k experiments for 650D / T4i
« Reply #1 on: September 22, 2020, 07:54:29 AM »
ISOless 8 & 15 Dual ISO fix.

ISOless PH err(2) prevails. Anything I can contribute to nail it?

LeandroFreitas

  • New to the forum
  • *
  • Posts: 21
Re: Bilal's crop_rec_4k experiments for 650D / T4i
« Reply #2 on: September 22, 2020, 08:59:57 AM »
Hey Walter.
There is problem with shutter speed for 1440p and higher resolution.
Its standard 29.98fps that make its hard to record due high frames, i set to 24fps override and the shutter speeds go to 1/250 or higher.
That would really be nice to get that fix.
I dont have this issue with EOS M running Dannes Build or with Billal custombuild of Dannes 100D branch he did for 650D that iam running, the only thing that dont work is the buttons zoom and info etc
.

Walter Schulz

  • Contributor
  • Hero Member
  • *****
  • Posts: 7991
Re: Bilal's crop_rec_4k experiments for 650D / T4i
« Reply #3 on: September 22, 2020, 09:01:30 AM »
I have and had this problem in photo mode with all tested builds.

theBilalFakhouri

  • Developer
  • Hero Member
  • *****
  • Posts: 657
Re: Bilal's crop_rec_4k experiments for 650D / T4i
« Reply #4 on: September 23, 2020, 06:01:49 AM »
ISOless PH err(2) prevails. Anything I can contribute to nail it?

RIP-00000" border="0

Reproduced the error on 700D:
ISOLess-PH-err2" border="0

It seems getting the right addresses aren't enough on your 650D, so you need to tweak these too in dual_iso.c:

Code: [Select]
        FRAME_CMOS_ISO_SIZE  =       0x22;

        PHOTO_CMOS_ISO_SIZE  =       0x10;

Error happens but it's err(8 ) now!
This result looks promising, let's try again:

1-Grab this adtg_gui.mo, should work on crop_rec_4k builds, you can try on my latest build for 650D too.
2-Set ISO to 100, Let's start from video mode, set FPS override to 12 FPS to achieve more speed when enabling the module.
3-Load only adtg_gui.mo module, restart the camera, enable the module, Highlight "ADTG Registers" then press Q button, highlight CMOS 0, you need to get the address, write it down.
4-Switch to Photo mode, none-LiveView, take a picture, write down the new address from CMOS 0.

Redo the steps from 2 to 4, but this time Set ISO to 200.

This is the part containing the address we need in the white box:
VRAM0" border="0

Provide the new addresses for ISO 100 and 200 for both photo mode and video mode, let me check them . .
700D 1.1.5 | no more ISOless LV err 8 / SDR104 @ 240 MHz - Constant! | Fixed Scrambled LiveView in Higher resolution | Real-Time correct framing in the Way

a1ex

  • Administrator
  • Hero Member
  • *****
  • Posts: 12564
Re: Bilal's crop_rec_4k experiments for 650D / T4i
« Reply #5 on: September 23, 2020, 07:04:06 AM »
Refer to issue #1781.

Hardcoding addresses is simply *not* going to work on 650D, 700D and maybe other models. It may fix the problem for some cases (possibly depending on what camera mode is used at startup, or other settings, or simply because of the multitasking startup process used by Canon) and will fail otherwise. The address you are looking for is allocated dynamically, so one has to find a pointer to it somewhere in a RAM dump - that pointer is likely at some fixed address, regardless of the exact startup sequence.

Suggested approach:
- find a few cameras / startup configurations that result in different CMOS[0] addresses (at the very least 2, but I'd recommend 3 or more)
- hardcode the address from adtg_gui and confirm it's actually working reliably (predictably, deterministically) on each of those cameras and/or configurations
- get a RAM dump from each of those cameras and/or configurations
- annotate each RAM dump with the address that was confirmed to work
- blindly scan for a pointer to that address (the CMOS[0] address will be specific to each RAM dump, but the pointer address should be the same in all dumps)

Walter Schulz

  • Contributor
  • Hero Member
  • *****
  • Posts: 7991
Re: Bilal's crop_rec_4k experiments for 650D / T4i
« Reply #6 on: September 23, 2020, 05:40:57 PM »
- find a few cameras / startup configurations that result in different CMOS[0] addresses (at the very least 2, but I'd recommend 3 or more)

Count me in.

LeandroFreitas

  • New to the forum
  • *
  • Posts: 21

theBilalFakhouri

  • Developer
  • Hero Member
  • *****
  • Posts: 657
Re: Bilal's crop_rec_4k experiments for 650D / T4i
« Reply #8 on: September 24, 2020, 04:23:20 AM »
I will follow a1ex suggestion in the future, this may take a while, not sure when . . my experience with finding stubs or addresses in ROM dumps wasn't very successful, didn't look at RAM dump before . .

So for now I can help with a temporary solution for Walter's 650D, hardcoded values for his camera.

+1

@LeandroFreitas
Do you have problem with dual_iso module on your 650D, ISOless PH err(2) ?
700D 1.1.5 | no more ISOless LV err 8 / SDR104 @ 240 MHz - Constant! | Fixed Scrambled LiveView in Higher resolution | Real-Time correct framing in the Way

theBilalFakhouri

  • Developer
  • Hero Member
  • *****
  • Posts: 657
Re: Bilal's crop_rec_4k experiments for 650D / T4i
« Reply #9 on: September 30, 2020, 11:13:31 AM »
New Build for 650D
Download link: crop_rec_4k.2020Sep30.650D104.zip

-All Changes in the following post:
https://www.magiclantern.fm/forum/index.php?topic=25784.0
700D 1.1.5 | no more ISOless LV err 8 / SDR104 @ 240 MHz - Constant! | Fixed Scrambled LiveView in Higher resolution | Real-Time correct framing in the Way

masc

  • Contributor
  • Hero Member
  • *****
  • Posts: 1824
Re: Bilal's crop_rec_4k experiments for 650D / T4i
« Reply #10 on: October 04, 2020, 06:54:03 PM »
@theBilalFakhouri: For 700D I found your focus pixel maps and added them to the MLVApp repos. This has the advantage, the latest MLVApp versions (inofficial) are able to auto download the maps, if MLVApp loads the clip. Do you also have the maps for 650D? If you have, let me know, I can add them to the repos. Or we add you as contributor to our repos for such uploads! :)
5D3.113 | EOSM.202

theBilalFakhouri

  • Developer
  • Hero Member
  • *****
  • Posts: 657
Re: Bilal's crop_rec_4k experiments for 650D / T4i
« Reply #11 on: October 05, 2020, 10:13:53 AM »
Hello @masc

Yeah I have noticed that, it was quick, for 650D it *might* be identical focus maps to 700D maps, but with 650D ID in maps naming . . Not quite sure yet I don't have clips from 650D to check . . I will try to get some clips.
I have no reason to refuse your offer, that would be cool :)
700D 1.1.5 | no more ISOless LV err 8 / SDR104 @ 240 MHz - Constant! | Fixed Scrambled LiveView in Higher resolution | Real-Time correct framing in the Way

Fuguha

  • Just arrived
  • *
  • Posts: 1
Re: Bilal's crop_rec_4k experiments for 650D / T4i
« Reply #12 on: October 12, 2020, 10:31:48 AM »
Hello everyone,

@theBilalFakhouri

I've been playing with your builds on my 650D, if there's anything I can provide to help, let me know.

Walter Schulz

  • Contributor
  • Hero Member
  • *****
  • Posts: 7991
Re: Bilal's crop_rec_4k experiments for 650D / T4i
« Reply #13 on: October 17, 2020, 12:20:45 PM »
I'm not Bilal,
- Clips needed to build a focus map. See post above yours.
- Hunting down "ISOless-PH-err". See reply #2273 #5 (and more ... and more ...)

PT-pics

  • Just arrived
  • *
  • Posts: 1
Re: Bilal's crop_rec_4k experiments for 650D / T4i
« Reply #14 on: October 21, 2020, 07:50:54 AM »
I’m new here. 650D/t4i user - been searching through the posts looking for the right settings in ML to get the 29:59 restart to work correctly.  I have the LUA build running with firmware version 1.04.  I’ve gotten the Movie Tweaks, Movie Restart set to “On”.   Everything else having to do with Auto Shut-Off is turned off.  Also, I am NOT recording, but using my camera as a webcam.   I have the camera on movie mode, with a direct power source, and it is shooting in Manual Mode. 

I’m sure there is something I am overlooking.  Thanks for responding.  I do ASL Interpreting (now from home) and I really want to utilize my completely capable DSLR to get a clear pic. 

Walter Schulz

  • Contributor
  • Hero Member
  • *****
  • Posts: 7991
Re: Bilal's crop_rec_4k experiments for 650D / T4i
« Reply #15 on: October 21, 2020, 10:32:37 AM »
Prefs tab -> Config Options -> Restore ML defaults
Restart cam
Prefs tab -> Powersave in LiveView -> 30-minute timer Disable

eeeebalo

  • New to the forum
  • *
  • Posts: 2
Re: Bilal's crop_rec_4k experiments for 650D / T4i
« Reply #16 on: November 03, 2020, 10:50:56 AM »
Do you also have the maps for 650D?
for 650D it *might* be identical focus maps to 700D maps, but with 650D ID in maps naming . . Not quite sure yet I don't have clips from 650D to check

I got 650D! Thank you for your work on this camera and MLVApp guys.
Focus dots removal in MLVApp seem to be working perfectly: https://imgur.com/a/uwYCNRK

What kind of clip is needed to build the focus dots map? I just recorded a dark one, if it's of any use: https://mega.nz/folder/vjoljYRR#mCCwyhPtCzCfscNXXoNIBA

eeeebalo

  • New to the forum
  • *
  • Posts: 2
Re: Bilal's crop_rec_4k experiments for 650D / T4i
« Reply #17 on: November 03, 2020, 06:12:45 PM »
Provide the new addresses for ISO 100 and 200 for both photo mode and video mode, let me check them

I got some notes regarding ISOless error (2).

Whenever I get this error the address (for both photo and video modes) is always shifted exactly by 0x100: https://imgur.com/a/2Uw6Sra.

I can confirm that workaround suggested by linuxhenhao six years ago still works.
I compiled mentioned workaround based on Bilal's build for 650D and 700D. I can't tell if this works on any other model, but so far it works consistently on 650D: https://mega.nz/folder/CvY2DJZL#wBCp4bONkizG_e2a8qeoIQ

Walter Schulz

  • Contributor
  • Hero Member
  • *****
  • Posts: 7991
Re: Bilal's crop_rec_4k experiments for 650D / T4i
« Reply #18 on: November 11, 2020, 07:44:08 AM »
@theBilalFakhouri
Minor bug in build dated 2020Sept30:
Loading MLV_REC.mo and MLV_PLAY.mo (only those two) will result in link error
Code: [Select]
tcc: error: undefined symbol 'lossless_decompress_raw'
Workaround: Activate MLV_LITE.mo

relations99

  • Just arrived
  • *
  • Posts: 1
Re: Bilal's crop_rec_4k experiments for 650D / T4i
« Reply #19 on: November 28, 2020, 10:19:31 AM »
Hey all.

I cannot seem to get focus peaking to work with an external HDMI.
The resolution captured by my capture card is 1080p.
It keeps drawing in the middle only. See below.




If I change the recording resolution down to 720p, it seems to work.




Focus Peaking works perfectly fine in 1080p on the build-in LCD.
I've noticed that the spot metering location is also offset when displaying over HDMI in 1080p.

I'm running the build posted above, crop_4k 2020Sep30.
I have tried the version on the website as well, 2018Jul03

TenJedyny

  • New to the forum
  • *
  • Posts: 2
Re: Bilal's crop_rec_4k experiments for 650D / T4i
« Reply #20 on: December 01, 2020, 03:53:51 PM »
Have the same problem. Also loaded mlv_rec.mo but sometimes shuter is blocked on screen. I have set shuter +fine tuning and have 48 shuter but on screen i see 25, histogram says over. Need restart many times and load one again some modules to work again. For now i will back to more stable ml.

TenJedyny

  • New to the forum
  • *
  • Posts: 2
Re: Bilal's crop_rec_4k experiments for 650D / T4i
« Reply #21 on: December 09, 2020, 08:54:40 PM »
My observations:
Using last build witch is Nightly.2020Nov03.650D104
I have loaded modules:
+bench,
+crop_rec
+mlv_lite
+mlv_snd
+pic_view
+sd_uhs > 160MHz

NOW in Overlay > Global Draw i set LiveView and i am able to run mlv_rec without errors on screen (because of "LiveView") when turning on the camera.
Shutter error 1/25 still occurs sometimes. The above modules load almost correctly but the clips recorded by mlv_rec have no sound.

Icaab

  • New to the forum
  • *
  • Posts: 21
Re: Bilal's crop_rec_4k experiments for 650D / T4i
« Reply #22 on: March 07, 2021, 09:02:53 PM »
New Build for 650D
Download link: crop_rec_4k.2020Sep30.650D104.zip

-All Changes in the following post:
https://www.magiclantern.fm/forum/index.php?topic=5951.msg231434#msg231434

Hello. The other day I switched to this firmware on my 650d. Very pleased! But unfortunately I encountered the following problem:
When I rotate the wheel to adjust the shutter speed, it is always the same. So visually the light on the picture changes, and the value remains the same as on the attached screenshot. :-[


theBilalFakhouri

  • Developer
  • Hero Member
  • *****
  • Posts: 657
Re: Bilal's crop_rec_4k experiments for 650D / T4i
« Reply #23 on: March 11, 2021, 02:46:34 PM »
Hi @Icaab

This issue affects some 650D models, Your camera have different "Shutter Blanking" addresses, current solution to hardcode your camera's shutter blanking addresses in the source code, and compile the build.

I may do it in the future (Compile and Upload two 650D builds), but currently I don't have time.
700D 1.1.5 | no more ISOless LV err 8 / SDR104 @ 240 MHz - Constant! | Fixed Scrambled LiveView in Higher resolution | Real-Time correct framing in the Way

Walter Schulz

  • Contributor
  • Hero Member
  • *****
  • Posts: 7991
Re: Bilal's crop_rec_4k experiments for 650D / T4i
« Reply #24 on: March 11, 2021, 07:12:25 PM »
@Bilah: Is this a good time to address the underlying issue?

- find a few cameras / startup configurations that result in different CMOS[0] addresses (at the very least 2, but I'd recommend 3 or more)