Messages

[0x05]

Here's a build using 0x05 wich is also free:

src/edmac-memcpy.c

Code Select Expand

#elif defined(CONFIG_7D)
uint32_t edmac_read_chan = 0x19;
uint32_t edmac_write_chan = 0x05;

https://www.dropbox.com/s/0452mx31r43pegx/ML_britom_10_Sep_2013_.zip

Updated here too: http://www.magiclantern.fm/forum/index.php?topic=7987

Hopefully we get less pink frames. Still need to test from 0x07 to 0x15
Built from 1% repository https://bitbucket.org/OtherOnePercent/tragic-lantern-6d/overview

USE AT YOUR OWN RISK!

EPIC! thanks a lot mister! updated also here http://www.magiclantern.fm/forum/index.php?topic=7987

Audio meters are working in the bin slave version for me!

@DragonFlame @%1

I had tried fps override but didn't work for me. I defined in features.h

Code Select Expand

#define FEATURE_FPS_OVERRIDE

Maybe i'm missing something.

WOW this app is amazing! thank you so much Dmitry and Yaroslav. It's Working on my 7D, but i forgot to test with raw recording enabled. Will test later.

buy a 5diii/5dii/7d/50d/6d

@1% Wow, you mean 1056x704!?!? i thought it was atleast 1500xsomething upscaled to 1080p :/



Talking about the 5x cropped mode, doing line skipping every 2 lines means we can cover the equivalent area of 5024x2400 or 5024x1200 pixels (the full sensor is 5184x3456) but still recording as 2512x1200. Am i right?

[5184/3=1728]

@thoma

Because the sensor size in pixels is 5,184 × 3,456

In 1x non crop mode the procesor does line skipping using 1 line every 3 lines (it's designed like this so the camera can supposedly output 1920x1080 h264 video without a lot of processing power)
In 5x cropped mode the processor takes a small part of the sensor without line skipping, therefore is possible to record at a higher resolution but not with the whole sensor (2512x1200)

Just divide 5184/3=1728 (max horizontal resolution in non cropped mode).

Now this comunity has acheived unbeliable things, so it's ok to dream that this barrier can be overcome.

  Thanks a lot 1%, it's looking good, i'm about to test it.

I've posted a build with your modules here: http://www.magiclantern.fm/forum/index.php?topic=7987

Sounds great devs!

Do you think if someday we discover how to take a full frame silent pic, we can save a line every 3 lines and perform deinterlacing in post (if necesary) resulting in less moire video or higher non-cropped resolution video?

It sounds like a classic crazy impossible question like "When can we get 4k 120fps video?" and i'm aware of the hardware limitations but if its really imposible i just wanna hear the "stop doing drugs" coming from you guys

Anyway big thanks for your work!

Me too! +1

This feature was available with 422 files but got deleted when the whole raw thing came up. I'm wondering if using the same method (capturing multiple 5x liveview images to cover the whole frame and then merging them to a single frame) can we get hi-res raw silent pics?

I'm getting 1.2mb/s slower speeds with nick's build, but i'm using an slow transcend 400x. My 7d is from 2009 too.


@feureau
Yes we need to organize our builds so i've started this thread and i'll update it from time to time. http://www.magiclantern.fm/forum/index.php?topic=7987

I've updated gcc-arm to version 4.7.4, but i dont know to wich level it affects the building process.

I've compiled my latest build from the hudson repository https://bitbucket.org/hudson/magic-lantern/. This repository has the normal raw_rec module that outputs .raw format.
Some compilation problems have been fixed within the code so the only thing you need to touch are probably the lines 24 and 25 in makefile.user.default to specify the right path to your gcc-arm

Code Select Expand

GCC_VERSION=-4.7.3
ARM_PATH=~/gcc-arm-none-eabi-4_7-2013q2

Instead if you want you can compile like nick did from the g3gg0 repository that has the new raw_rec that outputs .mlv format https://bitbucket.org/g3gg0/magic-lantern

I can confirm that this is +3MB/s faster than the previous .raw recording version. But the one nick posted several posts before records to the new .mlv format and that is about +20MB/s faster than this version on my cards.

Btw, how do you people get your sources? I mean, you're both compiling a new version with small hacks on but you each get two different versions. And I can't seem to compile the one I just cloned from hudson.

What's going on here? Too many forks?

I have not tested latest Nick's build yet. I've compiled my builds from https://bitbucket.org/hudson/magic-lantern/ on VM Virtualbox xubuntu using latest gcc-arm 4.7.4. I think is a good idea to have as many builds compiled on as many computers as possible so we can test and know wich configurations and compilers give the best results. And we can have something to play with until some developer give us a hand

[My Latest build:]

My Latest build:

- Latest fixes (Small hacks, some ettr stuff) from https://bitbucket.org/hudson/magic-lantern/commits/all
- All the modules working including autoexpo.mo
- I'm getting faster speeds compared to my latest build (+3mb/s) using the normal raw_rec that outputs .raw format (Can anyone else confirm?)

https://www.dropbox.com/sh/ifyeas56ltuqhg1/P-PeTt1Llz

USE AT YOUR OWN RISK

[To clarify, there are 3 latest builds:]

This made my day:


To clarify, there are 3 latest builds:

brunonicko's build with liveview freeze workaround
http://www.astronautafilmes.com/clientes/magiclantern/ML_BrunoNicko_26_Ago_2013.zip

my first build with liveview freeze workaround, some other updates from the bitbucket repo and latest original raw_rec module that outputs RAW format
https://www.dropbox.com/sh/ifyeas56ltuqhg1/P-PeTt1Llz

my second build with liveview freeze workaround, some other updates from the bitbucket repo and g3ggo's latest raw_rec module for testing that outputs MLV format (Some people reported higher write speeds with this build)
https://www.dropbox.com/sh/2358mdjq2msnivd/9jZlmta0Fb

Just select one build you want and download and extract (if you downloaded a zip file) all the files to your CF card

USE AT YOUR OWN RISK

Hey 1%, while we get you a 7d is there any way i can help you get the 7d firmware? You mean a firmware update from canon or some dump from the camera?

stevelucky read from here http://www.magiclantern.fm/forum/index.php?topic=7122.0

[Alternative build using the latest raw_rec from g3gg0:]

Alternative build using the latest raw_rec from g3gg0:

- Write speed increased (Can anyone confirm?)

https://www.dropbox.com/sh/2358mdjq2msnivd/9jZlmta0Fb

Dont forget to copy all the files and enable the raw_rec and ime_base modules or it wont work.

USE AT YOUR OWN RISK

I think we need to find the reslock stub

[Compiled under linux with latest changes:]

Compiled under linux with latest changes:

- Raw video works good
- Removed autoexpo.mo because of some conflicts.

https://www.dropbox.com/sh/ifyeas56ltuqhg1/P-PeTt1Llz

USE AT YOUR OWN RISK

I get 3 frames from the end of the video written in the begining on frame 2, 3 and 4... everything else is perfect

I did the same thing but with hudson/magic-lantern repo and under cgywin, compiling was a lot slower than on linux and also noticed some weird alerts while compiling. But this is great news man, thanks a lot!

Wow this really works! So i guess i wasn't wrong but probably compiling on cgywin screwed something Can you share more details on how you compiled?

Do you manage to compile the auto ETTR update?

Forgot that one!, done and updated!

I don't know how to delete this post lol

[Here is autoexec.bin with latest changes:]

WOW great news! I'd be happy to donate some money for developing this monster

Big thanks to arrinkiiii for your help coordinating, to Magic 7D for your great donation and to 1% for your awesome brain

Here is autoexec.bin with latest changes:
- Menu: experimental dynamic page that shows modified settings [usability] (Thanks a1ex!)
- Updated autoexpo and ettr modules (Thanks pravdomil and a1ex!)

https://www.dropbox.com/sh/ifyeas56ltuqhg1/P-PeTt1Llz

USE AT YOUR OWN RISK