Canon 1100D / T3

Started by a1ex, June 14, 2012, 04:50:54 PM

Previous topic - Next topic

0 Members and 1 Guest are viewing this topic.

AneMotion

I'm sorry guys IF my question maybe offends you, BUT I just have to find out somehow.
I'm not a coder, but would study it in order to make ML work on Canon 1200D.

Does anyone know IF the 1200D is the replacement for the older 1100D Canon?
If so, would the 1100D ML work on the 1200D?

If not, is there somebody who would create ML for the Canon 1200D, PLEASE!!???


haulric

Hello there!

First things thanks for your time.
I've spent a good part of my evening here trying to install ML on my 1100D but without success, I get the update file error :

Update file error.
Please check the memory card and reload the battery and try again.


I've :
* Formated my sd card (low level)
* use a 4Gb sd-card
* Executed the make_bootable script
* Tryed with three different build (latest, april and one from march).
* Updated my canon firmware to 1.0.5
* Re-downloaded firmware and try again.

My guess is that maybe there have been a new sub-build from canon for the 1.0.5 which is not compatible with ML (?) if so maybe did someone still have the canon firmware he used to upgrade before installing ML ?

Thanks in advance.
Excuse my engrish. :-X

Walter Schulz

Format card in cardreader.
Format card in cam.
Copy extracted nightly build contents to card.
Run firmware update.

Report back.

haulric

Wow it works, thanks very much!

hugo57

Hello everyone  :) I have one problem with ML.
I can't see ML's HUD on the LV, but this problem is new, I do not have it a week ago.
I need your help someone
I have the last bluid(magiclantern-Nightly.2015May03.1100D105.zip) for information

DonJuan18

Hi everyone
I just downloaded ML on my 1100D and it works fine :)
Now i really want to use a external monitor but my monitor only supports that old AV.
I can order that AV-400 cable for eos but on all sites the specifications says that 1100D is not supported.
Do you guys think that ML gave my camera the option to use AV output. Or can software don't change things like that ?

Greets

ejfuhr

just installed magiclantern-Nightly.2015May03.1100D105 - a build from over a year ago. the current builds are "BUILD FAILED". any word on a downloadable executable that  will work?
i would try but am terrible in C and cygwin

thanks

ejfuhr

mathias

try it with other memory card.
if same error ocurrs the problem is hardware not ml

dfort

Hey 1100D / T3 ML users. Saw this in another topic:

Quote from: a1ex on December 16, 2016, 09:41:16 PM
- the raw buffer is not allocated all the time (as it's done on the new models) -> easy to fix, see the 1100D raw branch

There's a working branch of the 1100D? I thought it has been broken for over a year!

Nobody ran this test yet?

http://www.magiclantern.fm/forum/index.php?topic=1009.msg146321#msg146321

Floont

I thought the 1100D development was dead, too.  I missed the earlier request from a1ex;  here are my results:

(magiclantern-Nightly.2015May03.1100D105, clean fresh install)

First, 5x zoom had no effect on the timer values I observed.

In photo mode: after a fresh boot, timers were 0 (-1), 0 (-1).  Switching to photo mode after being in video mode, timers remained as they were in the video mode.

FPS override off:
1280x720@30: 960 (+0), 1112 (+0)
1280x720@25: 1000 (+0), 1280 (+0)

FPS override @35:
1280x720@30: 872 (-88), 1050 (-62)
1280x720@25: 872 (x0.87), 1050 (x0.82) <- this was the only setting that showed 'x' instead of '+' or '-'; also the only one with decimals

Let me know if there are other tests to be run, and I'll try to help.  I'd love to see a new version.

dfort

Great. Don't know what to do with that but hopefully we'll have 1100D back on the nightly builds.

I made some test builds for testing 10/12bit raw and included a build for the 1100D. Nothing is fixed yet and the new features are probably not working but we need to start somewhere. Let's see if we can breath some new life into this camera.

For brave testers only:

https://bitbucket.org/daniel_fort/magic-lantern/downloads/magiclantern-raw_video_10bit_12bit.2016Dec18.1100D105.zip

Sorry -- needs some more work on it before it is ready for testing.

dmilligan

How did you make this build? Raw video does not work on 1100D except on a very old experimental branch. That branch no longer cleanly merges with current unified or other recent developments like 10/12 bit.

The 1100D nightly build was intentionally disabled because a1ex noticed some incorrect values for FPS override timers. I've never cared enough to try and fix it (the Maths involved hurt my head :P ), I just compile it with the old incorrect values.

dfort

Well if it the math hurts dmilligan's head then it must be way over my head!

I just compiled it with the incorrect FPS timers and didn't notice the missing modules. However, I was following this lead:

Quote from: a1ex- the raw buffer is not allocated all the time (as it's done on the new models) -> easy to fix, see the 1100D raw branch

So it looks like I need to hunt down an old branch to find out what he means. Do you happen to recall the approximate vintage of that branch?

dmilligan


dfort

@dmilligan
Thanks -- again! :D

FelixJongleur42

As I understand it: 1100D is still intentionally broken by undefining TG_FREQ_BASE as it seems.
But this also breaks "unified" (?) build, because reboot-all.c needs 1100D.105/magiclantern.bin.

As for now, I put
#undef FEATURE_FPS_OVERRIDE
in
platform/1100D.105/features.h
and I re-enabled the wrong
#define TG_FREQ_BASE 32070000
in
src/fps-engio.c
in my sandbox in order to be able to build "unified".

Basically, I wanted to try a unified build for two models I have, 60D and 550D. Am I correct that this is not possible at the moment?
Any suggestions where I can put in some work? I am willing to help and I know C and make.
I am still working my way through the build process - as for "unified" - that should possibly be another thread...

Straight_Shooter

Quote from: dmilligan on December 19, 2016, 02:41:02 AM
The 1100D nightly build was intentionally disabled because a1ex noticed some incorrect values for FPS override timers. I've never cared enough to try and fix it (the Maths involved hurt my head :P ), I just compile it with the old incorrect values.

Hi dmilligan,

being both a software developer and an owner of the 1100D, I would be willing to help here. However I have no idea what exactly needs to be done with the values found by Floont in Reply #1410. You mention something about complicated maths. Is there some formula that we need to compute certain values with? Could you please point me in the right direction about what needs to be done? Then maybe I can figure out a way how to do it.

dmilligan

Study the FPS code and understand how it works (the ML code itself is basically the best documentation we have on the hardware). That's what I don't have time to do.

dfort

@Straight_Shooter

Wow, another developer with a 1100D. There are multiple issues going on that platform. Here's one we're trying to solve but keep getting lost tracing lv_raw_dump:

http://www.magiclantern.fm/forum/index.php?topic=5601.msg176935#msg176935

a1ex hinted that solving the 1100D might help with other platforms:

http://www.magiclantern.fm/forum/index.php?topic=5601.msg176545#msg176545

The math problem headache we're talking about here is:

src/fps-engio.c
#elif defined(CONFIG_1100D)
    #define NEW_FPS_METHOD 1
    #undef TG_FREQ_BASE
    // #define TG_FREQ_BASE 32070000 - incorrect, see http://www.magiclantern.fm/forum/index.php?topic=1009.msg146321#msg146321
    #undef FPS_TIMER_A_MIN
    #define FPS_TIMER_A_MIN (ZOOM ? 940 : 872)
    #undef FPS_TIMER_B_MIN
    #define FPS_TIMER_B_MIN 1050
    #define SENSOR_TIMING_TABLE MEM(0xce98)
    #define VIDEO_PARAMETERS_SRC_3 0x70C0C


We had a similar issue with the EOSM that was resolved. I'm not a developer so I can't tell you exactly how it was figured out but the discussion starts around here:

https://www.magiclantern.fm/forum/index.php?topic=14959.msg148768#msg148768

Straight_Shooter

From what little I know about this issue, I seem to understand that the value for TG_FREQ_BASE is not correct. To solve this, somehow a better value needs to be obtained. But I have zero idea what exactly needs to be done to achieve that, hence my question.

What kind of computation is needed here? Is this explained at some point in the code, maybe in a comment?
I guess I'll try having a look at the file: fps-engio.c

dfort

Check Reply #1429. It points to a test a1ex asked to be run which someone did run and posted the results. How to calculate the correct TG_FREQ_BASE is beyond me but last time I said that was because I didn't look at the code carefully enough.

a1ex

FYI, there is an experimental build for 1100D with raw video support over here: https://builds.magiclantern.fm/experiments.html

It is incomplete, so you are welcome to fill in the missing bits. Some of them can be found in this thread, just follow dfort's links.

With this build, please check whether the FPS values displayed by ML match the value selected from Canon menu, in all video modes. Also check whether selecting 35 fps in ML menu gives valid image in all video modes.

Metronome

Hi. First timer here.

I updated my 1100D firmware to 1.0.6 (14 Nov 2016 release). I just want to ask if the latest ML build for 1100D "magiclantern-Nightly.2015May03.1100D105" will work with the latest 1.0.6 firmware? Or is it strictly for 1.0.5 only? Thank you.

dfort

1.0.5 only. You can find an archived firmware on Pelican's website. The 1100D.105 is here:

http://pel.hu/down/eos1100d-v105.zip