Messages

Alright since I wanted to use my EOS M as a secondary timelapse camera I wanted to fix ETTR on it and finally dive into ML hacking. Here are a few things I discovered so far:

• auto_ettr_step_lv_slow() and auto_ettr_on_request_task_slow() work, the quick fix would be to use them instead of the *_fast functions, but that's not what I want
• the *_fast functions use CBR_VSYNC_SETPARAM to modify the exposures for each frame
• in auto_ettr_vsync_cbr() the exposure is adjusted for each frame and on my 6D changes are visible immediately, on the M however, exposure is smoothly changed after around 1 second of delay, get_frame_iso() and get_frame_shutter_timer() still return the old value from the previous frame even though set_frame_iso() and set_frame_shutter_timer() have been called
• the *_fast functions expect the exposure to change within 2 frames, it does not happen for some reason and so it over-compensates in each step
• HDR video's hdr_step() uses a similar mechanism to change the exposure for each frame, although not exactly the same because it's not a module, ISO seems to change correctly and immediately for each frame

I'm stuck here... the quick fix would be to use the *_slow functions obviously, but it must be possible to fix auto_ettr_vsync_cbr(). I'd really appreciate it if someone pointed me into the right direction as this is my first time diving into ML code and I have no clue where to start. Is it possible that LV just behaves differently in movie mode? Can I put the camera into this mode if that's true? What other possibilities should I look into?

I tracked down this bug today...

Where are the values of the Prop_shooting_mode defined?

M is not being picked up correctly... both of the shooting mode props are 3

P is 0
TV is 1
AE is 2
M is 3

This is why Auto ETTR is broken... also Audio remote shot works the same way as 700D et all.

heh, this matches other cameras and it works fine in AE/TV... wtf

This is just a wild guess, but maybe it's always "3" because the EOS M chooses the mode in software and the hardware switch only has Auto, Manual and Video positions?

[The Problem:]

Hey guys,

I've been following this thread since the beginning and I've been using the Tragic Lantern nightlies on http://tl.bot-fly.com/
As of this post, I'm using the Nov 07 Build. My interest in ML for the EOSM is almost exclusively for enabling time lapse creation. In particular, I'm interested in using the EOSM for day to night time lapse using the the "Flicker Free ETTR Timelapse Workflow" here: (http://www.magiclantern.fm/forum/index.php?topic=5705.0). It's not written for the EOSM but in theory it should work the same.

Auto ETTR works OK for the most part on the EOSM but it hunts for the proper exposure a lot. Tweaking the minimum shutter speed or SNR limits will usually allow you to get it to hit the right exposure but it's usually by trial and error. It's also much slower to converge than Auto ETTR on my 6D (1%'s Oct 14 1.1.3 Build) Once the exposure converges, ETTR on the EOSM seems to do what it is supposed to do for non time lapse shooting (e.g. pressing the shutter button manually with your finger).

The Problem: The bigger issue comes up when I try to use Auto ETTR in combination with the ML intervalometer (Tested with 10-30 second intervals). Once the intervalometer is enabled, Auto ETTR no longer functions. The exposure will remain the same (at whatever it started with) throughout the time lapse sequence, regardless of any changes in light. For example, if Auto ETTR picks ISO 100, 1/125 @ f/16 for the first exposure in bright sunlight, it will remain at that exposure throughout sunset and into night. I've also tried testing this using a variable ND filter to simulate the ambient light change.

I have tested this problem extensively and on various builds and the behavior is always the same when the intervalometer is enabled. I've also tried the other Auto ETTR settings other than Always On such as Auto Snap and even Half S DblClck and manually double pressing the shutter half-way between intervals to try to get it to start ETTRing but it doesn't do anything; it's as if Auto ETTR is disabled.

Any thoughts?

If Auto ETTR worked with the intervalometer, the EOSM would be the ultimate compact timelapse machine.

Any news on this? I'm having the same problem and it works fine on my 6D. I want to use the M as a secondary camera when shooting timelapses. I'm willing to test and send logs!

After messing with Lightroom and LRTimelapse a long time (export to DNG with corected WB, then normal LRTimelapse workflow) and then manually removing some remaining flicker, I ended up with this:

18 EV ramp down and up. There were a few unnecessary adjustments, especially when adjusting ISO along with shutter speed, but nothing you can't correct.

I'll try again this weekend, weather permitting, and use the sidecar files before converting to DNG.

Thanks, yes I did use ML post deflicker but I couldn't get it to underexpose during the night - it was trying to brighten up the images considerably resulting in a lot of noise so I started using LRTimelapse which does luminance analysis and exposure compensation as well, or did I miss something?

Here are the sidecar files:
https://www.dropbox.com/s/m2pzmchfl1en38q/IMG_6295.XMP
https://www.dropbox.com/s/m3zmhhvr1svufu5/IMG_6296.XMP

In any case, the EXIF shows that these two RAW files should have a similar histogram because they are exposed exactly the same. Could it be that some in camera post processing like High ISO speed NR, Auto Light Optimizer or something is to blame? I think I turned everything off but there might be something I missed.

Is it possible to modify the ETTR module so it only adjusts in full stops and not in thirds? Or even better, make it delay exposure changes a frame or two so there are no unnecessary changes in the exposure settings, that would make it considerably easier to post process manually. If there's a way to make this work perfectly it'd be fire and forget for holy grail timelapses.

Hi! I tried this feature last weekend and it worked well - it kept ETTRing all night and it resulted in this:

http://www.youtube.com/watch?v=Fe3Jw1wUxQA

post processed using LRTimelapse, After Effects and GBDeflicker but as you can see I was not fully able to remove flickering, during the transitions there's still some flicker left and yes, I adjusted the frames that have different exposure settings so there's no flickering but here's my problem: two frames with exactly the same exposure have a different histogram. I'll attach those two raw files so you can see:

https://www.dropbox.com/s/q2ns6wv7vbc2pis/IMG_6296.CR2
https://www.dropbox.com/s/6kxr5msgzls3pf9/IMG_6295.CR2

As you can see shutter speed was slow so it should not be caused by shutter flicker neither should it be aperture flicker because I unscrewed the lens while holding down the DOF button. Of course I can fix these two frames manually but there are many like these and they are hard to find, and neither GBDeflicker nor LRTimelapse were able to detect and remove it automatically. It's driving me crazy - is there some setting I forgot?

I was using a 650D with a EF 10-22mm lens.

Hi!

I've been using the latest alpha for a while and I noticed two corrupt RAW files when taking time lapses. Here is one:

https://dl.dropboxusercontent.com/u/10119318/PublicPrivate/IMG_3322.CR2

I'm not sure if ML is the cause, the only know that this did not happen before I started using ML. I left live view before starting the time lapse shots and I don't have any special settings enabled (no Zebras, no Focus Peak, no Magic Zoom, no Cropmarks, no Ghost image, no Spotmeter, etc.) Let me know if you need more information or if this definitely does not have anything to do with ML.

EDIT: I just found another one: https://dl.dropboxusercontent.com/u/10119318/PublicPrivate/IMG_3121.CR2

Count me in as well!