Messages

Just a strange bug in the broken3x-12bit build - I've seen it twice with different cards. It destroys the \ML\modules\ folder (or does something wrong to the filesystem). Finally it fails with "Can't find 5D2_212.sym"

Warp Stabilizer plays its own tricks: in the static scene about 3 min. 10 sec. the mountain is breathing. Mocha would be much more controllable.

I see a problem with ghost images in this build. It shows the overlay, but immediately prints error log.
Looks like it was resolved in Unified. https://bitbucket.org/hudson/magic-lantern/commits/e0c1b48a2558975e23545eaeea9eced6de72ab97

Code Select Expand

ML ASSERT:
count <= 8192
at ../../src/fio-ml.c:528 (FIO_ReadFile), task cls_task
lv:1 mode:3


Magic Lantern version : raw_video_10bit_12bit.2017Jul03.600D102
Mercurial changeset   : c38f583da163+83173fc0953b+ (raw_video_10bit_12bit) tip
Built on 2017-07-02 22:50:32 UTC by jenkins@nightly.
Free Memory  : 208K + 795K

It sounds quite on the contrary to the existing solution but...

We know how programmable EMF chips work – they just return several predefined values (f-number, FL, focus prompt compensation). Instead, let's imagine one could hack this values directly in-camera. We'll get immediately working focus confirmation (+ focus correction) and enough EXIF data to identify the lens (you can add the proper lens names with Exiftool later).

As for the AF-MF switch - it's nothing but a software switch. It's easily overridden (with API commands, for example). Newer EMF chips simulate AF mode now.

Maybe the whole scene should be set up beforehand. Try to follow the risky but successful way of Jonny Armstrong.

http://strobist.blogspot.ru/2015/09/on-assignment-studio-in-wild.html
www.jonnyarmstrong.com

A plenty of broken hardware and a bunch of brilliant photos )

Defining Focus End Point works OK - it reports the same great number of steps as computer. But using left-right arrows (FF+/FF- to be sure) I can pass the same distance in a few clicks (3 times less steps). I guess something is wrong while reading stepsize/step_size variable. Or it just works how it was intended?

Upd
Finally I've solved the problem after resetting all CFn from canon menu. Strange, but setting AF to anything but halfshutter gives such unpredictable focus steps.

> ML Menu > Focus > Focus Settings > Step Size
It's already set to 1.

P.S. I'll do some deeper testing later.

I've just compared focus step size while shooting tethered and using follow focus. Shooting from PC gives 3 times smaller steps, compared to ML follow focus. If I'm not mistaken, the step size was defined as the minimal step size from canon API. Maybe it's an important update or just another command for smaller steps. Is there any way to sniffer that command from PC or from ML? It may have a great value for macro photography.
600D + Tamron 60mm f/2.0
Software: CaptureOne 9.2 (with latest canon API, I believe)

I'm not very familiar with Lightroom. Comparing CaptureOne and Adobe Camera Raw I'm sure that C1 is much more powerful. Noise removal is perfect, shadow/highlights recovery looks more natural etc. It's functionality is regularly updated. Can't say this about Adobe.
But be careful with DNG's: dng editing is a bit limited in C1 - it's always recommended to use native RAW formats.

Yes, it costs much, but 3 month trial may be a nice experience. Even more: you can install C1 v8, use it 3 months and then upgrade to v9 and get additional 3 months of trial.

[Focus Infinity Module]

You can use Focus Infinity Module. This complied binary works OK with the latest nightly. Finally it's your choice which distance from "infinity" to set (according to what you consider to be your hyperfocal).

The module requires Live View ON, doesn't require Follow Focus ON. Use recommended step sizes and delays.

It seems, that the problem with Sound Recorder (discussed here) was solved by Rufustfirefly. The patch was published 8 months ago (https://gist.github.com/jbuchbinder/2af8856ca1e8276210d7). Is it possible, to apply the patch the main branch? Unfortunately, no one reported success or fail testing it.