Topics

Santa delivered this year and I'm now the proud owner of an Irix F/4 11mm Blackstone lens.

This lens has the widest FOV with the lowest distortion of any lens I'm aware of.

In case anyone is interested in this lens, I've posted a few insights on my blog:http://photography.grayheron.net/2017/12/understanding-irix-11mm-depth-of-field.html

I'm writing this short post in the spirit of passing on my humble learning, in order to help others trying to script.

As 'the ML gurus' know I am far from a scripting expert, nevertheless, through a lot of experimentation, and guru support, I've picked up enough to 'get by' in scripting.

BTW I noticed lately a few 'negative' or 'curt' comments appearing on the forum, and this saddens me; as 'the gurus' give their time freely and people like me have a lot to learn from them. I hope others posting respect the hardwork of 'the gurus'.

As for the reason to post this entry, it was to feedback that scripting is all well and good,  ie getting the syntax and 'equations' right; however, a good script can be let down by a poor/weak User Interface (UI): which I think covers: usability and user experience.

Having now reached the 'script and maths' is right stage, I realized my two main scripts (The Focus Bar and The Toggler) were letting themselves down wrt their UI.

Of course I can only talk about this through my (old) eyes and how my brain works, but I was clear to me (pun intended) that the UI needed attention. For example, font size and font colours (foreground and background), plus message positioning.

I've therefore attempted to get a 'better' user experience and welcome any feedback on what I've tried to do.

The Toggler is at https://gist.github.com/pigeonhill/508a3722b9dc319b64ca4c5972b9e50b

The Focus Bar is at https://gist.github.com/pigeonhill/cd1dbe0191ca1264fc0aa07aaafb515d

As I've spoken a lot about the technical side of my Focus Bar, ie using blur info, I thought I would give a simple demonstration.

In this case how the focus bar helped me capture a 12 image non-macro focus bracket set: http://photography.grayheron.net/2017/12/extreme-focus-stacking.html

Hope someone can educate me.

At the moment one of my scripts, on loading for the first time, uses an IF-THEN-ELSE, which first detects that a lens is AF capable, and if it is it carries on with the script, ie loading variables, menus and function.

If the lens is not an AF one, the script simply shows a console message and does nothing else.

This means I need to put the main part of my script in the (main) IF-THEN-ELSE, like this:

Code Select Expand


IF test-true THEN
carry on with main script
ELSE
console-message
END


Is there a way to put a test, up front in a script, that checks things and then doesn't load the script?

Cheers

Garry

PS having been prompted by @dmilligan, I hope this is an X question, rather than a Y one 

@dmilligan @anyone

I wish to hide a menu item from users and am using the following:

Code Select Expand

        {
            name = "update",
            help = "Not under user control",
            choices = {CON,COFF},
            value = COFF, -- start up default
            hidden = true
        }

However, the menu item, ie update, remains shown.

Any ideas where I'm going wrong.

Cheers

Garry

Use Case: look out for changes in a specific named menu or sub-menu.

Why: one script changes the state of another and this needs to be detected.

Suggested format: extend lua module property to handle the 'ML side' of life or even an event handler.

Although I have picked up a bit of scripting knowledge, thanks to the help I've received on the ML Forum, the one area I'm weak in is io.

For example, what is the easiest/simplest way to have my script's 'state', ie as set in the script's menus, recorded like the menu states in the ML core, ie at camera switch off. So my script starts in the same configuration as I left it at camera switch off, rather than a hard coded state.

I would welcome some hints or pointers.

Cheers

Garry

Just in case anyone else has found my Toggler script useful, here is an update that works with the latest Lua-fix nightly (Latest Build (2017-12-03 22:09)).

Although I wrote the Toggler for the EOSM, which is lean on buttons, the Toggler should work on other cameras using the Build above.

You can download Toggler from here: https://gist.github.com/pigeonhill/156ea0e55cd0a029bc447f8bf5593224

Why use Toggler?

If you find yourself switching ML states a lot, then Toggler could help.

The script you download is configured to use the PLAY, INFO and MENU buttons:
[MENU] = normal use of [MENU] if not in imode (interactive mode)
[INFO] = normal use of [INFO] in not in imode
[PLAY][PLAY] = normal use of [PLAY]
[PLAY][MENU] = enter Toggler interactive mode
In imode [PLAY] and [INFO] will toggle back and forwards through your ML condition states giving you an on-screen message
In imode [MENU] will set that condition and leave interactive mode.
In imode pressing any other key, eg [SET] (not [MENU], [INFO] or [PLAY]) will leave interactive mode

You can change the script to use other button combinations, ie that work better on your camera (the 'default' ones work well on the EOSM).

The script, as I've written it for my use, allows me to switch my Focus Bar on and off, switch the FB stacking mode on and off, open the ML menu, switch on ETTR or Auto-Bracketing and the list goes on.

As usual I welcome feedback, especially to make the Toggler idea better and make my coding better 

We have close() to close the ML menu, but unless I've missed it, we don't have open(), ie to open the ML menu from a script.

On the next Lua fix update it would be great to see this feature.

Cheers

Garry

I am running two EOSMs with two identical scripts.

But they respond differently and I believe it to be how the scripts are loaded, as I'm using keypress event handlers in both scripts.

On one camera script A is listed first and on the other script B.

Is there a simple way to ensure which scripts load first?

Cheers

Garry

I have two scripts running and I need to get one script interacting with another.

What I have tried, but it doesn't seem to work, is to simply get one script to change the menu settings of the other script. But nothing seems to happen.

Here is the code in one script, with Focus Bar being the ML menu created by the other script:

Code Select Expand

        if menu.get("Focus Bar","Focus Bar") == "ON" then
            menu.set("Focus Bar","Focus Bar","OFF")
        elseif menu.get("Focus Bar","Focus Bar") == "OFF" then
            menu.set("Focus Bar","Focus Bar","ON")
        end

Is the problem that Lua can't change a script created ML Menu?

Cheers

Garry

For those interested in what you can do with an EOSM, you may be interested in my lasted post: http://photography.grayheron.net/2017/12/thank-you-sony.html

Ok I know I'm sometimes stupid, but I'm at a loss.

Just got another EOSM and I've reset all camera settings, tried different lens and with and without ML card the focus distance remains fixed on one number.

Does anyone have any ideas, ie seen this before?

Cheers

Garry

Using Lua property one can 'detect' changes in the shutter, aperture and ISO.

Could we also add a property to detect changes to focus distance and focal length?

Cheers

Garry

Currently we can control Mirror lock-up in two ways.

Always on responds as per Canon, i.e. Two shutter presses, with the first 'just' moving the mirror.

The second being a delay when using the intervalometer.

Some other cameras allow the shooter to select a delay between the mirror going up and the shutter activating; thus reducing mirror vibration issues when not in LV mode. 

So a thought. Can we introduce a user selectable delay in the ML mirror menu, say in 1/10 sec up to 1sec.

And/or bring in mirror control in Lua, unless it's there and I've missed it. My idea being, via the shutter property and/or keypress event, control the mirror and shutter through Lua.

@A1ex or anyone

I'm trying to make use of a long halfshutter and need to differentiate it from a LV AF action.

I thought I could use the Lua property handler, eg accessing LV_FOCUS_STATE or another LV_FOCUS property, ie to handle the AF condition.

But the LV_FOCUS property handler always seems to return 0.

Does this mean it is not yet exposed in Lua?

Cheers

Garry

I've pulled my hair out all afternoon.

I have two EOSMs and one, all of a sudden, exhibited a strange behavior: the Aperture/Exposure button stopped functioning, ie I could not use it as a right toggle in menus (ML or Canon) and could not toggle between aperture and shutter.

I've tried different lens: no difference.

I've removed ML, via firmware reset: no difference.

Has anyone else had a similar problem and fixed it!

Cheers

Garry

I have a script that draws to the screen and all is well, until I switch to Canon menus, eg INFO or MENU.

I'm detecting these screens via a property change, ie:

Code Select Expand

function property.GUI_STATE:handler(value)
    if value == 1 then
        fb_dirty = true
    else
        fb_dirty = false
    end
end

Both MENU and INFO return the value 1.

The issue I have is how to decouple me not (sic) drawing and the Canon side drawing MENU or INFO screens.

Has anyone worked out how to manage the Lua and Canon side graphics? That is cleanly doing one or the other.

@A1ex or anyone

Is there a way to detect that the console is not showing, eg the start-up screen has finished.

I just had a strange occurrence when using my 50mm 1.8 II, with my 5D3.

All my lenses work with my focus bar script, however, the 50mm generates an Err 70.

If I switch off my script all is OK.

Here is the CRASH LOG.

Code Select Expand

ASSERT: hLvJob->hJpegMemSuite
at ./Epp/Vram/VramStage.c:891, task Epp
lv:1 mode:3


Magic Lantern version : lua_fix.2017May01.5D3123
Mercurial changeset   : ea6d752aee38+744f5868a308+ (lua_fix) tip
Built on 2017-05-01 16:16:09 UTC by jenkins@nightly.
Free Memory  : 130K + 3399K


Does anyone have any ideas?

Cheers

Garry

@A1ex

Hope you can help.

In my focus bar script I'm writing text to the screen then, when not required, erasing.

I erase by simply overwriting in the transparent color.

Unfortunately this negatively interactions with ML drawing, eg the spot meter.

Is there a way to deal with this, ie after I've erased then ML, eg the spotmeter, doesn't know I was writing at that point.

Hope this s clear.

Cheers

Garry

@anyone

When I get a chance I will attempt to measure this myself, but until then, does anyone know th reference plane for the focus distance reporting?

That is the sensor or the lens principal plane?

I know it won't make that much difference in calculating depth of field etc, but I'm being picky ;-)

Cheers

Garry

Don't know if anyone else has noticed, but the Lua API documentation is broken.

I thought some may be interested in this claim: ie a world's first

http://flip.it/98nASR

@A1ex

I saw on another camera that is was possible in Bulb mode to monitor the 'liveview' as the exposure 'developed'.

For LE work this seams a great feature, as it is often 'hit and miss' that you have an optimum exposure.

I was therefore wondering if it was possible in ML to consider a feature that provides realtime feedback on the exposure, say in bulb mode.

The feedback would ideally be a LV image and the RAW histogram. Or even 'just' the RAW ML histogram displayed on a blank LV screen.

Do you think the idea is workable? That is for an ML C coder ;-)

Cheers

Garry

Thanks to @dfort I managed to compile on my laptop and prove that my pull request to fix the DoFs had an error, ie:

Code Select Expand

   const uint64_t  temp = lens_info.aperture*coc*(fd-fl) // note aperture and coc in tenths of their units, hence the 100000 factor below

should be:

Code Select Expand

   const uint64_t  temp = lens_info.aperture*coc*(fd-fl); // note aperture and coc in tenths of their units, hence the 100000 factor below

But I still 'dont get' how to correct pull requests and I'm not going to keep randomly trying things out.

Bluntly, I'm personally OK, ie I have my focus bar up and running on all my cameras with the optimum DoF info and I love it.

But I can't work out how to help others, eg I made a pull request the other week (albeit with a silly ";" error), but I just don't have the confidence/skills to do any more.

For now I think I'll stop posting on all 'my problems' and watch out to see how others get on.

Bottom line: thanks Dan, but for now I think I go silent for a while ;-). Plus I'm away on business this weekend for a week, ie no laptop.

Those that track 'my fun', will know I've implemented a focus bar in Lua.

In this post I'm releasing an enhanced focusing feature that allows you to get optimum focus stacking.

The only requirement is that the lens should report focus distance (which is assumed to be 'accurate').

As the current ML DoF reporting is broken (pull request submitted), I've made this release of the FOCUS script self contained, ie it does not rely on ML DoF info,

The script can be downloaded from here: https://gist.github.com/pigeonhill/cd1dbe0191ca1264fc0aa07aaafb515d

The use the focus stacking/bracketing feature, simply switch it to DoF mode (option -1). The script default for focus stacking is on (you can reverse this if you wish).

Here is a screen dump having just taken an image:



The focus bar is telling us what the optical (sic) DoFs are, and the script (at the bottom) is telling what the diffraction corrected DoFs, ie where the ML DoFs would normally be (switched off when running the script). The focus bar end (in DoF mode) show the 'worst case' DoFs, ie the optical ones that ignore diffraction. The green dots are the diffraction aware DoFs and the red dot the focus point.

In focus stacking mode, the DoFs at the time the image is taken are captured, so that we can compare these with other DoFs as we refocus.

Here is a screen dump showing refocusing towards the infinity end:



And here is a screen grab if we focus the other way, ie towards the macro end:



In both cases, the magenta shows the amount of focus overlap, which you will need to ensure is visible, ie  if no magenta is showing then you don't have focus overlap ;-)

Bottom line: I hope some find the focus stacking feature of value. Although I wrote it for 'landscape photographers', because you can use it 'both ways', ie relative to the image you are taking, it may be of use to others.

As usual, I welcome feedback.

Cheers

Garry

@dmilligan

I'm a little stuck as I need to detect when an image has been taken, so I can record some variables for later use.

Can you suggest the best way in Lua to detect an image has been taken. I can't see how to do it with event or property.

Or maybe you can't do it, i.e. a feature yet to be realized.

Cheers

Garry

Looking into ML chdk-gui_draw.c tells me that the ML draw coding is sourced from CHDK.

Looking at the coding, which is in C, and I'm not a C programmer, tells me it should be relatively simple to add in a few enhancements, which then can be exposed in Lua.

For example, the draw_circle is coded like this:

Code Select Expand

void draw_circle(int x, int y, int r, int cl)
{
    uint8_t* bvram = bmp_vram();

    int dx = 0;
    int dy = r;
    int p=(3-(r<<1));

    do {
        bmp_putpixel_fast(bvram,(x+dx),(y+dy),cl);
        bmp_putpixel_fast(bvram,(x+dy),(y+dx),cl);
        bmp_putpixel_fast(bvram,(x+dy),(y-dx),cl);
        bmp_putpixel_fast(bvram,(x+dx),(y-dy),cl);
        bmp_putpixel_fast(bvram,(x-dx),(y-dy),cl);
        bmp_putpixel_fast(bvram,(x-dy),(y-dx),cl);
        bmp_putpixel_fast(bvram,(x-dy),(y+dx),cl);
        bmp_putpixel_fast(bvram,(x-dx),(y+dy),cl);

        ++dx;

        if (p<0)
            p += ((dx<<2)+6);
        else {
            --dy;
            p += (((dx-dy)<<2)+10);
        }
    } while (dx<=dy);
}

I've tried doing this myself, but I remain mystified at some of the C coding.

Also why do we have to duplicate certain things, eg like this:

Code Select Expand

        bmp_putpixel_fast(bvram,(x+dx),(y+dy),cl);
        bmp_putpixel_fast(bvram,(x+dy),(y+dx),cl);

My 'idea' was to use draw_circle to create a function that draws an arc, ie from angle a to angle b.

Looking at the existing coding, my thought was to simply use the existing draw_circle coding and 'simply' put in a test for the colour, ie cl.

If that part of the circle was outside the drawing angles, then don't draw or draw transparent.

Anyway, I'm afraid at this time, in my C development, I am unable to do this.

But the feature request remains: can we introduce a draw_arc into the ML chdk-gui_draw.c, and expose this in Lua?

Cheers

Garry

In case anyone, like me, uses ML supplied DoFs, which are not working as they should at the moment, here is a version of my focus bar that displays the DoFs correctly.

https://gist.github.com/pigeonhill/e6c40fe4c9a8d9941123c0f3e6b664ec

Make sure you switch off the ML DoFs.

Otherwise use this script as instructed in the script.

The DoFs are displayed where they normally would be.

This is a temp fix: a pull request has been submitted.

Cheers

Garry

@a1ex, @dmilligan

As I finalize my focus bar work, I'm noticed some 'strangeness' that points to Lua and core ML potentially being different.

I see the differences when calculating the near and far dofs and comparing these with the ML calculated ones.

The equations are the same and I'm concluding that, other than rounding errors, the ML reported dofs may use different (?) source data for certain key variables.

For example, ML calculated dofs use lens_info.x, where x could be, say, aperture, ie lens_info.aperture. Or focal length or focus distance.

Lua gives me access to the aperture data via camera.aperture.value

The question in my mind is, at any given time, are these variables created from the same source? That is Lua camera.aperture.value = ML C lens_info.aperture.

Cheers

Garry

In case someone is interested, I came across this today, that might be useful if one is trying to calibrate mechanical shutter action.

https://www.camsst.com/cam-lite

Now I have the focus bar running (on my 5D3 and EOSM), I needed to tweak the EOSM Toggler to ensure script-2-script compatibility in the UI.

The EOSM Togger was written specifically for the EOSM, ie this version is not a generic Toggler.

The updated EOSM Toggler (https://gist.github.com/pigeonhill/156ea0e55cd0a029bc447f8bf5593224) now looks like this, ie with the focus bar.



In the above screen dump we see the focus bar and the Toggler in interactive mode, showing that the ML state currently selected (ready to change to) is AutoETTR. If you select this state, then the Toggler will change all the ML Menus and exit interactive mode. In interactive mode, you can go backwards and forwards through your Toggler states and have as many as you like.

The EOSM Toggler works without the focus bar.

As usual I welcome feedback.

@A1ex/@dmilligan

I've hit a final challenge in my focus bar script.

I wish to switch off the focus bar (eg switch of the display) if the ML menu is showing or the Canon Menu. I seem to be able to do this with the following snippet:

Code Select Expand

if (menu.visible or camera.state ~= 0) or (not showing) then -- don't show bar'
        if not hidden then
            display.clear()
        end
            hidden = true -- now hidden so don't need to keep hiding'
    else
            hidden = false -- show the focus bar

The variable 'showing' is a one of the focus bar inputs, ie to turn on or off the focus bar.

So far, so good.

What I can't seem to do is work out how to detect various other GUI state, eg I would like to switch the focus bar off if the user is using one (or all) the info screens, ie not just the Menu screens (Canon and ML).

I've looked into the source code but can't seem to see the info I'm after, eg the GUI state codes.

Would the above be possible...and, of course, how ;-)

Cheers

Garry

@dmilligan

David I hope you can put me out of my misery.

I'm needing to right justify some text and need to the width of a string.

I've tried using font:width(text) but I'm obviously using it wrong.

Could you help me please, ie how do I get the pixel width of a piece of text.

Sorry to be a pain.

Cheers

Garry

If you have been following my scripting, you will know I've been trying to get a focus bar running; which looks like this (based on specific setting):



Thanks to help/hints from @A1ex and @dmilligan I have now got a (near) standalone focus_bar function, ie you can use it in your scripts.

The demo script, showing how to use the focus_bar function, is here: https://gist.github.com/pigeonhill/2c0ebbf59f0b57f272a808aed9562b71

The demo script has more info about the focus_bar, so please read this before using the script and/or function.

Note: this version is 'nearly' standalone, ie you need to use one global variable and have one additional function.

A few things to note:
- I decided to keep the visualization of the focus field to a linear representation. I personally get more out of it this way.
- In addition to 'seeing' the focus field, and the fp and dofs, you can decide if you wish to see the near and far distance info shown below the bar
- For landscape photography, once the far dof is greater than 'infinity' the focus_bar distance reporting switches to total blur units (optical and diffraction in quadrature) at infinity for the far; and the dof distance based on the total infinity blur for the near dof
- Blur reporting at infinity is green if above the sensor limit (2xsensor pixels) and red below
- The multiplier allows you to see the focus field beyond the far dof
- A multiplier of -1 switches the focus_bar into dof mode. Note in all modes the focus_bar functions during Canon zoom. Hint: this mode may help portrait photographers or videographers looking to 'ensure' the focus field is 'right', ie use Canon zoom and focus on the eyes (say) and check the focus field
- A multiplier of 0 switches the focus_bar into fp mode, ie see the focus field from zero to the fp
- I think it 'best' to have ML diffraction aware on (diffraction impacted green dots showing dofs) and switch diffraction off in the focus_bar. This way you can see the impact of diffraction in the focus field. The focus field is black at unity blur (optical or with diffraction if selected)
- Unity blur is that set in ML

As usual I welcome feedback.

Cheers

Garry

I just posted a new script that I wrote for landscape photography, ie where you are seeking to nail the focus from a near field position to infinity,

This helper script uses a 'new' approach to focusing, ie managing infinity blur.

What I also included was a focus bar that visually allows you to see the focus field through the scene.

Because of the non-linear way our lenses behave when we focus, ie look at the focus scale on your lens, finding a way to specify a distance is a challenge.

The method I used was to calculate the distance at which the infinity blur equals the sensor pixel size, eg about 6.3 microns on my 5D3.

This then allows you to zoom out and now in.

To zoom just change the multiplier in the menu, ie 3 = a zoom of 1, < 3 means zoom in (0.5 or 0.25 multipliers) and > 3 means zoom out (multipliers of 2, 3, 4 etc).

This approach is a a bit clunky and will tidy this up when I move out of beta testing.

For all (sic) photographers and videographers, the focus bar in this landscape script allows you to dynamically visualise optical (ie without diffraction) focus.

As we know, diffraction is flat through the scene, while optical blur changes through the depth of the scene.

The colours tell you about the 'focus density' based on a unity blur criterion that you set in the script, eg 30 microns:

- Green = unity blur to half unity blur
- Yellow = less than half unity
- Grey = blurs above unity (Black being the max blur you wish to see, set by the user in the script)

Here is the link to the script: https://gist.github.com/pigeonhill/cd1dbe0191ca1264fc0aa07aaafb515d

Bottom line: Anyone (photographer or videographer) interested in 'seeing' what focus looks like through (sic) the scene, can use my landscape focusing helper script, ie use it as an education tool to plan better focusing strategies. It's great fun to play with the script in the comfort of your favorite chair

As usual I welcome any feedback. For instance, one idea could be to create a focus bar (alone) script that all can use. As I say, your feedback is most welcome.

NOT SUPPORTED ANYMORE; see https://www.magiclantern.fm/forum/index.php?topic=23036.0

Previously I have posted details of my Landscape Focus Helper, which now allows a 'novel' way to control infinity focus via dynamic LV access to blur data.

This beta version 3 introduces another feature to help you with focusing, namely a focus bar.

The script has all the info you need and here is a screen view of the script in action.



The script has been testing on a 5D3 and can be downlaoded from here: https://gist.github.com/pigeonhill/cd1dbe0191ca1264fc0aa07aaafb515d

This is a beta version, ie I'm still playing with ideas, so I welcome any feedback on the approach and any thoughts on making it better.

Cheers

Garry

I've just up-issued the above script and you can read about it and access it here: http://photography.grayheron.net/2017/05/landscape-focusing-helper-script.html

Basically the script fully automates moving an AF lens to the HFD.

It then provides you infinity diffraction and optical blur information so you can fine tune your focus.

As usual I welcome any feedback and/or improvement ideas.

Cheers

Garry

I hope some kind soul can put me out of my misery, as I've not used string.format before.

I've read up on line and I think I'm doing things correctly.

'All' I'm trying to do is format a string to print: here is the demo code:

Code Select Expand

b_diff = 45.6577
info_message = string.format("Blur_Diff = %f", b_diff )

But all I get back when I print is

Code Select Expand

Blur_Diff = f

I've tried %0.2f and still get the same as above.

Where am I going wrong?

In the spirit of helping fellow photographers, and because I mention ML, I thought I would share my latest post on landscape focusing: http://photography.grayheron.net/2017/04/getting-best-out-of-hyperfocal-focusing.html

I'm not sure if this is the right/best location for this post, but I'm sure it will get repositioned if required :-)

The purpose is to allow ML users the chance to flag (to the developers) potential Lua API extensions, eg accessing ML data, as accessing core Canon 'stuff', eg exposure etc, is well handled at the moment.

To kick things off, I would like to propose the following, to 'open up' the ML Spotmeter:

1. Get [read] the current ML spotmeter position in LV space, eg:
get_spotmeter.pos_x
get_spotmeter.pos_y

2. Get [read] the current ML spotmeter value (as set in the ML menu), eg:
get_spotmeter.val

3. If possible (I accept this may not be possible, eg linked to canon focus point) enable the following Put [write] capability for the spotmeter, ie move the spotmeter location, eg:
put_spotmeter.pos_x
put_spotmeter.pos_y

I hope others add to the above, as Lua has more things to offer.

Cheers

Garry

@A1ex

I have a use case that ML may (sic) be able to help with.

As a stills photographer I don't always use LV, ie I use the eye piece a lot.

I often find myself wishing to focus at the HFD, but to do this I need to switch to LV and use ML feedback or scripts I've written to automatically move the lens to a specific HFD state, eg fully open, optimised for DoF or sweat aperture (open + 2 stops).

One thought I've had is explore if ML could allow me to drive the lens with the mirror down (sic), ie I'm using/accessing the normal eye piece functionality/feedback.

My idea is to use a script and allocate this to a key of my choice, thus allowing ML (if possible) to drive the lens to the HFD whilst I look through the eye piece, ie composition is the one thing I wish to control.

I've explored the current lv functionality in Lua, but this doesn't help me.

So the question is, could ML be used to allow the AF drive, whilst the camera's mirror remains down. Or is this idea simply 'silly'.

I appreciate this is not top priority at the moment, but I would like to know if it is possible in principle.

This may have been requested before, but as if just came up in a recent post I would like to (re)register a request.

The ML spotmeter it a very powerful feature, however, IMHO it has two 'limitation'.

First, it evaluates a fixed area of the sensor. An enhanced approach would allow two things. First, provide the user a variable to decide the spot size on the sensor. The second would be to make the sensor intelligent, and estimate the degree spot size according to the FL, eg 5 degree, 1 degree and, say, 0.5 degree.

Second, the ML spot doesn't work in Canon zoom mode, and maybe this is a fundamental ML-Canon limitation. If it is not, then making the spot work in Zoom mode would be a killer feature, especially with the additions above.

Finally, as a personal challenge I would like to look at the current coding of the spotmeter, but don't know where to find it in the ML code. Could someone point me at the right place.

Cheers

Garry

On the off chance someone has experienced this problem and can hint at a solution.

I have a 50D and have been using ML on it for a long time, I also have the latest nightly.

I am well aware of the CF-battery protocol ;-)

All of a sudden all, and I mean all my batteries, kept discharging really quickly in the camera. I have a couple of Canon batteries and four others. I have tried three different chargers and the internal battery seems OK, i.e. date is maintained without the main battery.

I just tried two new third party batteries, with good Amazon reviews and these two exhibit the same phenomenon, i.e. battery seems to charge OK, when first in the camera all is OK, then the charge crashes.

Has anyone any ideas? I'm at a loss.

I know there have been a few posts related to ND filters and VND filters.

In case you haven't seen this, another approach is a sensor mounted filter.

http://stcoptics.com/en/clip_filter/

Potentially a dumb question, but is it possible to have ML ficus peaking enabled when Canon zoom is also being used?

I thought I would share a post I just did following a recent trip to Iceland.

I do this to give those new to ML-based photography an idea about how radical it can look to some.

http://photography.grayheron.net/2017/03/first-reflections-from-iceland.html

Cheers

Garry

The ML focus peaking is a great feature for manual lenses, however, the area that is being highlighted is often very difficult to see.

Although I recognize this may not be a top priority, may I request the 'developers' consider adding a feature that amplifies the visual feedback, eg a stronger visual indication of the in focus area that stands out more that the current feedback.

I'm thinking about how to implement ICM (intentional camera movement) using, of course, ML Lua.

The idea is simple.

During an open shutter action, move the lens, i.e. coincident with the image being taken.

The ICM technique implies a longish exposure, eg using an ND filter.

Does anyone know if this parallel action is possible?

If it is not, could one of the gurus consider implementing 

Cheers

Garry