Messages

Slightly late to the party, but whatever....

I saw the original launch of the fundraiser and was intrigued but though it'll just be an interesting technology project without a decent OS.

Saw on NFS today about ML coming on board.....



This is now very cool, and I'm sure both Apertus and the ML gang will step up and produce a wonderful cambination!! (sod it, not fixing this typo - it works)

Batch 3 camera ordered and 30 euros in the pot for ML to get a camera!!

I've just posted what we found out about the 30 min limit in the main Dev area as there's not been any replies for a few days here - http://www.magiclantern.fm/forum/index.php?topic=12594.0

Hi all,

I'm cross posting from the 50D thread to aid visibility on something we've discovered and there's been no postings there since Saturday.

kolen came up with an interesting idea which seems to work, when shooting flick into the ML menu and then back to Live View and for the 50D recording seems possible over the 30 minute limit now.

It would be great if this could be done without touching the camera - so an internal timer blips the ML menu after 25 mins of recording.

For Raw FPS adjusted timelapse this would be absolutly fantastic!!

Possible?  Safe?  Thoughts on this?

I've just done a quick test, normal timelapse but at 25 mins flicked the menu button.

F*ck me!!

My battery wasn't fully charged but the 50D stopped at about 38.46 mins when it ran out of juce - so it's definitely recorded longer than normal!!

Fantastic idea, i wonder if the Devs can just blip the menu display just before 30 mins to make it continuous without touching the camera?

OK, thanks! So I think I should order a 1050x 128GB Komputerbay CF card today...

What about my other question - does the 50D has a RAW video (or MLV) playback module?

Yep, look in the Raw Video sub menu and you should see an option for playback if there's something on the card - there is for Raw, not sure about MLV as I don't use it.

Thanks PhilK! What is the biggest CF card size that I can use for installing ML ?
So before I spend 10 bucks on a cheap card just for installing, I would get another smaller Komputerbay card.
32GB? 64GB?

I also heard the 1050x 64GB Komputerbay card is faster then the 1050x 128GB Komputerbay card, is that true? Because it says the same write speed...

And does the 50D now have a playback module for playing back raw videos?

Thanks for all the help!

I *think* the biggest is 64GB, but I would check this as it's been a while since I had to do this.

If the 1050x Komputerbay cards are the same as the 1000x ones then the 64GB is faster, but again might be best to check the other forums to confirm this as I've not used the newer card.

I use my 2GB card for trying out latest ML versions so it does get some use though, as well as some pictures sometimes - so it's not a total waste.

JUST BOUGHT A EOS 50D ! 

So can now somebody tell me what CF card should I get for raw video?
Can the 50D use the fast speed of 1050x cards? And I read about problems installing it on 128GB cards, others say they use 128GB...so what is the answer?

Thanks!

Get a cheap 2 gig card and install it from that, then use your big card :-)

However, at any decent resolution, recording raw will fill up any nowadays most biggest cf cards far before reaching the 30mins limit...

Unless you are doing Raw Timelapse, then this 30 min restriction is a pain.  This is the reason I bought a BMPCC to go with the 50D, longest I've done with the BMPCC is just short of 5 hours with an external battery before the 64GB card was full (but the 50D with ML would be a lot better in a few situations like low light without this limit).

Of course, this 30 min restriction is Canon's, or the EU's, fault (different V.A.T. tariffs for video cameras than stills that can do under 30 mins video) - not Magic Lantern's

Mixture of 50D and BMPCC, wide was the 50D.

In ptp.c there's a few PTP functions set up already, one for starting recording (0x9997) and one to return 0xdeadbeef (0x9999).

With gphoto I'm getting nothing returned back on these - I know the code is working as I can get & set 0xd406 or named params like shutterspeed fine from both the gphoto shell and via C code.

I can't put any debug code in ML to see whats happening as I can't compile from source at the moment (lines 520/522 in shoot.c are crapping out) so at this point I'm unable to even get the existing ML PTP handlers to work!

Just getting 0x9997 working should be enough to get me started.

Bumping thread.

There's seems to be a lot of activity going on with remote controllers at the moment (for example:  http://petapixel.com/2014/01/16/diy-take-wireless-control-canon-dslr-using-30-router-android/) and a couple of threads here open on this topic.

From what I've been able to find out none of the ML parameters expose themselves to PTP, it's just the raw Canon ones.

So, questions.

1.)  Does ML expose anything to PTP?
2a.)  If not, why not - what's needed?
2b.)  If it does, what am I doing wrong!

[/main/capturesettings/aperture/main/capturesettings/shutterspeed]

Hi Paulo,

I'm trying something similar to you with a Pi, but I'll probably be using another language (C in my case).

But, coming back to your question what you've posted looks slightly wrong - here's what I have from my 600D.

gphoto2: {...ern/libptp2-1.1.10/src} /> list-config
/main/actions/uilock
/main/actions/syncdatetime
/main/actions/autofocusdrive
/main/actions/manualfocusdrive
/main/actions/eoszoom
/main/actions/eoszoomposition
/main/actions/eosviewfinder
/main/actions/eosremoterelease
/main/settings/datetime
/main/settings/reviewtime
/main/settings/output
/main/settings/movierecord
/main/settings/evfmode
/main/settings/ownername
/main/settings/artist
/main/settings/copyright
/main/settings/customfuncex
/main/settings/capturetarget
/main/settings/capture
/main/status/model
/main/status/ptpversion
/main/status/batterylevel
/main/status/lensname
/main/status/serialnumber
/main/status/shuttercounter
/main/status/availableshots
/main/imgsettings/imageformat
/main/imgsettings/imageformatsd
/main/imgsettings/iso
/main/imgsettings/whitebalance
/main/imgsettings/whitebalanceadjusta
/main/imgsettings/whitebalanceadjustb
/main/imgsettings/whitebalancexa
/main/imgsettings/whitebalancexb
/main/imgsettings/colorspace
/main/capturesettings/focusmode
/main/capturesettings/autoexposuremode
/main/capturesettings/drivemode
/main/capturesettings/picturestyle
/main/capturesettings/aperture
/main/capturesettings/shutterspeed
/main/capturesettings/meteringmode
/main/capturesettings/bracketmode
/main/capturesettings/aeb
/main/other/d402
/main/other/d407
/main/other/d406
/main/other/d303
/main/other/5001

I couldn't see the two lines bolded above in your list-config output.  Assuming this is just a copy/paste issue here's the next steps.

Check what values you can set like so:

gphoto2: {...ern/libptp2-1.1.10/src} /> get-config /main/capturesettings/aperture
Label: Aperture
Type: RADIO
Current: 4.5
Choice: 0 4.5
Choice: 1 5
Choice: 2 5.6
Choice: 3 6.3
Choice: 4 7.1
Choice: 5 8
Choice: 6 9
Choice: 7 10
Choice: 8 11
Choice: 9 13
Choice: 10 14
Choice: 11 16
Choice: 12 18
Choice: 13 20
Choice: 14 22
Choice: 15 25
Choice: 16 29

So, here's how I set the aperture fully closed:

gphoto2: {...ern/libptp2-1.1.10/src} /> set-config-index /main/capturesettings/aperture=16

Or fully open:

gphoto2: {...ern/libptp2-1.1.10/src} /> set-config-index /main/capturesettings/aperture=0

You'll need to play around with the config params as some are only read-only.

Interesting ideas!  I'm looking at something similar - here's what I have so far.

Wireless FF - http://www.kickstarter.com/projects/1608311032/lenzhound-wireless-lens-motor-control-system.  This is a complete system so there's no fun to be had in building it yourself, but should be a solid system for use.

Going down a more DIY path.

9" HDMI screen - http://www.kickstarter.com/projects/697708033/hdmipi-affordable-9-high-def-screen-for-the-raspbe.  This is designed for the RasberryPi but works as well as a field monitor.  The latest discussion is there will be space in the back of the case for a Pi, and the ability to switch between inputs so you can toggle between LiveView and the Pi.

Input device - http://retail.contourdesign.com/?/products/22.  This would provide enough input to be useful in the field without using a keyboard and mouse being plugged into the Pi, select what you want to change with the buttons (aperture, focus or whatever) and then dial in the value you want.

The bit I'm missing at the moment is what is exposed in ML via the USB PTP interface - which you should be able to access using ptpcam or libgphoto2.  I've asked in the general dev section but I've not got an answer yet (http://www.magiclantern.fm/forum/index.php?topic=9723.0).

While my solution isn't wireless it could offer more features.  Feel free to bump my question up at the link above if this is something interesting to you as well!

[really]

Test, as in really test. As if you were QC working for a software company. Don't just use ML in normal situations and submit problems when things go wrong (and don't just do another ML RAW vs. H264 test, there are more than plenty of those out there). Try to break a feature. Think of as many possible scenarios as you can. Throw everything you can think of at a feature to break it. Try every value of every setting. Try it in extremely unusual scenes or lighting. Write down your results like a scientist doing an expirement. Then share your results, even if nothing went wrong. It's also helpful for devs to know when something actually works. Test for the sake of testing, with specific intention, not for the sake of making your 'budget short film'.

What could be better for the Devs and easier for people to contribute to is producing test scripts.  A simple and clearly documented set of steps with the results easily quantifiable.

Something like this (a very basic example).

• Enable the raw.rec module [pass|fail]
• Enter the Movie menu and then enable Raw Video (when in Live View mode - pick resolution suitable for you card) [pass|fail]
• Render video and inspect it to check for bad frames [pass|fail]

While you sometimes can't go wrong with free-form testing, a more structured approach means you can measure the test coverage and split this activity up so many people can contribute and still have some control.

As new features are added, or defects found, new tests should be created - essentially moving to a Test Driven Development model.

As the website is PHP based there will be off the shelf applications that could be deployed to the site to automate this easily, or just use Goggle Docs and track it in a spreadsheet.

I've just had a quick play with libgphoto2 to see what params I can access via USB, I looked at libptp a while ago and libgphoto2 seems to be much better. 

The stock Canon ones seem fine but nothing from ML seems to be coming back when I query the cameras (50D/600D).

Any ideas on this, is it possible to read/write ML params via USB? 

The 50D now has a VAF available, can the first entry be edited to reflect this?

http://mosaicengineering.com/products/vaf/50d.html

Great news about the VAF-50D Levi. 

Do you know if there are any differenced between the prototype unit you've been testing and the production units?

OK, got a strange problem, or I'm doing something badly wrong - I suspect the latter 

I compile ML from source and this works fine for the 50D but for the 600D I get SOS flashing on the activity light!

Compile process is make clean && make 50D or 600D in the root and then cd into platform for the camera and make clean && make zip, then copy over the contents to the card.

The 600D works fine when using the nightly builds using the same card for the compiled one (completely removing each ML first), any pointers?

Looks interesting, it's designed for the RasberryPi micro Linux system but can be used as a fields monitor, resolution is 1280 x 800.

http://www.kickstarter.com/projects/697708033/hdmipi-affordable-9-high-def-screen-for-the-raspbe

Most of the pledges have gone, but there's still a few left.

Is there a link to the test shots? I'd like to see them . . .

Levi S. Davis should pop his head back into this thread at some point, he's been testing the prototype.

Here's his video for the prototype: http://vimeo.com/74470756

For me this is worth the money, especially at 200 bucks!

Any news for the Mosaic Engineering's VAF for 50D ? The 70D version is already available on their website ...

Bump as I'm interested as well.

Thanks A1ex, FIO_CreateFileOrAppend worked well enough for now, but if I get time I'll look at the trace module.

Quick question, is there a 'correct' ML way of dumping info to a logfile or card?  Should I use FIO Create/WriteFile?

I'm looking for the best way to printf the values of a few params for investigating.

Thanks

Fixed that compile error.

Yep confirmed, thanks.  But it took ages for the pull to register the updated file - I first tried when I got in about 3 hours ago and only in the last 30 mins did it register.

Looking at the Bitbucket site you committed the fix yesterday!

Is anyone else seeing this error building from source?

../../src/gui-common.c:29:12: warning: 'bottom_bar_hack' defined but not used [-Wunused-variable]
[ CC       ]   chdk-gui_draw.o
[ CC       ]   movtweaks.o
[ CC       ]   menuhelp.o
[ CC       ]   menuindex.o
[ CC       ]   af_patterns.o
[ CC       ]   focus.o
../../src/focus.c: In function 'focus_stack':
../../src/focus.c:260:29: error: 'focus_rack_enable_delay' undeclared (first use in this function)
../../src/focus.c:260:29: note: each undeclared identifier is reported only once for each function it appears in
make: *** [focus.o] Error 1

It's the same compiling for 600D, everything was OK yesterday and I updated just before I saw this.