Author Topic: Bolt trigger module test results (bolt_rec.mo)  (Read 132592 times)

Yuppa

  • Freshman
  • **
  • Posts: 61
  • 60D + ML (Nightly)
Re: Bolt trigger module test results
« Reply #100 on: July 12, 2013, 10:36:17 PM »
Body: 60D
Build: Latest compiled 60D Raw + BOLT_RAW

If you've racked your brain as to why you can't lower the shutter below 1/24.7 even with "Exposure Override" set to ON:

From mk11174 (excellent primer BTW):

http://www.magiclantern.fm/forum/index.php?topic=6303.msg55463#msg55463

"3. Turn on FPS to 15fps Low Light which is 14.985fps."

So...

This is in the "FPS override" menu (select the Video icon). Set "Desired FPS" to 15 THEN make sure to set "Optimize for" to "Low light".  This will automatically reduce the shutter to a range between 1/15.3 and 1/39.8 so you can set ~360 degree shutter.

I'm restating this again as this is critical to (my) success!  Just playing with hot light, I'm getting much better results, i.e. complete frames.  Only problem: where did the lightning go?

Thanks again to g3gg0 and mk11174 and...
When you care more about capturing DATA, as opposed to WONDERMENT, you've lost your creative SOUL.

Yuppa

  • Freshman
  • **
  • Posts: 61
  • 60D + ML (Nightly)
Re: Bolt trigger module test results
« Reply #101 on: July 13, 2013, 11:10:01 PM »
Not the best view, but decent for my 1st attempt.

Body: 60D.
Global Draw: off.
Lens: old 18-55 kit lens.
State: hand held.
Threshold: 7500 or so.
Final: JPG, 15% compression.
Results: some pink and broken frames (12 of 64--taking 3 post frames for no reason really), but they were in ADDITION to the good frames, so it was a non-issue.  As far as I'm concerned, it worked like a charm!

When you care more about capturing DATA, as opposed to WONDERMENT, you've lost your creative SOUL.

mk11174

  • Contributor
  • Hero Member
  • *****
  • Posts: 809
Re: Bolt trigger module test results
« Reply #102 on: July 14, 2013, 12:04:41 AM »
Not the best view, but decent for my 1st attempt.

Body: 60D.
Global Draw: off.
Lens: old 18-55 kit lens.
State: hand held.
Threshold: 7500 or so.
Final: JPG, 15% compression.
Results: some pink and broken frames (12 of 64), but they were in ADDITION to the good frames, so it was a non-issue.  As far as I'm concerned, it worked like a charm!

Nice! Cool to see someone else shooting lightning! I just had a cell blow through here in West Central Florida, I think I got a good one, still waiting for storm to die out to check, here is one from yesterday!

500D/T1i  550D/T2i  600D/T3i  700D/T5i

mk11174

  • Contributor
  • Hero Member
  • *****
  • Posts: 809
Re: Bolt trigger module test results
« Reply #103 on: July 14, 2013, 12:06:30 AM »


Another from earlier today!
500D/T1i  550D/T2i  600D/T3i  700D/T5i

mk11174

  • Contributor
  • Hero Member
  • *****
  • Posts: 809
Re: Bolt trigger module test results
« Reply #104 on: July 14, 2013, 12:13:41 AM »


Here is the one I just got 30 minutes ago!
500D/T1i  550D/T2i  600D/T3i  700D/T5i

Kim.dh

  • New to the forum
  • *
  • Posts: 43
Re: Bolt trigger module test results
« Reply #105 on: July 14, 2013, 02:54:55 AM »
Awesome! Is it possible to automatically take a full resolution, shutter picture,after the raw -- so that you can overlay the bolt on the high resolution picture in post?

mk11174

  • Contributor
  • Hero Member
  • *****
  • Posts: 809
Re: Bolt trigger module test results
« Reply #106 on: July 14, 2013, 05:24:45 AM »
Awesome! Is it possible to automatically take a full resolution, shutter picture,after the raw -- so that you can overlay the bolt on the high resolution picture in post?
Not of the bolt, but yeah, should be able to make it take a picture after a detection, but this would slow things down and would not be good for multisequence, might cause a miss of the next strike in a sequence.

Are you hoping to keep the image the same size as the raw_rec capture and just use the CR2 to clean up everything else?

It might stop the raw_recording though, so that might be an issue to have it stop after every time it take a picture.
500D/T1i  550D/T2i  600D/T3i  700D/T5i

Kim.dh

  • New to the forum
  • *
  • Posts: 43
Re: Bolt trigger module test results
« Reply #107 on: July 15, 2013, 03:32:41 PM »
Not of the bolt, but yeah, should be able to make it take a picture after a detection, but this would slow things down and would not be good for multisequence, might cause a miss of the next strike in a sequence.

Are you hoping to keep the image the same size as the raw_rec capture and just use the CR2 to clean up everything else?

It might stop the raw_recording though, so that might be an issue to have it stop after every time it take a picture.

I was thinking of upscaling the bolt dng to cr2 size or the largest resolution the bolt itself looks good, then mask the bolt on the cr2.

Yuppa

  • Freshman
  • **
  • Posts: 61
  • 60D + ML (Nightly)
Re: Bolt trigger module test results
« Reply #108 on: July 15, 2013, 05:49:11 PM »
@Kim.dh

Why don't you just press the shutter button all the way RIGHT after the bolt gets recorded?  You have to do it, sure, and it'll be 16:9, which is somewhat of a bummer, but it'll also be @ whatever quality setting you've set.  Large fine yields 5184 x 2912 on 18 MP crop sensor bodies.

I would try THAT and figure out your post processing--masking, blending, etc.  If there's any wind at all, the clouds--or trees (if applicable)--will have moved regardless.

The issue I see is the photo mode is controlled mechanically, by the mode dial (60D) or switch.  What's the system going to do when you programmatically switch it to photo mode when the dial or switch is in video mode?  Has that ever been tried before?

CORRECTION: the 5184 x 2912 is JPG, but in RAW (CR2), it's 5202 x 3465! What is that, ~3:2?
When you care more about capturing DATA, as opposed to WONDERMENT, you've lost your creative SOUL.

Kim.dh

  • New to the forum
  • *
  • Posts: 43
Re: Bolt trigger module test results
« Reply #109 on: July 16, 2013, 01:14:19 AM »
I can do it manually of course, but I might get bored standing around waiting for a bolt, so an automatic trigger would be awesome. If it is hard to implement with full res cr2, a high res silent picture (from multiple crops) might be possible?

dprice970

  • Just arrived
  • *
  • Posts: 1
Re: Bolt trigger module test results
« Reply #110 on: July 17, 2013, 04:29:07 AM »
As a storm photographer all over the midwest I have used several lightning triggers and have ruined several lightning triggers ... I am excited to try this and by some of the result already it looks promising ... I say thanks the dev. of this and everything else I use on ML I am still lost on my 7d since it doesn't have nearly half the features as my others do now but that's ok I will live and LOVE this trigger feature if I can get it to work .. will keep ya updated if and when I get some day time lightning

mk11174

  • Contributor
  • Hero Member
  • *****
  • Posts: 809
Re: Bolt trigger module test results
« Reply #111 on: July 22, 2013, 03:21:17 PM »
I modified the trigger with a new auto exposure feature, its not a fast auto exposure feature for regular movie recording but it is plenty good enough for the bolt trigger and the settings best used with bolt trigger.

I added an option to turn the new feature on and off and a min and max level, best to only play with min level though, but from all my tests the settings I set are optimal.

This new feature will allow you to turn on bolt trigger press record and relax and let the trigger adjust for the sky changes while it captures day bolts.

I have it set all the needed settings for the cam to use trigger correctly, shutter, iso, and white balance.

Alls you need to do is before turning bolt trigger on, is to turn on FPS at 15fps at Low Light so it is 14.985 and also be in Movie Mode 1920x1080 30p.

This is a mixture of Raw_Rec and Bolt_Rec. This trigger would not be possible without g3gg0.

http://www.mediafire.com/download/nyx4tg13dc5na97/BOLT_RAW_July_22_2013.zip

Here is the Source: http://www.mediafire.com/download/a593ytr8wrpz8w2/BOLT_RAW_SRC_July_22_2013.zip
500D/T1i  550D/T2i  600D/T3i  700D/T5i

xRun

  • New to the forum
  • *
  • Posts: 11
Re: Bolt trigger module test results
« Reply #112 on: July 29, 2013, 07:40:41 PM »
Not sure if this was covered, but you'll get more than 1920x1080 even in bolt trigger mode. All the way up to 3.5K in 2.72 widescreen (3584x1320) by hitting the zoom button once for 5X before activating the recorder. With a wide lens it'll cover a wide area with more pixels, excellent for lightning capture. The framing is off though, width-wise what you see on screen covers about 40% in from the right edge of whats actually captured.
Even with 15fps (following your settings above) I got that to work with a 1000x 32GB Lexar, capturing beautiful 3.5K raw stills. Might be a bit much if there's a lot of ongoing activity though, haven't fully tested that.

mk11174

  • Contributor
  • Hero Member
  • *****
  • Posts: 809
Re: Bolt trigger module test results
« Reply #113 on: July 29, 2013, 09:53:21 PM »
Not sure if this was covered, but you'll get more than 1920x1080 even in bolt trigger mode. All the way up to 3.5K in 2.72 widescreen (3584x1320) by hitting the zoom button once for 5X before activating the recorder. With a wide lens it'll cover a wide area with more pixels, excellent for lightning capture. The framing is off though, width-wise what you see on screen covers about 40% in from the right edge of whats actually captured.
Even with 15fps (following your settings above) I got that to work with a 1000x 32GB Lexar, capturing beautiful 3.5K raw stills. Might be a bit much if there's a lot of ongoing activity though, haven't fully tested that.
Yep, did not feel the need to mention about crop mode because most people that use ML already know of crop mode. I use crop mode when I get low cloud base cells that I have no reason to shoot the upper area, so I go to crop mode to get wide views, my first upload here with double bolts was shot in crop mode. Bolt trigger works in both modes just fine.
500D/T1i  550D/T2i  600D/T3i  700D/T5i

mk11174

  • Contributor
  • Hero Member
  • *****
  • Posts: 809
Re: Bolt trigger module test results
« Reply #114 on: July 29, 2013, 10:06:21 PM »
By the way, here is the latest version of the module with a fix to the auto exposure feature I have added, I noticed it was a bit flakey when using more then 1 scan line for obvious reasons, so I separated the auto expose feature from the main scanline code and now you can use up to 6 scanlines  for the triggering and the auto expose will only use a center line. You may need to tweak the MIN and MAX levels for your cam, but for 550D 5400 MAX and 4300 MIN work best.

Module
http://www.mediafire.com/download/zy4pya312kty9tm/BOLT_RAW_July_28_2013_Tweak.zip

Source Files http://www.mediafire.com/download/hcd7j1gadhqu1ef/BOLT_RAW_SRC_July_28_2013.zip
500D/T1i  550D/T2i  600D/T3i  700D/T5i

xRun

  • New to the forum
  • *
  • Posts: 11
Re: Bolt trigger module test results
« Reply #115 on: July 29, 2013, 11:43:05 PM »
Great work! :)

Two questions:
1. Live view powersave, would it be possible to have that during Bolt Trigger mode? I know a lightning storm may last only an hour so battery probably wouldn't run down during that. Just a thought cause the triggering seemed to stop on my 5d3 when I tried it.
2. Would there be any benefit or usability issues in using shutter speeds lower than 15fps? Is it possible to get a (near) 360 degree shutter as low as 1fps? Or 180 degree at 0.5fps? I'm thinking about both lightning capture at night on a tripod as well as something else I'm into - time lapse. In my experience the display gets sluggish at those low speeds though.

mk11174

  • Contributor
  • Hero Member
  • *****
  • Posts: 809
Re: Bolt trigger module test results
« Reply #116 on: July 30, 2013, 12:12:06 AM »
Great work! :)

Two questions:
1. Live view powersave, would it be possible to have that during Bolt Trigger mode? I know a lightning storm may last only an hour so battery probably wouldn't run down during that. Just a thought cause the triggering seemed to stop on my 5d3 when I tried it.
2. Would there be any benefit or usability issues in using shutter speeds lower than 15fps? Is it possible to get a (near) 360 degree shutter as low as 1fps? Or 180 degree at 0.5fps? I'm thinking about both lightning capture at night on a tripod as well as something else I'm into - time lapse. In my experience the display gets sluggish at those low speeds though.
Yes power saver works fine with bolt trigger, I use the Dim feature it really helps, but dont have it turn off display just dim feature and global draw off is ok.

And Bolt trigger is not really useful at the slower shutter speeds the detection reacts to slow at that point, but I like to use 4fps at night to capture lightning frames when I want to do something differ then stills, but for Bolt Trigger 15 is as low as I would go for sure and only way to avoid the rolling shutter effect with lightning anyway. If it is not 359 or 360 the rolling shutter will cause major split frames to where the bolt will be split and only be captured on half the image and the other half lost, so a full 360 is a must! but 4fps allows you to capture crawlers good just dont bother using it with bolt trigger.

500D/T1i  550D/T2i  600D/T3i  700D/T5i

xRun

  • New to the forum
  • *
  • Posts: 11
Re: Bolt trigger module test results
« Reply #117 on: July 30, 2013, 03:07:45 AM »
Yes, that's what I'm seeing too. Works with the display dimmed, but not with it all the way off in my experience. Not an issue really, unless there's suddenly a 6 hours lightning storm and we wanna capture all of it. Not gonna happen even with the display off.  ;D

mk11174

  • Contributor
  • Hero Member
  • *****
  • Posts: 809
Re: Bolt trigger module test results
« Reply #118 on: July 30, 2013, 03:57:15 AM »
Yes, that's what I'm seeing too. Works with the display dimmed, but not with it all the way off in my experience. Not an issue really, unless there's suddenly a 6 hours lightning storm and we wanna capture all of it. Not gonna happen even with the display off.  ;D
yeah cause it works in live view so it has to be on but yeah would be nice to have it shut lcd off but have live view still actually going.
500D/T1i  550D/T2i  600D/T3i  700D/T5i

mk11174

  • Contributor
  • Hero Member
  • *****
  • Posts: 809
Re: Bolt trigger module test results
« Reply #119 on: July 30, 2013, 10:42:53 PM »
Here is one other update, my friend says the auto expose feature was flakey on his 5D2 so I changed it up a bit and it seems to work fine on his and my 550D.

This link includes the src files and module:
http://www.mediafire.com/download/8x6t0iav8167n44/BOLT_RAW_SRC_July_30_2013.zip
500D/T1i  550D/T2i  600D/T3i  700D/T5i

Kim.dh

  • New to the forum
  • *
  • Posts: 43
Re: Bolt trigger module test results
« Reply #120 on: August 01, 2013, 10:08:41 PM »
Mk, what settings do you use for night time captures on the 550d, rural low light pollution? I assume you turn AE off, use 15 fps with 1/15 shutter. But what ISO and aperture?

Thanks for the builds and answers.

mk11174

  • Contributor
  • Hero Member
  • *****
  • Posts: 809
Re: Bolt trigger module test results
« Reply #121 on: August 01, 2013, 11:01:39 PM »
Mk, what settings do you use for night time captures on the 550d, rural low light pollution? I assume you turn AE off, use 15 fps with 1/15 shutter. But what ISO and aperture?

Thanks for the builds and answers.
For night, I just shoot stills in M mode, dont use live view either just regular stills, 5 to 10 second exposures if really dark and not active I go longer or use bulb and manually open and close shutter. The aperture I set from F5.6 to F14 depending on how close or far away the action is. Oh and always use ISO 100.

If I do try video at night using the RAW REC module, I use FPS Feature at 4fps let shutter be controlled by FPS should be like 1/4 shutter, and aperture I open up all the way unless there big strong bolts, but bolt trigger is no good for night stuff, raw rec part of it makes for really cruddy noisey images only at night though. Bolt trigger is best used DAY only in my opinion from all the tests I have done judging by the quality and the rolling shutter issues.
500D/T1i  550D/T2i  600D/T3i  700D/T5i

mk11174

  • Contributor
  • Hero Member
  • *****
  • Posts: 809
Re: Bolt trigger module test results
« Reply #122 on: August 09, 2013, 05:45:15 AM »

Time lapse created using the bolt trigger and using a time lapse feature I added to it so it will capture a frame whatever interval you choose, for this storm I set to every 6 seconds and it also captured the bolts like it was made to do. I snagged 3 bolts during this storm, bolt trigger worked perfect!
500D/T1i  550D/T2i  600D/T3i  700D/T5i

Tichro

  • New to the forum
  • *
  • Posts: 13
Re: Bolt trigger module test results
« Reply #123 on: August 09, 2013, 04:01:42 PM »
Excellent work mk11174.

1 question, is this working with the new builds now?

mk11174

  • Contributor
  • Hero Member
  • *****
  • Posts: 809
Re: Bolt trigger module test results
« Reply #124 on: August 09, 2013, 06:13:22 PM »
Excellent work mk11174.

1 question, is this working with the new builds now?
Yes, it should be, I keep my build up to date and module still works fine.

Here is latest module with Time Lapse feature and Autoexpose feature.

http://www.mediafire.com/download/z61z6ucgodgq1qc/BOLT_RAW_Aug_9_2013_Update.zip
500D/T1i  550D/T2i  600D/T3i  700D/T5i