Topics

Hi all !

Is there an answer or an update to the luma shifts that we are experiencing when tweaking dng or raw file "Highlight / Clarity" sliders in Adobe Camera Raw yet ?

I know this was a well recognised issue some time ago ... But i'm using latest version of Adobe Photoshop CC and it still seems to be a problem.

Here's an older thread ... but no recent answers : http://www.magiclantern.fm/forum/index.php?topic=6981.0

Any advice and updates would be really helpful as i love using ACR as my raw converter.

Hi !

I'm playing with Auto-ETTR for shooting timelapses of sun set / rise and then into night time with stars.

What is the suggested setting for "Highlight Ignore" ?

I normally would asume that 0.1% is best as i want to hold as much highlight detail as i can for sunset ... However, i read somewhere that as much as 5% is needed ... But surely 5% burns out the sun and makes my sunsetting a white blob ? What about at moonrise - should i then use 0.1% to hold the small highlight detail a full moon shows etc ?

Can anyone explain how the "Highlight Ignore" function in practical settings a little more for me please ? Should i also tweak the Midtone SNR & Shadow SNR as not quite sure how this works either as usually leave these at default.

Finally ... are there any instances when i'm best not to use A-ETTR and better to use Aperture Priority AV metering instead ?

Fabulous ! Thanks in advance ...

Hi all ...

I'm exploring the incredibly clever Auto ETTR feature for day to night timelapse.

I understand the concept and have had some success, but my timelapses always start with fast shutter speeds of around 1/500 sec. This can drop to say 30 secs when the light levels drop too, but I'm left with the ugly staccato effect during the day.

Is there any way around this - via user settings ? Or alternatively, can i use a vari nd filter cranked up to max to start with (reducing exposure approx six stops) and then turn the filter as light levels drop to reduce the ND effect.

If i can play with a vari nd filter - how will this affect flicker and the consistency of Auto ETTR module ?

Your advice would be greatly assisted to help me improve my day to night timelapses.

Thanks in advance

Hi !

I'm new to ML but have recently experienced problems on my 5D2 using magiclantern-Nightly.2015Jun09.5D2212.zip

The camera froze and said card error i needed to use new card. Was just using h.264 not raw.

Where do i submit the Logs as i don't see how i can attach to this post ?

Hope this helps you guys make an even better tool for my ageing 5D2.

Thanks ...

I'm thinking of buying a new Canon 5D III and installing Magic Lantern for raw video etc. However, the latest Canon firmware is 1.3.3 and ML only works up to 1.2.3.

I also see that on the Canon website under download new firmware it states - "Once the EOS 5D Mark III camera is updated to Version 1.3.3, it cannot be restored to a previous firmware version (Version 1.0.7 through 1.2.3)."

Does this mean that if i buy a new Canon 5D3 i'll never be able to use Magic Lantern as there is no new firmware support ? If so, have Canon introduced this to stop ML from stealing sales of their C series video cameras etc ?

I know i'm not supposed to ask ... But what likelihood is there of ML support for 1.3.3 and future Canon 5D series - is it possible or is the firmware now 'locked' ?

Want to buy new 5D3 but only if it's got the amazing ML raw video support.

Hi !

I've started playing with the .mlv raw video on my Canon 5D II which creates max size 16:9 1856 x 1044 pixels.

I'm generating ProRes 4444 .mov files from MLRawViewer but ultimately i need to upsize slightly to full 1920 x 1080 HD.

What's the best way to do this and then sharpen the final output using FCP X ?

Am i best to set the 'New Project" to Custom 1080p / 1920 x 1080 pixels / 25fps and then just apply some Sharpen filter (2.5 default) on the clip and Export to Quicktime 1920 x 1080 or am i best to resize using the transform or crop tool then apply some sharpening then Export ?

Not sure of the 'right' way to do this so any advice would be appreciated. Thanks ...

Hi everyone ...

Just wanted to make users aware of a great little app that seems to help my own ML Raw workflow.

It's called Cliphouse by Pomfort and is $49 / year or $129 for a permanent licence. Free 14 day trial too.

http://pomfort.com/cliphouse/

It allows for copying and back up of files, minor 'in app' editing / grading / adding flat profile and most importantly Export of cdngs and MLRawViewer dngs to various video formats such as ProRes 4444, h.264 etc. It's still necessary to generate the cdngs in the first place with MLRawViewer (or similar), but avoids the use of Resolve or After Effects.

I'm playing with it at the moment with the trial.  Not really heard it mentioned on the ML forum and it seems pretty cool. Saves me using Quicktime Pro 7 for Export and is pretty fast too. I think it can batch process clips too.

It does seem to occasionally crash, but if more of us used it and bought it, i imagine the stability and functionality would be improved more with updates.

Anyone got any recent experience with this - Jan 2015 onwards ?

Hi everyone !

Sorry to ask but I'm a photographer and a video newbie and i'm not sure how to use the fabulous MLRawViewer with my Canon 5D2 and ML Raw video.

I've downloaded the latest version 1.4.3 for Mac and read all the Read Me instructions.

I can open my .mlv files in MLRawViewer but they never seem to look like they should. When i cycle through the standard LOGs, some look better than others but many simply look dark, crushed blacks, high contrast etc. I know I can increase the exposure using the Up and Down cursor keys, but this seems to overexpose the preview image. Is this right as i'm dealing with a log file - not a traditional raw photo ?

Ideally i'd like a nice flat profile to work with, so i downloaded VisionLOG for ACR and the VisionLOG to Camera Neutral Rec709.cube files from VisionColor. I've also imported the VisionLOG to Camera Neutral Rec709.cube 3D LUT to MLRawViewer using the Shift L keys but the histogram looks underexposed. Should i now increase the exposure with the Up and Down keys to make the histogram look good or use the settings at default ?

Which LOG is best to output from ?

I'd like to export to Pro Res and grade in Final Cut Pro as I don't have Adobe Premiere or AfterEffects. I'm guesing this means first outputting 16Bit tiff files for each frame using Adobe Camera Raw and the VisionLog Camera Calibration / Profile ? Then making an Image Sequence with Quicktime Pro 7.

Also ... what camera profile should i load as a picture style for my 5D2 - ideally one that will show the full contrast range of the scene i'm shooting.

I've spent days reading through all the processing forum posts etc, but still trying to work out the best way to do all this. Any help would be greatly appreciated. Thanks ...

I'm trying to create a fast and efficient workflow for my ML raw video.

Atm, I'm using the fabulous MLRawViewer 1.4.3 to convert the raw footage to .dng. I then export my dngs to 16 bit tiffs with VisionLog dcp profile added in Adobe Camera Raw. Finally, i create a ProRes HQ movie using Image Sequence via QTP7 - but it looks slightly lighter due to gamma shift.

Is there any way to stop this gamma shift ? I'm guessing the slight QTP 7 gamma shift is less of a problem since I'm using the VisionLog flat profile ?

I edit footage in FCP X as i don't have After Effects or Premiere and don't want to use Resolve as i'm a photographer and still getting my head around FCPX !

Any advice appreciated.

Hi !

I'm getting to grips with MLV raw and using MLRawViewer 1.4.3 to process my files.

However, I see there are several differemt log settings to choose. Which one is best to edit in FCPX ?

Tbh, most of them look pretty dark or very flat but i assume this still allows a great deal of grading in FCPX as (i think) the files are ProRes 4444 ?

I'd like to use VisionLog cube LUT if there is one as i use the VisionLog ACR dcp profile when playing with dngs. Is this a good idea ?

Is MLRawViewer 1.4.3 also the latest version as not much seems to be happening lately.

I'm a bit confused about all this log and lut stuff so all advice is appreciated.

Hi !

I'm not that familiar with Lightroom as i only own version 3.6. However, i know Lightroom 3.6 could be used to output slideshows and i think it could output the movies as Pro Res video. It used to require a template for 24 , 25 or 30 fps.

This ability seemed to stop with Lightroom 4 and 5, but does anyone know if this is possible in the latest version of Lightroom 6 ?

I know that it can be done using LR Timelapse 4 Pro but i need to buy that and it would be great if it could all be done in Lightroom.

All information is of interest.

Hi all !

I'm currently using a Canon 5D II with a VAF-5D2 anti aliasing filter for Magic Lantern raw video. Aliasing is still apparent, especially when converting to .dng and tif output.

I was wondering if the anti-aliasing and moire were any better with the Canon 5D III ? They are a pretty good price now so was not sure whether to upgrade or not.

What are people's experience with aliasing and moire on Canon 5D3 using ML raw video ? Is it likely to be any better than my Canon 5D2 with the VAF fitted ?

All thoughts appreciated ...

I'm still tring to find the best way to process .mlv raw video files to edit in Final Cut Pro X ...

Like many of us, after using MLRawviewer (which is awesome!),  I love the ability to edit the exported .dng files in Adobe Lightroom ... so it's a shame to have to swap from one app to another.

I know that in the past (until version LR 5 i think), Adobe used to support 'Slideshow Templates' and we could output timelapse sequences (or video files) with 24, 25 or 30fps using LR Timelapse templates.

Sadly, the video files were mpeg4, so quality suffered. Also LR Timelapse & Lightroom 5 no longer support the older slideshow templates, but in concept it seems a good idea.

What is the likelihood of Adobe returning support for these 'Slideshow Templates' ... idally with Pro Res output and option to attach the .mlv WAV sound files too ? Do Adobe listen to feature requests ? If so ... maybe we should all suggest it !?

It's possible to currently output high quality Pro Res files using .dng's within the fabulous "LR Timelapse" http://lrtimelapse.com/, but in an ideal world we'd just use Lightroom or Adobe Camera Raw for the entire conversion process.

What are your thoughts ?

I do have a screen grab of Lightroom 3.6 that shows 'Slideshow Template' support but couldn't work out how to upload image from desktop ! Please help !

Hi !

I'm loving ML raw video and appeciate all the development being put in ... so thanks.

I'm currently using an old Canon 5D2 but am thinking about upgrading to a Canon 5D3. Can you answer these questions :

1 ) I understand the 5D3 records full HD with ML - but is this continuous on a fast card ?

2 ) Playback on the 5D2 is obviously a bit jittery and has skipped frames. Is playback smooth on the 5D3 due to faster processor etc ?

3 ) Does 5D3 playback also have sound or is this a possible development for future (sorry to ask !  ) ?

4 ) I have heard rumours that there is a newer 5D3 firmware ? i'm guessing this doesn't work with the ML Nightly Builds atm, so can i install older firmware on a new 5D3 if it ships with updated firmware ?

5 ) Is the raw video on 5D3 generally that much better than my 5D2 is currently using ?

All advice and input would be much appreciated !

Thanks in advance ...

Hi ! I'm a relative newbie so please excuse my lack of knowledge ...

I'd like to try recording Raw video with my Canon 5D2.

I've read most of the forum posts, but as raw video develops it's hard to keep on top of things.

What's the latest on recording video with 5D2 ? Which nightly build is best to use ? Can i record audio at same time using MLV or is that the realms of the 5D3 ?

I'd like to use Final Cut Pro X for post processing ... So i guess i convert the raw video files using Raw2Dng then use Quicktime Pro to create Pro Res clips ? If not, what alternative workflow is recommended with 5D2 ?

Thanks in advance ...

Hi !

I'm fairly new to Magic Lantern, so excuse me if this can already be done ...

I'm using ML on my Canon 5D2 to generate 7 frame exposure brackets for HDR spherical panoramas.

I have selected Mirror Lock Up and it allows a two second delay between frames, but it would be great to have three secs or longer delay to eliminate all camera shake when mirror goes up.

My spherical panorama head causes slight vibration and a longer delay option would really help ...

Is this possible ?

Also, is it possible to shoot the exposure bracket range in the same sequence as standard Canon firmware (on my 1Ds3) - Ie : Exposure at 0 and -,-,-, then +,+,+ ?

Thanks so much ! 

Hi !

What are the advantages of increasing the bitrate in ML h.264 to 2X or 3X default bitrate ?

Is it just more colour info in video files which helps with grading ? Or are there any other advantages of selecting faster bitrate for h.264 video on 5D2 ?

I use a Komputer Bay 64Gb 1000X card ... i'm guessing this is ok for 3X bitrate recording ?

Sorry to ask such a basic question but would appreciate your advice ... 

Hi !

Just wondered if it would be quicker (and possible), to output a lower res / quality video file when processing video in raw2dng ?

I primarily use the ProResHQ output just to see the video content and not for any other reason - and it seems to take ages to compress the high quality video file.

Be nice to have the option to chose - although I know i'm asking a LOT !

You guys are fantastic to have achieved what you have and just thought i'd give some feedback to try and help streamline the functionality.

Keep up the great work !   

Hi !

What is the best video resolution when recording raw video on Canon 5D2 ?

I see options for 1880 or 1872 for full HD - is one better than the other ?

Also - what is the best way to playback the raw video in camera ? The default playback function is slow due to slow frame rate - is this the best way to view what i've just recorded on camera or am i best to dial in different settings ?

Advice would be greatly appreciated ... 

Hi !

I'm just starting to play with raw video in my Canon 5D2 but need a little advice on post processing.

After converting the .dng to jpgs (or tifs), I'm importing "Image Sequence" into Apple Quicktime 7 and then saving .mov file. Problem is, the raw video looks great in the on screen QT viewer but loses quality when output / saved and compressed.

I'm editing in Apple Final Cut Pro X - so wonder if there is an easier / better way to process the raw video ?

I don't have Adobe After Effects or Premiere and would rather use the software I have if possible.

Any advice ?

Much appreciated  ...

Hi !

After a bit of general advice really ...

I've just started using ML raw process and it's amazing.

I'm in UK and normally shoot 25fps for PAL. However, with full 1880 x 1058 pixel HD and a fast KomputerBay 64Gb 1000x CF card, I'm hitting limitations on recording times.

The ML raw recording duration is obviously better with 24fps at this resolution.

This makes me tempted to shoot 24fps as opposed to my normal 25fps.

Most of my video is for the web, but if it were to be selected for tv broadcast, would 24 fps be much of a problem in UK or elsewhere ?

Just interested in advice please ... as every second counts at this high resolution raw ! 

Hi !

I've been playing around with high speed fill flash sync outdoors in the sun to allow large apertures and shallow DOF. Perfect for fill flash portraits in full sun.

However, whilst my Canon 580EXII flashgun allows me to sync up to 1/8000 sec and at wide apertures like f2.8, the portable flash power limits the distance i can shoot to about 1 or 2 meters with 85mm lens on 5D2.

I realise ML can't increase the power of my flash gun (!), but if the shutter exposure was undertaken with electronics rather than focal plane would this effectively allow for increase in distance from hand held flash gun ? I'm guessing it could only be performed with mirror up or in live view mode, but could this work ?

I'm not 100% sure how the high speed sync feature works on my Canon flash gun, but i understand it's limited by the mechanical movement of the focal plane shutter.

Is this right ?

If ML can use electronic shutter (rather than mechanical focal plane shutter), could it effectively make our lenses leaf shutter lenses ? These are previously limited to the expensive world of medium format.

Perhaps somebody smarter than me can give me an answer or implement this as new feature in the awesome Magic Lantern firmware ?