5D2 RAW video Builds 14-Bit

[ted ramasola]

thank's Ted, I will see further on MLV. I've read the topic of reddeer to extract RAW+wav, but I need a topic for synchro in after effect or premiere. So 1050x cards are full functionnal on 5D2 ?

To get the full potential on the 5D2 I recommend the 64gig 1000x, the 1050x will function on the 5d2 BUT due to its write speeds you get smaller resolutions that are continuous. But yes, it will function but you might not get longer record times.

For synchro, use resolve. It is faster and do not go the after effects route, its time consuming and the acr is not designed for long videos but photos. Its just ok for short clips but in the long run, resolve is designed from the ground up for raw videos.

To sync audio refer to pages 122-124 of the resolve manual.

To extract MLV videos with wav, use either g3gg0's MLV browser ro tonybeccar's mlv converter.
http://www.magiclantern.fm/forum/index.php?topic=8447.0

http://www.magiclantern.fm/forum/index.php?topic=10198.0

[reddeercity]

I should mention that right now the fastest stable MLV+Audio build comes from the nightly build
(magiclantern-v2.3.NEXT.2014Feb16.5D2212)
1872x936 2:1 23.976p + audio, Is continuous with 1000x card on 5d2 fill buffer set to "0"
I will be using this tomorrow on a News-story/interview 

Edit: I do my audio syncing in A.E., but I also record a master track on a Zoom 4Hn
I just use the audio in mlv to sync with master. Resolve is Ok, but ACR plugin is the best out there,
Its just a slower workflow but better result, at least that's what I think anyways.

[PeterR]

You can manually link the audio in resolve so you can export them as individual clips, if you don't link them, the clips will be exported without sound. When you export the resolve timeline as a SINGLE clip audio is included.

Look at page 122-124 of the Resolve manual PDF. There is a "one click" method to link all wav file to the clips however, I found out that longer clips will have a linked wav file that suddenly has no sound half way, and you can see in the resolve timeline the linked waveform volume is gone halfway. what I did was manually link only the problematic clips so I don't have to link all.

Thank you, Ted. I'll try this in the morning.

[xvince1]

I should mention that right now the fastest stable MLV+Audio build comes from the nightly build
(magiclantern-v2.3.NEXT.2014Feb16.5D2212)
1872x936 2:1 23.976p + audio, Is continuous with 1000x card on 5d2 fill buffer set to "0"
I will be using this tomorrow on a News-story/interview 

Edit: I do my audio syncing in A.E., but I also record a master track on a Zoom 4Hn
I just use the audio in mlv to sync with master. Resolve is Ok, but ACR plugin is the best out there,
Its just a slower workflow but better result, at least that's what I think anyways.

To get the full potential on the 5D2 I recommend the 64gig 1000x, the 1050x will function on the 5d2 BUT due to its write speeds you get smaller resolutions that are continuous. But yes, it will function but you might not get longer record times.

For synchro, use resolve. It is faster and do not go the after effects route, its time consuming and the acr is not designed for long videos but photos. Its just ok for short clips but in the long run, resolve is designed from the ground up for raw videos.

To sync audio refer to pages 122-124 of the resolve manual.

To extract MLV videos with wav, use either g3gg0's MLV browser ro tonybeccar's mlv converter.
http://www.magiclantern.fm/forum/index.php?topic=8447.0

http://www.magiclantern.fm/forum/index.php?topic=10198.0

Thank's for both of you 

[dariSSight]

I should mention that right now the fastest stable MLV+Audio build comes from the nightly build
(magiclantern-v2.3.NEXT.2014Feb16.5D2212)
1872x936 2:1 23.976p + audio, Is continuous with 1000x card on 5d2 fill buffer set to "0"
I will be using this tomorrow on a News-story/interview 

Edit: I do my audio syncing in A.E., but I also record a master track on a Zoom 4Hn
I just use the audio in mlv to sync with master. Resolve is Ok, but ACR plugin is the best out there,
Its just a slower workflow but better result, at least that's what I think anyways.

Thanks for the tips like always, now the magiclantern-v2.3.NEXT.2014Feb16.5D2212 file is the second Feb 15th download. Do you primarily use the 2:1 1872x936 setting.

[reddeercity]

Thanks for the tips like always, now the magiclantern-v2.3.NEXT.2014Feb16.5D2212 file is the second Feb 15th download. Do you primarily use the 2:1 1872x936 setting.

Yes 2:1,  I don't like any A.R. less then that as it can be re-framed to full HD
Or As is for Cinematic Super-Scope or also known as Univisium (http://en.wikipedia.org/wiki/Univisium)
So its a win win scenario, And for todays Interview I did with that build, everything when perfectly no problems.
It was to a point where I no longer worried about MLV/Raw Stop recording unexpectedly .
But I did not use any HDMI devices as this was a Run & Gun Style interview, so I don't think I would get Continuous
at 1872x936 with hdmi but I haven't tried that yet.

[melimelo]

Hi,

I understand Dual ISO RAW video is yet to be ported to 5D2, right?

Any alternative idea for raw video in situations where dual iso video would apply? will dual iso video ever be ported to 5D2 or is there any physical/technical limit?

Thanks

[PressureFM]

Hi,

I understand Dual ISO RAW video is yet to be ported to 5D2, right?

Any alternative idea for raw video in situations where dual iso video would apply? will dual iso video ever be ported to 5D2 or is there any physical/technical limit?

Thanks

Please read the Dual ISO white paper to understand why it only works on a few camera models.

[melimelo]

Please read the Dual ISO white paper to understand why it only works on a few camera models.

I'm not too good at very technical stuff, but from what I can see in this doc http://acoutts.com/a1ex/dual_iso.pdf, nothing is said about dual ISO for video vs photography, and Dual ISO does work in picture mode for the 5D2.

[reddeercity]

Duel ISO just photo for 5d2, the CPU can not keep up.
Try HDR video-Raw it will come close to duel iso but at a cost of half of source frame rate.

[a1ex]

The horizontal resolution is restricted to multiples of 8 bytes and 16 pixels according to latest findings. This restriction is valid at least for 5D3 and 6D (didn't do much testing on other cameras), but I'd like to keep the code portable without camera-specific exceptions.

So, before including this change in nightly builds, I'd like to ask you which is better: 1888 with 8 pixels of black border that you will have to crop, or 1856 without any border pixels? Between these 2 values, there are no valid resolutions that respect the alignment restrictions.

This change was discussed here:

http://www.magiclantern.fm/forum/index.php?topic=3904.msg106087#msg106087
https://bitbucket.org/hudson/magic-lantern/pull-request/438/raw-recording-force-line-size-to-be/diff

[romainmenke]

From a workflow point of view it would be better not having to crop in post.
Keeping things "simple" is important I think, to keep usability as high as possible.
Cropping exactly on the pixel can be tricky, and if you forget to crop and enlarge or sharpen the black border will effect the image.

If people worry about those lost pixels, they should buy a purpose build film camera with native Raw support.

[reddeercity]

My Vote for the max pixels 1888 ! I don't have problem cropping in post etc..

[ted ramasola]

I vote for a resolution that requires less post manual adjustment and can be scaled equally so that NLEs can automatically scale to fit into "standard" aspect ratios.

[gabz]

If people worry about those lost pixels, they should buy a purpose build film camera with native Raw support.

If people don't worry about those lost pixels, they should stick to their webcam

[a1ex]

Please show us the difference in image quality between 1856 resized to 1920 and 1880 resized to 1920.

Blind test preferred.

[guentergunter]

The question is:
Quality (24 (effective) pixels more, no crop)   VS   A few seconds less clicking in post

I can understand, that any more steps in post drive you nuts - at least when you're in a pressure.
But I'm producing my own shorts (or other's as a camera man) without any time pressure in post. So, from this point of view, it would highly demand 1888 as for the higher resolution and no further crop.

Anyway, both options do not have much impact. But, for me it's only one thing, that drives me nuts:
Reviewing this decision from 10 years in the future, I will think: "I should have kept the reolution!" Because that is what lasts!

[a1ex]

I repeat. Show me the difference. You need to convince me it's worth the extra coding effort.

[Audionut]

There will be no visual effect on 1856 vs 1880.

I'd go for 1856, as it's a simple resize to 1920 or 1280 or whatever.

[ted ramasola]

to pixel peepers, here's proof that there is virtually "NO" difference between 1856 and 1880 upscaled to 1920.

A downscaled image for posting:


A link to bigger image:
https://scontent-a-dfw.xx.fbcdn.net/hphotos-prn1/t31.0-8/1622320_452213114909478_1345754329_o.jpg

closer view of the vertical scales.

[Africashot]

Absolutely agree, have stopped using 1880 ages ago and never regretted it, no difference whatsoever.

[Kharak]

Please make it 1856!

Having to re-align aspect ratio, as little as it may be.. Is just one more slow step in the post. I'd rather avoid it.

And thank you, Ted for pointing it out

[reddeercity]

I see a sharpness difference! 1880 is sharper with cleaner edges Then 1856.
1856 is softer image and never be as sharp as 1880, If you look closely at Ted's
Posted image you can really see this.
I Can't believe we are Really talking about this!!
This should be a No Brainer! Max. Resolution, since it take no extra performance out
Of the camera there is no need to limit the the frame size.
I just finish a shoot a few days back with nightly from feb16th
and filmed at 1872x936 continous . MLV is to a point now
That a can trust it not to fail. So anyone that's says that
There is no difference , you are just fooling you self in to thing it's OK.
And when come to Green Srceen/VFX you Need the Sharpes image
You can obtain .   

[ted ramasola]

I would be dishonest to say there is no difference. That is why I carefully wrote in my post, "VIRTUALLY NO DIFFERENCE" .

And I can see it as well in the vertical scales between 600 and 700 lines. This is why I always rely on charts and just roll my eyes at those that say, "well, we don't shoot charts".

I have presented the exhibits/evidence in this "TRIAL" , I leave the judgement to the judge and jury. 

And for kicks lets add how all that fare against a 2048 downscaled to 1920.

[a1ex]

The reason for changing the resolution is portability: in order to keep 1872, one has to either add 5D2-specific exceptions in the code, or remove the safeguard in some other cameras (e.g. both 6D and 5D3 are known to crash on resolutions that are not multiple of 32 pixels, and this crash is a memory overflow that may result in anything, including permanent camera bricking).

I can choose between 1856 and 1888 with no coding effort. Between these 2 numbers there are no values that respect the alignment restrictions from EDMAC (it happens to work on 5D2, but I don't want to rely on this for all other cameras).

Camera-specific exceptions are not always desirable (they make the code harder to maintain and they have a negative effect on testing coverage). I've asked you to convince me that the extra effort to add the exception (and to maintain that exception from now on!) is worth it. I'm trying to simplify things, since this will reduce the burden of maintaining the code base (and this is one of these things that can be kept simple and portable).