hello there,
I'm french and I would like to use my 5Dmk2 in RAW mode to shoot.
But I find a lot of differents way to install the RAW module.
what difference between raw_rec and raw_mlv?
which module the most recent and stable to install?
Need more modules or parameters in particular to use the RAW module? ( I heard mlv_snd, play ... what else? )
I imagine that the subject has already been treated, but I would have real good answers to not install anything.
I have already installed the latest version of ML for 5Dmk2.
v2.3.NEXT.2014May14.5D2212
Really thanks for your help.
Ludo.
I would suggest installing the latest nightly. Although nightlies are very stable, the general consensus is that the official public release and alpha versions are considered completely stable.
As for recording raw video, there are two main modules, raw_rec.mo (RAW) and mlv_rec.mo (MLV) and two supporting modules for MLV raw (mlv_snd.mo and mlv_play.mo), one is for recording sound, the other for in-camera playback.
The main overt difference between the two is the raw module (raw v1) does not record audio whilst the mlv module (raw v2) does support audio recording via mlv_snd; there may be some other internal differences which are unknown to me. Also, some say the RAW module is more reliable than the MLV module. As far as post production, many apps exist for making it very easy for processing either version of ML raw.
Ludo,
Are you preparing for an important day shoot? If so sure you try end-to-end post workflow with whatever nightly build that you're using beforehand. It will give you a good idea of what you're getting yourself into. There are things that can go wrong during the shoot that will ruin your day if you are not prepared. Just making sure... :)
mlv_rec is the newer format that supports audio but if at all possible don't use audio on-camera - the recording speed to memory cards is the bottleneck and it'll mean a lower possible resolution. If you're on a mac, the free version of RawMagic (ML post processing) won't support mlv_rec, only raw_rec, but the paid version is only $30 IIRC.
Install all modules, you will have to activate them later in modules tab.
For video recording, i only use Mlv rec, mlv snd, mlv play and file manager.
Mlv_play also works with Raw rec.
Mk ii struggles when sound recording is activated, as the write speed is limited. Disabling audio recording will gain you more speed and that gives you higher resolution on continues or longer record times on non-continues resolutions.
Reddeercity, he's an avid ML Mk ii user and has managed to get continues 16:9 at max resolution, 1880x1058. I dont remember what his precise settings were though, but that was on raw_rec module, october 24th build.
Global Draw off and every other extra feature, think also Sraw was selected.
But mk ii records normally 16:9 1880:1058 for atleast 30-40 seconds without hazzling of disabling too many features.
This is ofcourse if you have a 1000x CF card.
ML Raw will seem quite confusing at start, but if you do some reading on the forum, especially the 5D mk ii thread in Camera Specific Section, you will learn a lot.
Good luck and prepare to be amazed.
thank you to you for the clarification.
I prepared a shoot, but I wanted to test before the RAW configuration.
I just installed the latest nightly builds (2014Jul25.5D2212)
I just tried to shoot in Raw_rec and Raw_mlv with Lexar 128GB x1066
And I could shoot opnly 4 seconds ...?!?
Anyone know why ? ..
I'm on a MAC, but raw2dng.app is not great?
thank you to you.
What resolution and frame rate?
1856x1044 at 25p
ML notifies me 80.8 MB/s but when i shoot there wrote 25..MB/s
Dont understand this result :o
normal ??
Benchmark complete.
ML Hightlgy.2014Jul25.5D2212, aa3fb0f81249 (unified) tip
Mode: movie 1920x1080 25p, Global Draw: OFF
Write speed (Buffer:16884k): 17.4 MB/s
Read speed (Buffer:16884k): 9.8 MB/s
Write speed (Buffer:16884k): 17.4 MB/s
Read speed (Buffer:16884k): 9.2 MB/s
Write speed (Buffer:16626k): 17.6 MB/s
Read speed (Buffer:16626k): 9.8 MB/s
Write speed (Buffer:4896k): 16.8 MB/s
Read speed (Buffer:4896k): 9.2 MB/s
Write speed (Buffer:8986k): 16.4 MB/s
Read speed (Buffer:8986k): 9.1 MB/s
Write speed (Buffer:2848k): 16.2 MB/s
Read speed (Buffer:2848k): 9.2 MB/s
Write speed (Buffer:1968k): 16.8 MB/s
Read speed (Buffer:1968k): 8.9 MB/s
Write speed (Buffer:128k): 8.5 MB/s
Read speed (Buffer:128k): 6.1 MB/s
Well, the numbers kinda match between test and what ML is saying, but they both seem WAY too low.
yes i think it's too low, but it's not really strange the result for this card (x1066) ?
Also, max resolution for continuous recording on 5DM2 is 1880x940@24p, so you're a bit over the limit, back down the resolution for continuous raw recording.
Do you have a decent USB 3.0 cardreader (recommend Kingston FCS-HS3) connected via USB 3.0?
If you're using a PC you might want to run a benchmark programm with CrystalDiskMark or ATTO.
It seems you purchased a faked card.
Good point, where did you get card?
25 fps will make max continues resolution smaller. I was talking about 23.976 earlier. Hihgher framerate requires higher write speed.
But you should definitely be able to record more than 4 secs. Try only activating raw_rec module, it gives best write speed.
It would also help a lot if you posted your exact settings under Raw tab.
Here is the 5D MK ii specific thread: http://magiclantern.fm/forum/index.php?topic=11205.0
Firet post pretty much contains everything you need to know.
Dont have cardreader 3.0
And i understand well why the card is too low now.
i buy on internet.. so i think it's a fuckin faked card..
benchmark of old card x400 16GB
Benchmark complete.
Nrite speed (buffer:16384k): 21.4 MB/5
Read speed (buffer:16384k): 48.1 MB/5
Nrite speed (buffer:16384k): 23.6 MB/5
Read speed (buffer:16384k): 48.3 MB/5
Nrite speed (buffer:15625k): 23.2 MB/5
Read speed (buffer:15625k): 48.3 MB/5
Nrite speed (buffer:4896k): 21.3 MB/5
Read speed (buffer:4896k): 38.? MB/5
Nrite speed (buffer:3986k): 21.9 MB/5
Read speed (buffer:3986k): 38.1 MB/5
Nrite speed (buffer:2848k): 21.4 MB/5
Read speed (buffer:2848k): 36.1 MB/5
Nrite speed (buffer:1953k): 28.8 MB/5
Read speed (buffer:1953k): 35.3 MB/5
Nrite speed (buffer:128k): 8.9 MB/5
Read speed (buffer:128k): 12.3 MB/5
You don't have to transfer results manually. Just access card's root and look for BENCHx.PPM files (x=number). Convert it to JPG/PNG, upload it to a picture hoster and link here.
Further indication your card is faked: Card information is blanked. It should look like
http://postimg.org/image/l34zfmr13/
how can i upload image ?.. ::)
http://postimg.org
lexar 16GB x400
(http://s29.postimg.org/7bjrw1l93/bench0.jpg)
FUCKIN FALSE Lexar x1066
(http://s2.postimg.org/5sdx0o36x/bench0.jpg)
DELETE
andludo, should tell us where you buy the fake card from so others can avoid. damn, that sucks
Fast cards will be faked, regardless which seller is involved. Faked products are a common problem and no company can guarantee that all their cards are original ones.
Said this I don't believe it makes much sense to avoid companies in general (if not very strong indications exist that a given address is indeed a scam shop). Let's see how the seller reacts to this.
I strongly suggest for all buyers to benchmark their cards first. Product piracy exists and will do so in the future.
the seller is sorry and will send me another card ... I hope. I've say him I denounce him for fraud. :D
I will keep you posted on the rest of the adventure. ;)
So, I just give you some news!
I sent my card to Lexar for testing, apparently it was a real card, but it was a problem, so they sent me a new card and it works perfectly!
So if you have any problems with your cards, don't hesitate to send them, they are very nice !! :)