Magi Lantern and RAW workflow doubts

rawfa

rawfa

Veteran
Haya fellas,

So, with all this talk about how Magic Lantern is enabling RAW recording on the MarkIII (and possibly other Canon cameras) I have some questions regarding the entire workflow, but specially post:

1 - I´ve been shooting a web series and each episode (around 5 min each) ends up with around 100 minutes of footage. How much disk space are we talking about here?

2 - What is the more standard editing workflow for raw? Do people use an intermediate codec like cineform or just go with an ultra powerful computer?

2.1 - Is most people go with an ultra powerful computer, what is the system requirement smoothly edit RAW? I´m talking CPU, Cores, RAM, Disk space and speed, GPU, etc. I have limited space at my place and I was wondering if it is remotely possible to edit RAW on a powerful laptop or if it´s just crazy.

3 - This is more of a development question and being so soon in the development process it might not have an answer now. Is it possible to record Full HD 24p RAW into fast SD cards? I keep hearing everybody talking about CF x1000...which basically would mean that you have to go with either the MarkII or MarkIII (I think these are the only Canon video DSLRs that have CF ports).
 
Last edited:
* 100 minutes of RAW recording at 1920x1080 and 24 fps means 522 GB
* por optimal quality, import the DNG stream directly into AE; if that's too painful to edit, do your color correction there (not the grading, just the first light), and export to prores (quality will be a lot better than recording prores from the HDMI)
* even with a decent workstation (intel 3470, lots of RAM, nvidia graphics card, and SSD) playback in AE will be at less than 2fps; no idea about slower systems
* SD cards: still unknown, but it seems like the SD controller will be a bottleneck
 
rawfa said:
Haya fellas,

So, with all this talk about how Magic Lantern is enabling RAW recording on the MarkIII (and possibly other Canon cameras) I have some questions regarding the entire workflow, but specially post:

1 - I´ve been shooting a web series and each episode (around 5 min each) ends up with around 100 minutes of footage. How much disk space are we talking about here?

2 - What is the more standard editing workflow for raw? Do people use an intermediate codec like cineform or just go with an ultra powerful computer?

2.1 - Is most people go with an ultra powerful computer, what is the system requirement smoothly edit RAW? I´m talking CPU, Cores, RAM, Disk space and speed, GPU, etc. I have limited space at my place and I was wondering if it is remotely possible to edit RAW on a powerful laptop or if it´s just crazy.

3 - This is more of a development question and being so soon in the development process it might not have an answer now. Is it possible to record Full HD 24p RAW into fast SD cards? I keep hearing everybody talking about CF x1000...which basically would mean that you have to go with either the MarkII or MarkIII (I think these are the only Canon video DSLRs that have CF ports).
Click to expand...

We don't even know if it will work on anything but the Mk2 and Mk3.
 
Samuel, how much time do you need to transcode from AE to prores on a good computer? lets say 1 hour of material?
 
rawfa said:
1 - I´ve been shooting a web series and each episode (around 5 min each) ends up with around 100 minutes of footage. How much disk space are we talking about here?
Click to expand...

Beyond that, could you shoot your web series using only 55 second takes? Because ML currently has that limitation. And there are only going to be two workarounds:

1) ML automatically starts up recording again. This will result in a loss of frames every time, probably unacceptable.

2) ML changes the way it records to being single files in a folder, similar to a BMCC. I have no idea if this is possible or not.
 
Gary Huff said:
Beyond that, could you shoot your web series using only 55 second takes? Because ML currently has that limitation. And there are only going to be two workarounds:

1) ML automatically starts up recording again. This will result in a loss of frames every time, probably unacceptable.

2) ML changes the way it records to being single files in a folder, similar to a BMCC. I have no idea if this is possible or not.
Click to expand...

If 2) is possible, then 1) without lost frames is possible too (99% sure of that)
 
Gary Huff said:
Based on what? Your gut? Do you understand the bandwidth requirements necessary in order to accomplish spanning?
Click to expand...

Yes. Basically, the same ones as those required for writing a different file for every frame.
Specially since in this case the numbers of frames you can write before spanning is predefined. If you can write a file for every frame, you can write a file for every two frames. And if you can do that, you can write a file for every 1320 frames.
(I'm not saying this will definitely be possible, just that if 1 is possible then 2 is possible too, 99% sure)
 
Samuel H said:
Specially since in this case the numbers of frames you can write before spanning is predefined. If you can write a file for every frame, you can write a file for every two frames. And if you can do that, you can write a file for every 1320 frames.
Click to expand...

It sounds like you have it backwards. When you span, the system has to have two files open for a brief moment of time, and for that brief moment the needed bandwidth is doubled. And if something fouls up because the card cannot be written to in that brief span of time that two concurrent file operations are going on, then you have the recording stop.

It seems like the 1000x cards are just fast enough for writing out the Raw files, not for supporting twice that needed bandwidth.
 
Gary Huff said:
It sounds like you have it backwards. When you span, the system has to have two files open for a brief moment of time, and for that brief moment the needed bandwidth is doubled. And if something fouls up because the card cannot be written to in that brief span of time that two concurrent file operations are going on, then you have the recording stop.

It seems like the 1000x cards are just fast enough for writing out the Raw files, not for supporting twice that needed bandwidth.
Click to expand...

Do you know this or are you stating this based on your gut too?

Creating a next file doesn't require doubleing bandwidth.

if size is equal to 4gb,
then create a file named current number + 1
file name = current file name + 1
record next frame to file name.
 
bumkicho said:
Do you know this or are you stating this based on your gut too?
Click to expand...

Why do you think certain SD cards could handle a GH2 hack, but would fail upon spanning? Because it was enough for the single file and not for two files concurrently is the answer.

How do you think you make a seamless division between two files that are recording information at a rate of, at least, 24 frames per second? By handing off to a file that's currently open and ready to receive data even as the previous file is being closed up.
 
I'm not saying extra bandwidth is not needed for that brief moment when you (may) have two files open at the same time. All I'm saying is that if you have those extra requirements for spanning, you have them too for single-frame-file writing. So, almost sure, it's all or nothing.
 
Samuel H said:
I'm not saying extra bandwidth is not needed for that brief moment when you (may) have two files open at the same time. All I'm saying is that if you have those extra requirements for spanning, you have them too for single-frame-file writing. So, almost sure, it's all or nothing.
Click to expand...

I don't know this for certain, but I don't think it does because the process should be fast enough that all those files are written sequentially. No need for starting the write of a second frame while finishing up the first.
 
I would love to have it in one big file. Formatting a card in EXFAT is not a big deal to me. You shoot, download footages, format a card in EXFAT, and pop it back in.
 
bumkicho said:
I would love to have it in one big file. Formatting a card in EXFAT is not a big deal to me. You shoot, download footages, format a card in EXFAT, and pop it back in.
Click to expand...

The problem is that you're more prone to losing everything in the event of a failure rather that just a portion. I much prefer spanning to a single file.
 
If two files with 1 frame each can be written sequentially, then two files with 1320 files each can be written sequentially, almost sure. Anyway, not that important. We'll see.
 
While the RAW news is certainly an awesome achievement, I think most users are getting WAY ahead of themselves by getting excited about RAW. Even in best the cases, you'd probably just want to save this for extremely demanding, high-detail b-roll. It's just so big and so processor-intensive. Even the BM Pocket Cam has compressed RAW, which should be easier on your machine than the RAW DNG's this ML hack will create.
 
You must log in or register to reply here.
Back
Top