My experience with ML/Crop Mood and why I decided that I will not film using it

[Wannabe]

Mod note:
Some of the statements made below are either misleading, or false. You can find clarifications in the following posts.


Let me start by saying that I just removed Magic Lantern / Crop Mood form my EOS 100D. I am not sure if I will ever go back to using it again.
This is not a knock on the ML/Crop Mood developers. You guys are absolutely great.

I wanted to try ML/Crop Mood out, because I have a short film project coming up and I got attracted by the idea of shooting in RAW.
After messing around with ML/Crop Mood and MLV App for a few weeks, I decided that I will not use it at all, and that I will shoot my short film using the regular Canon video firmware with a C-LOG profile installed. This will be more than sufficient for my color grading needs and will allow me to shoot without worrying about crashes, interrupted recordings and lost material.
I will then upscale the 1080p result up to 4K, using artificial intelligence software.


Anyway, here are my thoughts on Magic Lantern / Crop Mood:



Pros of Magic Lantern / Crop Mood:

1. Donate-ware. Free to download and install. The user decides how much he wants to pay for it.
2. Has a great forum with lots of friendly and helpful people.
3. Glitches and bugs get ironed out almost instantly by the developers.
4. Could be an interesting project for geeky programmer people who want to mess around with code and try some firmware hacking.


Cons of Magic Lantern / Crop Mood

1. Not exactly user-friendly. Way overloaded with options, too many menus and sub-menus most of which rarely come in to play. Takes a lot of trial and error, a lot of reading, and a lot of YouTube tutorial watching before you can become comfortable with it. Suitable for geeks, but frustrating for the average Joe. Needs to be simplified.
2. Not reliable. Video recording could stop at any given time. For example: the user can experience an interrupted recording while shooting an important movie scene or interview. Material gets lost and/or ruined forever.
3. Requires an expensive fast SD card that often costs as much as the camera itself (or more). If you don't have such card already available to you, be prepared to drop some extra $100-$150.
4. Requires an expensive computer that can handle and speed up file conversion and editing of RAW files. If you don't have a powerful PC already available to you, that would be an extra expense that you will have to make, and another $1000+ down the drain. The total you are going to pay for this Magic Lantern / Crop Mood setup can reach ridiculously high numbers. In many cases you will not be saving any money, but you will be spending money instead. It is like trying to pimp a really old car - it ends up costing you same money as a brand new car.
If you are starting to build your ML / Crop Mood setup from scratch, it would be way better to just spend all that money on a proper cinema camera (second hand) or just shoot 8 bit video with a LOG profile.
5. Steep learning curve for the post-processing stage. MLV App may be intimidating and confusing to some users.
6. Time consuming post-process (mostly the conversion from RAW video to usable video files). Takes long hours or in some cases even days to convert all your material.  And when you are done converting, you have yet to start editing the converted files. Not suitable for people who are in a hurry to finish a project that has been ordered by an impatient client.
7. Auto focus does not work in Crop Mood. Yes, I know that most cinematographers don't use auto focus anyway, but not everybody has a dedicated focus-puller person at hand.
8. Does not offer real 4K video, but basically upscales 1080p and this results in a soft image which does not have the level of detail that real 4K normally delivers.
9. Some builds are still work in progress and can cause camera crashes (learned this the hard way) and some MLV App features may be missing for your camera model. For example there's still no focus pixel maps for EOS 100D.


That's just my 2 cents on this subject. This is my personal point of view: the point of view of an ordinary guy and an ordinary user.
Once again: no disrespect to all the brilliant people who have put countless hours of hard work in this interesting project. Rock on, guys!

I will be following the further development of ML / Crop Mood with greatest of interest.

[names_are_hard]

I'm glad you found the community helpful!  It's true, ML is more complicated and less polished than a normal cam UI - this is to be expected given the resources involved, and we do explain this right at the start.  For some people it's worth working with, for some it isn't.  If you need very high reliability, it's probably not a good fit.

I would like to correct some factual inaccuracies, for other people reading.

Donate-ware. Free to download and install. The user decides how much he wants to pay for it.

ML is not donate-ware.  It is free (libre & gratis) software, nobody has ever been asked to pay for ML.

Requires an expensive fast SD card

Untrue.  You do need a card capable of certain speeds for raw video, but:
- not everyone needs to shoot video at all, ML has lots of stills features
- required speed varies depending on the mode you want to shoot in, so you never *need* a fast card
- cards are not as expensive as you say

be prepared to drop some extra $100-$150

I have no idea where you're buying your cards but you're getting ripped off.  Amazon is not the cheapest, and they have Sandisk Extreme Pro 128GB for $22 currently (EDIT: $14 for the microSD variant, including adapter).  And that's an expensive brand, better than you need to hit max speeds!

Auto focus does not work

It does work, although there are some limitations depending on mode.  And, the Canon cameras that ML supports are fairly old, with not great AF to begin with.  Notable exception being 70D, which has DPAF.

Does not offer real 4K video, but basically upscales 1080p

This is simply mistaken.  Capabilities vary per cam but resolutions *higher* than 4k can be obtained.

No offence taken, just wanted to get the facts straight for anyone reading in the future.

[Wannabe]

ML is not donate-ware.  It is free (libre & gratis) software, nobody has ever been asked to pay for ML.

Correct me if I am wrong, but Magic Lantern was created about 14 years ago as a paid addon for the Canon cameras. So, yes, people have been asked to pay money for it before.
Also, it is my understanding that Crop Mood was built by raising funds from donations (nothing wrong with that, of course), and I think people are still encouraged to donate funds and equipment for its further development.
This is why I labeled it "donate-ware". Maybe the label that I used was not very suitable.

I have no idea where you're buying your cards but you're getting ripped off.  Amazon is not the cheapest, and they have Sandisk Extreme Pro 128GB for $22 currently (EDIT: $14 for the microSD variant, including adapter).  And that's an expensive brand, better than you need to hit max speeds!

Amazon is not a good place to shop for an SD card. They sell copious amounts of fake SD cards. The Amazon guys know about it and they do nothing to prevent it. A good 256GB Sandisk Extreme Pro SD card that is guaranteed to be genuine will set you back at least $80. In some countries (like mine) much more.

- not everyone needs to shoot video at all, ML has lots of stills features

The vast majority of people use ML / Crop Mood for shooting RAW video.
For shooting RAW stills you can always use Canon's stock firmware. You don't need ML for that.

Quote

    Does not offer real 4K video, but basically upscales 1080p

This is simply mistaken.  Capabilities vary per cam but resolutions *higher* than 4k can be obtained.

Yes, in terms of pixel count you can get resolutions higher than 4K with ML / Crop Mood, but in terms of quality, the result is not real 4K. It is upscaled 1080p and a good portion of the detail gets smudged after the video gets stretched up to 4k+.

[Danne]

5diii for stability or an eos R6 for 4k compressed output. H264 on 100d, I think I rather use my phone to record with .
I still think a worked out 1x3 setting on the 100d will work very good working with 24fps.

[Wannabe]

5diii for stability or an eos R6 for 4k compressed output. H264 on 100d, I think I rather use my phone to record with .
I still think a worked out 1x3 setting on the 100d will work very good working with 24fps.

Yes, I am happy with the Crop Mood 4.4K preset on my 100D. It is stable in low light and at 10 bit. I have not experienced any problems with this preset. Really cool.
Still, I have to play it safe, so I will film with Canon's native firmware - it shoots in 8 bit, but it now has a C-LOG profile installed (for color correction) and it guarantees stability.

I hear you about shooting with a phone and I have considered this option as well. Unfortunately, a phone cannot create a blurred background, which is an important component of the general "cinematic look". I need a large sensor and a proper lens for that. Canon 100D offers both.

[names_are_hard]

I'm going to be brief but I'm not trying to be rude - I just don't want newcomers to be misled when you confidently state incorrect opinions as fact.

Correct me if I am wrong, but Magic Lantern was created about 14 years ago as a paid addon for the Canon cameras

You are wrong, it has always been free and licensed GPLv2.
https://trmm.net/Magic_Lantern_firmware/

A good 256GB Sandisk Extreme Pro SD card that is guaranteed to be genuine will set you back at least $80

You are wrong.  Sandisk have an official store on Amazon, which will not send you fake cards, and that's where I quoted the price of $14 for 128GB from.  You can also buy from Sandisk direct (prices differ due to different deals but are far below what you claim).  In addition, you don't need to buy Sandisk.  Multiple other, cheaper, brands have been tested to be fast enough.

The vast majority of people use ML / Crop Mood for shooting RAW video.

You don't know if this is true or not, do you?  How would you tell?  Regardless, your claim was that ML required a fast SD card.  It does not.  Shooting raw video *may* require a fast card.  ML does not.

In terms of quality, the result is not real 4K. It is upscaled 1080p and a good portion of the detail gets smudged after the video gets stretched up to 4k+.

You're making two claims that are both wrong.  You can shoot in true 4k if you want.  This will demand higher data rates and probably won't allow filming continuous shots, even on the 5D3.  But you can do short shots.
You can also shoot higher than 4k; full sensor size, limited to low frame rates.
You can also shoot in a variety of resolutions that are higher than 1080p, for continuous shots, e.g. 2048x1152 or 2560x1320.

[70MM13]

magic lantern is nothing short of miraculous.

the capabilities it adds to these cameras put them in a league of their own, plus it extends the usable lifetime of consumer oriented devices that would otherwise be in landfills further poisoning the planet.

i'm filled with gratitude and awe when making beautiful footage with a 10 year old outdated camera in this society of gear that is outdated 5 minutes after release...

that being said, it is not for everyone.

it's for everyone who wants to enjoy tinkering and "cheating death" with gear being pushed to the limit.

if you cannot tolerate a single unexpected stop or total crash requiring a battery pull, don't use it.

i've made hundreds of videos over the years with magic lantern (my channel has hundreds and a lot have been deleted over the years, in fact i will be culling them again soon) and i can tell you from vast experience that in the best case scenario you will experience lost takes at least 10% of the time.  sometimes there can be a battery pull failure followed immediately by another.  it's rare, but it happens.  it just did yesterday.  other times, there can be 20 safe recordings in a row.  obviously the longer the recording, the riskier the operation, which automagically makes it unsafe for very serious filming.  the last thing you need is to get increasingly worried about a camera failure the longer (and therefore more important) the scene you are filming...

in the end, it's a trade-off.  only you can decide what's best for you.

i don't care about losing a take.  it's a chance to do it better again.  but it costs me nothing.  i only wish the damned beeps were loud enough to hear.  i don't have a camera operator, and it is a seriously unpleasant experience to walk back to the camera after a 15 minute talk to find that recording stopped long ago.

this is the price we pay to use magic lantern.

i'm happy with it.  i love it.  i'm not interested in changing my camera until it dies.

i find it wonderful to look forward to future improvements in magic lantern.

but for each of us, this is an individual decision and an individual experience.

if you want the power of what magic lantern CAN DO, you must get a good camera such as the 5d3 and be prepared for the hassles and costs involved, but in the end it is capable of producing video that is world class if you can master it.

the quality of my video speaks for itself.

in the end, a camera is just a tool.

[DarkTrapStudio]

You want to make a film in H264 ? And I guess stock canon software is interframes H264 codec only ?
And you need Autofocus for that film ? You better shoot your film like Better Call Soul in that scenario
*instant edit : H264 Interframes codec is harder on the computer than RAW, + you can export in whatever codec you want as youll need MLV to transcode, I hope you will reconsider your opinion/knowledge, or please make argument where Im wrong !

[Skinny]

I don't know how stable 1x3 presets are, but on many cameras default 3x3 mode is usually very easy to use and almost never crashes, just press record and enjoy.. it will be the same resolution as native h.264 but it will have more details and of course more dynamic range and everything you will expect from raw. If you use 10 or 12 bits you don't even need a fast card.

It is just better than native h.264 in every aspect, except for:
You need more GBs
It takes some time to process, although it will be faster than 3k or 2,8k processing
And there is no smooth review of the recorded material in camera.



P.s. I understand why you want to use standard h.264 recording for your project, if it is more about the story and you don't have a lot of $$$ or time to make everything perfect, it is totally ok..

But even standard nightly build or any very stable ML build for your camera can give you some advantages when recording in h.264, such as - histograms, false colors(!!!), focus peaking, 10x zoom on half shutter (for focusing), disabling auto gain for microphone and so on.

[masc]

4. Requires an expensive computer that can handle and speed up file conversion and editing of RAW files. If you don't have a powerful PC already available to you, that would be an extra expense that you will have to make, and another $1000+ down the drain.

6. Time consuming post-process (mostly the conversion from RAW video to usable video files). Takes long hours or in some cases even days to convert all your material.

Apple Silicon computers start at 599$ and convert up to realtime MLV to e.g. ProRes.

5. ...MLV App may be intimidating and confusing to some users.

Tell us more. That doesn't help to improve. Or improve it on your own and give us a pull-request, so any user can profit.

9. ... and some MLV App features may be missing for your camera model. For example there's still no focus pixel maps for EOS 100D.

Focus pixel maps are no MLVApp feature. A map is a file MLVApp reads and/or creates. MLVApp removes pixels for any camera, even MotionCAM or whatever if you need and want. You can e.g. remove all pixels using your own map for any camera. Also here: you then could make a pull request and everybody can use it.

[DarkTrapStudio]

There is a full comprehensive MLV course on youtube, you waste people time, no pro edit software is less intimidating than MLV app stop wasting people time. This is not mean, this is objective.

[names_are_hard]

I think the original post was okay for the aspects that were actually opinions, or gave impressions about how someone was feeling, e.g., "Steep learning curve for the post-processing stage".  This is reasonable feedback and it's useful to get it.  Of course, it would be more useful with actionable detail, but feedback has to start somewhere.

Most of the post was "ML can be confusing, technically hard and unreliable" - this is true.  It's okay to not like using ML and not want to use it.  People have different tastes, priorities etc.

I wanted to correct the parts that were factually untrue, because that might mislead future users.  For the rest of it, maybe best to let someone who didn't get on with ML just walk away?  It's not a big deal.

[70MM13]

The impulse to be unkind is not a beneficial one, not to the aggressor, nor the "victim", the community, nor the reputation of the project.

People, when belittled and judged simply because they reported their experience,  tend to be harmed as a result.  Is this what anyone wants?

I speak from multiple experiences and I pity all parties involved.

To the original poster: forgive them for they know not what they do.

[Skinny]

It would be actually cool if MLV App had an option for someone who don't want to spend a lot of time, something like quick export everything in Arri Log and auto-level adjust for each clip so they are not too over or under-exposed. People say it is important to have proper exposure level before converting to log. Just like one button solution, select a folder and everything is done, exported in prores 4444 for example.

I know it is not ideal. But it is a very quick and easy start for people who work with other cameras usually shooting log profiles, so they can color grade in whatever software they like.
And if they want to have a better control over everything they can learn MLV App features..


MLV App is great but it's interface and sliders behavior is different from most software (premiere, resolve, even lightroom) and also there is a lot of features that can confuse new people, like RAW section and profiles with transfer function and everything. I get it when someone say it is difficult to achieve what they want in MLV App, you need some experience with this software. I usually tweak sliders and curves in certain way which I know will work in MLV App and only in MLV App, if I open the same frame in Adobe software I will do completely different things.

So you really need to know MLV App behavior to use it effectively, it is a fact. Which is another layer of complexity, while with other cameras you usually need just your NLE which you already know how to use.

[a.sintes]

or just don't use MLVApp at all: rely on MLVFS and work directly with Resolve? anyway you will have to learn a little about color science and color grading in all use cases...

[masc]

I hope we don't have agressors or victims here. I just wanted to say, that we need to know, what exactly could be better to make it better. Or even better: the user who needs a feature creates it and everybody is able to use it - ML is a community.

@Skinny:
Currently we already have Quick-Export receipts for MLVApp (find them in the repos) - just choose one of these receipts as default, open a folder of MLV, export to whatever you like. Ready. If someone needs something similar with e.g. ARRI-Log, create a receipt and create a pull-request. So everybody can use it.
Auto-adjustment indeed is not available atm - very complex theme.

@a.sintes:
Use whatever software you like - but you'll have to learn this one in all cases.

[DarkTrapStudio]

For my Part @namesarehard, at no given times I wanted to no let you go, what I was explaining is that you lack knowledge until proven otherwise, based on some things that you say that I pointed out precedently, I only seek for arguments/counterarguments and I value more the conversation when Im wrong than when Im right cause I can learn something from my errors, and that's what I wish for you.

[reddeercity]

nice A.I. response , who talks like that ?
I started using Magic Lantern (2011 i think or 2012) not for Raw Video , but for Clean HDMI feed and all the feature that make a Picture camera
in to a Video/Film camera (i think everybody knows what i'm talking about)
and with technicolor cinestyle picture profile i was in heaven !! and with the H264 bit rate controller
I was getting over 75Mb/s H264 video in cine log image , as long as you expose the image correctly
you had a amazing image that hold up even today , there said when the Japanese engineers first
saw "Vincent Laforet"  "Reverie" 
There all had tears in there eyes , the engineers where so moved !
https://www.youtube.com/watch?v=UPh4sy4shMI

https://en.wikipedia.org/wiki/Canon_EOS_5D_Mark_II#Independent_film_and_television
just a small list of what the 5D2 did without raw video in film & video production
and some i bet where using ML

Magic Lantern Can Be for everyone its really that easy , but you can make it as hard as you what
it all depend what you are liking for, if you what Raw Video like a "RED" or "Arri Alexas" then you must
be prepared for some work & new learning curves , if not use a Iphone     

edit:I usually read the forum 2-3 time a week so i can keep up 

[DarkTrapStudio]

nice A.I. response , who talks like that ?

Are you reffering to me ?

[iaburn]

Many of the cons are related to RAW processing in general, like the need of a fast computer, the expertise needed to get the look that you want or the increased time until you get the final video over an already backed-in source video.

To also mention "impatient clients", implying professional paid work, but you say it's the opinion of an ordinary guy and an ordinary user, which is odd.

From your comments I get 2 things:

First, RAW recording and processing in general is way beyond your needs and more than you can deal with, and this has nothing to do with ML.

Second, you didn't expend the time you needed to understand the pros and cons of ML, because you felt overwhelmed to soon.

So it's ok to express your frustration on the forum, but I'm happy that you stand corrected by the gurus here so other people reaching this post don't get scared 

[DarkTrapStudio]

Many of the cons are related to RAW processing in general, like the need of a fast computer

Again until proven otherwise this is false, He is mentionning that he want to use H264 on stock canon software (So Interframes/GOP I guess) and GOP is harder on any computer than RAW, even on my M1 mac if I begin one or two effect I can't get smooth playback at 1080p, this is basic theory as the computer have to decode theses GOP (Group of pictures) when RAW is an Intraframe "codec" meaning that every frame is a still and there no decoding to do, like every pro grade codec (I know Raw isn't really a codec Its for the it's for the convenience of the explanation) like Pro Res, Cinema Dng, DNxhD...


This is why I explain to him in that manner, don't blame ML, if you he can't teach himself the knowledge, anything will be harder than MLV App/ML in the pro video ecosystem, there no shortcut, know your stuff and acknowledge you ignorance rather than saying something is not good, this is not respectful for our time while everyone is here to help freely.

[DarkTrapStudio]

https://ibb.co/rdfBmVM

As you can see here Its H264, they don't mention H264 Intraframe so you can be pretty sure is Interframe, and Upscaling interframe also is a bad idea in my opinion but Im not a pro so may ask from someone with authority I just based what I said on the GOP movement artefacts this codec produce.

On top of that, if you want to upscale Intraframe I guess Its gonna be really painfully long if don't transcode but Im less sure of that.

[Skinny]

I don't know why but when I edit projects with stock h.264 footage I have realtime playback, I can do it even on very old computer, but real-time RAW was always something impossible. I used to have a laptop from 2012 and now my main PC is core 2 duo 2 ghz from 2007, with ssd but really old graphics card, it can't even handle realtime h.264 most of the time, although it works, I have about 4-5 seconds of somewhat realtime playback...
But RAW is just out of the question, and if I use MLV App for color grading and export everything in any of prores (even 422) then I can only edit everything without playing the project.
Anyway I export everything in prores 4444 and cut using my imagination So nothing can stop me lol

But it is more demanding if you have an old computer, maybe on relatively new hardware you have different behavior.

[DarkTrapStudio]

Interesting, I have smooth playback on H264, but not when I use differents effects, but also Its difficult to make objective conclusions if we don't make an A/B test with the same parameters (Old vs new computers with same footage/codec/timeline/video software).

[Skinny]

We can test it, if you want My computer vs yours on same raw footage
it doesn't even have to be the same, just uncrompressed raw (I don't have a camera with compression enabled), same resolution and bit depth

I can only test with MLV App and adobe premiere, (both raw cdng in premiere and prores or h.264)...
Can't run resolve here, because the graphics card is too old. Radeon something 4580, if I remember correctly, people say you need very ancient version of resolve (like resolve 8 ), and the oldest version I found is 11, it can't see the graphics card. It's already too modern hahahah