Messages

Not to worry, I found the name, it is stored in the meta under two different types which look like this:
File Index                      : 7716
File Number                     : 103-7716

Unsure as to the extra number on file number, but file index, just add IMG_ to the start and .JPG to the end.
Was doing a massive SD card recovery and wanted to not have the scrambled jpg names that the recovery produced.

Using exiftool (on linux) they can be renamed with this simple command, !however! be warned, I think it will overwrite duplicates if two images happen to have the same file index:

Code Select Expand

exiftool '-FileName<IMG_$fileindex.JPG' *

More a question of the actual canon firmware and how it writes its images, I guess irrespective of Magic Lantern.

Does the original file name of the jpeg, ie IMG_9876.JPG, get stored somewhere in the exif data?
It stores all the picture style and sharpness and what not for the sake of the camera reading that is the image number given by the camera stored anywhere in there?

Thanks.

(very long time since I've posted here, still a big user of ML though)

We do need to upgrade to 1.1.0. Didn't someone test and find no issues? It was a pretty minor update I thought.

Wow. Replies.
Including unhelpul +1 spams 
I don't think anybody is using discord as archives. At the time it was something I needed to help newcomers, everything else would've as usual stayed in any forum category. I don't think  there is any software development taking official feature requests through discord.

Thanks for the discord links, I will check them out!

Oh I thought you were struggling to get it to work on Linux, I read it wrong.

What I've heard for windows is, if you install the canon and eos software, eos utility and what not, the camera should be able to act as a webcam.
It's something like that, there is a driver for it to be a webcam.

it's an mjpeg stream that comes through it, I never figured out how to get it to act as a webcam in linux,
I could only get the stream through whatever method.
maybe search for canon dslrs and v4l2?

If you're writing a guide shouldn't it be stopping questions being asked? If a guide is created that creates questions wouldn't that be the opposite of a guide, some sort of ediug or something.

Just do it, just start it, people in here will probably go "I don't need it" that's because they're already using it. Any criticism; don't take harshly, any feedback; implement as needed. But if you post what you've posted and then wait for responses it's nothing but wasted time.

As far as your second question goes, if you have something, you can create a thread and link it and people will comment along the way, you can post updates in that thread and what not. But I don't think anyone will commit to a proof read of something that doesn't exist. Write the first chapter, start a thread, link it, get the feedback there, keep the thread alive, keep the project alive, etc etc.

After pondering on the first response and the image from the second response and combining the two responses I think i can see whats going on now.

I would guess it has something to do with the composition. Example: the horizon should normally not be placed ideally in the middle of the frame but rather 1/3 or 2/3 of its height (or maybe golden ratio). To satisfy this rule for all formats, it requires matching the position of 1/3 of their height, not their centers.

Ok, but why not frame the shot so that it is 1/3 or 2/3 the way down, instead of cropping off centre for uneven lens distortion.

Probably because the material is created one camera for various later playback formats...

This argues for why it shouldn't be placed there, those crops happen in the centre.

A question for anybody who has the correct answer;

When I see behind the scenes stuff for movies, they have their 16:9 monitor with multiple crop marks, but for the "cinema wide" crop mark (greater than 2:1 (the 2.39:1 or what not)) it is never positioned at the centre vertically...
Why?

Ah thanks very much, I had a slight feeling that it already existed so I just wanted to check, but I guess it doesn't then, I'll check out the links you've posted.

I don't understand the question, probably.
Because you can apply any pic style to raw files with DPP
https://global.canon/en/imaging/picturestyle/file/dpp.html

Thanks for answering with exactly what I said not to answer with.

Is there a way I can have the picture save on the camera into two different picture styles?
And in particular; two custom picture styles?
(inb4 "just use raw")

edit & explanation:
I want to take the photo in jpeg, and instead of it saving one photo in "neutral" or what not,
I want it to process and save 2 jpegs of the same picture, one in neutral and one in say standard (or custom picture profile).

Sweet! I'm Australian too. I'm on a 550D though.

Welcome.
I bought my 550D brand new back in the day, that was a while ago, installed ML on it straight away.
There wasn't much option for second hand back then but it's definitely the way to go nowadays (for most part).
Might get myself some more DLSRs of the same era in the future.

This should now be fixed, I got an update on both Linux and Android.
However a quick fix for it was to go to about:config and set xpinstall.signatures.required to false

If you put a strobe light in your toilet you'll actually see an animation of your internals while taking a piss.

I know that this sounds crazy, but is 8bit RAW possible? Sorry if I sound like a noob.

The best feature of shooting RAW on my 5D2 is the aspect ratio control. I can use a lot more of the sensor than shooting in H.264 allows me to. Shooting in 3:2 and 4:3 allows me more creative control in shooting for instagram/facebook because of the larger vertical resolution these sites are intended for. It also helps when shooting anamorphic because of the 1.33x horizontal squeeze!

If 8-bit raw sounds too weird, then is there a possible way to control aspect ratio when recording in H.264?

You won't get any benefit from it, the size is pretty much the same as 10 and 12 bit.
Plus I would assume it would result in banding.

You should probably update your ML firmware.
https://builds.magiclantern.fm/

The first digital camera I actually bought (and I was quite late to the party of digital cameras in terms of owning one myself), was a canon and I made sure it was one that CHDK could run on.
It was in beta for this camera but worked perfectly fine at the time, back many versions ago. However when I tried raw it revealed how terrible the sensor in the camera was by showing the colours it actually captures compared to what it should be made available through the picture style.

I did find it worked with my brothers EOSM mp4s,
but as far as the movs from my 550D, turns out it doesnt like the audio,
so it will only cut and export if i select it to export with no audio.

Unfortunately it doesn't seem to do my MOV clips from my 550D, they play in it though.
I think the playing is done through chromium, then the muxing is done in ffmpeg.
I'll have to play with it a bit to see if I can get it to work, plus i have to try it on some other clips as well.

Just came across this! Been wanting something like this for a long time.
(I have not yet tried it myself)

https://github.com/mifi/lossless-cut
"Save space by quickly and losslessly trimming video and audio files"

This is not an advertisement, I just found it and thought I definitely need to share.
It runs off ffmpeg, and I was actually unaware ffmpeg had such a cutting capability.

Anyway, hopefully this means no more re-encoding a lot of videos for me.

Good idea!