Show Posts

This section allows you to view all posts made by this member. Note that you can only see posts made in areas you currently have access to.

Messages - Ilia3101

Pages: [1] 2 3 ... 12
OMG 😍😍😍!!! waza57 your the person that made video dual ISO on 5D2 almols work rigtht??!

Can you record MLV yet with this confuguration? If not, how close is it??? And how could I help? The goal would be to put it in a single menu option in a normal build, how much work is that? I would love to help if possible. Or test it  ;) ;) ;) ;) ;) (as every noob always says)

I'm so excited!!! 2.8K—that's Alexa sensor resolution (it must be a sign)

@Andy600 Thanks for the explanations. So ColorMatrix2, when inverted, describes a transform from "CameraRGB" (debayered) to XYZ with D50 white point.(?)

So... when that image in XYZ is transformed to sRGB (for viewing) it will be the same as having the white balance slider set to ~5000k in a raw converter and looking at it?

And it makes a lot of sense that the sensor would respond differently at lower colour temperature, never thought about that before. But I'm a little confused as to what the temperature of ColorMatrix1 (the ones we have in ML code) is... where do I find out?

Oh wow all the code on her makes MLV App colour management look horrific :( It's a bit better than all of the unused code makes it look (well, in the whitebalance branch at least) :D

@Andy600 I have been wondering for a very long time, is ColorMatrix2 D65 white point? I have assumed this, and it seems to match(??), but not tested it with actual comparison. If you could tell me definitely if it is D65 or D50, that would make me quite satisfied. These are adobe matrices I think (is this right bouncyball?)

I love to test out the lower bit but Where can i get the 10/12 bit build you're using, also which build do you recommend for mlv_lite for 1880x1056?
Thanks for the Advice
You have to compile the 10/12bit branch yourself (it is the one that allows 1880x1056), I'll see if i can find a build on my computer, if not I'll compile it.
I'll send it to you.

Also are you guys looking at the new ProRes RAW Codec from Apple?
Seems like its 100% proprietary and will take a while before anyone except big companies can put it to use. So not yet (possibly not for ages who knows).

Also download the new version everyone!!!
I see it has very few downloads compared to the old one. Improvements and bugfixes are important.

@dariSSight masc's advice is good I agree with all of it. However I always use a 10/12 bit build for the 5D2, it's surprisingly stable, the reduced bits saves space when recording in crop mode(however it corrupts normal shots because it's not fully working yet). mlv_lite also gives a bit more horizontal resolution: 1880x1056 vs 1856x1044, but I don't use it.

Does this work for manual lenses?  :D 🍳🐣

Raw Video Postprocessing / Re: Question regarding ML RAW > ACES
« on: March 28, 2018, 12:13:47 AM »
Highlight/shadow control coming in next release which might even be really soon. Even before WhiteBalance branch is merged.

Camera-specific discussion / Re: Canon 5D Mark II / 5D2
« on: March 27, 2018, 10:51:20 AM »
So y ou're saying 12 bit fixes moire patterns on the 5D2? very excited to watch the video and test it out...

This means you have somehow changed the binning as a side effect?


Raw Video Postprocessing / Re: Question regarding ML RAW > ACES
« on: March 26, 2018, 10:36:10 AM »
That would be fun and interesting.
@Andy600 contact me if you have any questions if you do this.

For testing:
Use the "WhiteBalance" branch (soon to be merged) which should have correct sRGB/rec709 output. White balance in that is however done in XYZ, not yet any kind of LMS space. Me bouncyball or MASC can provide you a build for whatever OS your using.

Also the "log profiles" are nothing more than curves (no colour stuff).

Current mainline MLV app has absolutely no "colour accuracy", so colour testing should only be done in WhiteBalance branch or wait for merge.

There is also the parameter with color science which I don't fully grasp and where Mlv App is right now.

Colour science: MLV App is getting better in this soon. Current version has no camera calibration (those matrices not used), but in a version maybe two that will be implemented (as well as "correct" whitebalance stuff that g3gg0 was explaining ages ago) - You can have a preview of this stuff if you look at "WhiteBalance" branch on GitHub. Colour matches Davinci/Adobe quite well with new colour processing methods.

Output colour space: it's sRGB/rec709 I think. I hope we figure out how to encode in wider space with FFmpeg (already can with AVFoudnation), as output in other colour spaces will soon be supported in processing(watch that "WhiteBalance" branch).

@Danne You can launch MLV App from bash with mac open command
Code: [Select]
open -a "MLV App" file_1 file_2 file_3 This will launch default version of MLV app on the Mac and load all the specified files in to it.

One thing where I believe the Qt app is worse than cocoa app is the control sliders, you can't just click anywhere on the slider and start moving it, you have to click the knob, which feels a lot clunkier to me. Could you make the slider draggable from any position @masc?
Maybe those who like cocoa app would be happier if the Qt version had a simpler theme, but that isn't really a big deal for me.

As for Cocoa app coming back, probably not. It's currently in a bad mess and doesn't work fully (the last release is about as good as it is going to get). It's just too much work for what it does - it was written without XCode, no interface builder which I think Apple never intends anyone to do. Also written by me with no clue about how Objective C works (the real issue).

It's my fault that white balance is wrong. I'll have this fixed some time soon (I've written a really nice black body simulator which I'll put to use in white balance). Also matrices.... I've almost got to that :D

@IDA_ML I have ran it on a Lenovo x61 from 2007 with linux. It's probably something to do with windows being bad as usual.

Reverse Engineering / Re: ProcessTwoInTwoOutLosslessPath
« on: February 24, 2018, 02:00:03 AM »
I feel we're getting close to lossless on the 5D2.

Is DecodeLosslessPath necessary? Would it be dangerous to compile and run this? Try it and let us know if you and your camera survive.  ;D
I tried, but there seems to be a function that is missing from the 5D2 codebase that won't let it link. I'll try to get rid of that error tomorrow.

@reddeercity have you had any luck compiling, is it even time to try it yet?

I've also been thinking about a pattern matching script...

Reverse Engineering / Re: ProcessTwoInTwoOutLosslessPath
« on: February 20, 2018, 08:30:15 PM »
How necessary is DecodeLosslessPath for lossless recording? Also would it be dangerous to compile and run it "where we're at" right now?

Is the pattern matching just finding really similar blocks of assembly instructions in the disassembled code of different cameras?

Reverse Engineering / Re: ProcessTwoInTwoOutLosslessPath
« on: February 18, 2018, 11:18:43 PM »
Can confirm VSCode is only good thing Microsoft ever made.

I'm a bit confused as to how dfort found the addresses it seems more complex than just finding strings in the disassembly :( Can't imagine I'll ever figure this out.

There shouldn't be much of a difference between this and a LUT in terms of what it does(just does it a lot differently and slower :D), and it was even trained on some samples that someone I know filmconverted(shhhhh) WHOOps I mean real film

LUT support might be added at some point too, they seem quite simple actually.

The "AI" only has 3 inputs and 3 outputs... I would imagine its difficult to do any damage with just three floating point variables. Don't worry :D

Raw Video / Re: 5D Mk III - completely glitched out clips
« on: February 16, 2018, 02:19:14 PM »
What about dogecoin? It has some of the lowest fees.

I might have donated a bit just now (sorry not much) ;)

Reverse Engineering / Re: ProcessTwoInTwoOutLosslessPath
« on: February 16, 2018, 12:50:11 AM »
Ok this is starting to smell exciting again :o If the mk2 goes lossless... that will be really exciting

I can help with/test whatever tomorrow. And remember: upload the code if you have something :)

@a1ex How does it look to you?

Best release yet!

AI (neural network) powered film emulation filters!!! <-- Maybe a first in open source image processing software! (gonna overtake adobe at this rate)
Much better performance in export and playback (which is often realtime now)
All of the other problems complained about should be solved.
+ lots of other stuff <3

How I feel about this release.

@ZakWestbrook Features like that are coming next. I agree "clarity" and highlights/shadows are the main thing it needs right now to be even better.

Also the app on my OSX just randomly closed the second time I used it when I tried to export the file.

Which app version, which export settings? Explain as much as you can!!!

Yes try the github release page.
Internet explorer has been deprecated for years, so it's a vulnerability and may not support some javascript features that are used to fetch the releases through Github API.
I could be able to fix it if you tried to track down some what causes the problem specifically, but probably not worth the time.

Some improvments coming to playback...



and last but not least, is having only one lens hindering possibility of a 3D perspective?

Ummm no... I did this ages ago already ;D Create a lens filter blue on one side, red on the other - put it on a lens with a big aperture, and you have 3d, red and blue glasses ready video
Only downside is it looks awful.

If you disable Canon's vertical offset corrections (and whatever other position-dependent calibrations they might be doing), and re-apply them from scratch in post, yes. I remember g3gg0 has a proof of concept somewhere on youtube, maybe around 2012.

Unfortunately, the implementation will be camera-dependent (along the lines of crop_rec_4k). You can already play around with adtg_gui to find what registers may need tweaking for that.
Ah right, that seems a bit high level for me ::) And as it's camera-dependent maybe not even worth the time for such a non versatile feature.

Are you thinking at some sort of 4K with reduced temporal resolution at the borders of the frame, and full coverage in center?
Yep. I'm not sure why no one has really talked about it before, isn't it a pretty obvious solution for quite static shots? Or interview shots where only a subject in the middle moves a lot...

But as a first test I'm writing a script for taking panoramic video (one shot after another) along with a script to blend them together in the middle (though it would only look good in shots without action where the seam is).

Pages: [1] 2 3 ... 12