MLV App 1.14 - All in one MLV Video Post Processing App [Windows, Mac and Linux]

Started by ilia3101, July 08, 2017, 10:19:19 PM

Previous topic - Next topic

0 Members and 2 Guests are viewing this topic.

masc

Quote from: ultimatemale on May 01, 2018, 06:26:15 PM

https://drive.google.com/drive/folders/1K8DNFBCgiYBCglxdZe_ZsAA1QFhHTWU_?usp=sharing

Thx, I'll have a look. @bouncyball: could you also check what happens? File error checking was implemented by you ;) ... you're deeper in this than I am.
5D3.113 | EOSM.202

Danne

Yes full mlv path in a list is ok.
Check my edited answer above. Similar idea  :P

ultimatemale

Quote from: masc on May 01, 2018, 06:55:33 PM
Thx, I'll have a look. @bouncyball: could you also check what happens? File error checking was implemented by you ;) ... you're deeper in this than I am.
Cheers dude, awaiting your response

masc

Quote from: Danne on May 01, 2018, 06:58:03 PM
Yes full mlv path in a list is ok.
Check my edited answer above. Similar idea  :P
Hej Danne, check out latest revision. I changed:
- temp files are in /tmp/mlvapp_path/... now
- temp file names are changed to "better" names
- one temp file includes a list of all original mlv filenames+pathes
5D3.113 | EOSM.202

masc

Quote from: ultimatemale on May 01, 2018, 07:23:08 PM
Cheers dude, awaiting your response
I debugged with your file now: MLVApp is right, when it tells that it is corrupted. It includes a MLV block named "È\0"... that is not defined in the mlv definition. In past we ignored such blocks, but this was not always safe, if I remember right. I tried ignoring via a small code change and MLVApp did not crash and I could open the file. Maybe you should try another nightly build, which does not produce such errors (whyever it does, I am a noob here).
@bouncyball: is it possible to safely ignore such corrupted blocks in files, in a way we can be sure MLVApp does not crash?
--> video_mlv.c line 1450 and following...
5D3.113 | EOSM.202

Danne

Hell yeah :)
app_path.txt
file_names.txt
output_folder.txt
tif_creation


I need to rewrite some parts of my script now to fit with this refinements but I really like it better.

Now a question. Is it possible to erase and add path+names from the file_names.txt file while arranging in the session window? I see myself deleting and adding files back and forth in session window but not until I hit Export and Save button will I be sure what path+names will be used in the end. This is no biggie if it doesn´t work since I will use these paths to check for matching features so if the file is still in file_names.txt it will be ignored.

Thanks again for an open mind and helping coding hands :),

masc

@Danne: each time you import a masxml we check, if the files exist. If not, this mlv will not be loaded in MLVApp, and you won't get this filename into file_names.txt when exporting. So the only way to crash the app is moving the file while it is loaded into MLVApp. Until now we said, we don't want to control this... :P
You also only get the filenames chosen for export (must not be all files in the session).
5D3.113 | EOSM.202

ultimatemale

Quote from: masc on May 01, 2018, 08:44:56 PM
I debugged with your file now: MLVApp is right, when it tells that it is corrupted. It includes a MLV block named "È\0"... that is not defined in the mlv definition. In past we ignored such blocks, but this was not always safe, if I remember right. I tried ignoring via a small code change and MLVApp did not crash and I could open the file. Maybe you should try another nightly build, which does not produce such errors (whyever it does, I am a noob here).
@bouncyball: is it possible to safely ignore such corrupted blocks in files, in a way we can be sure MLVApp does not crash?
--> video_mlv.c line 1450 and following...

Thanks a lot dude

In your test would you say the new 4k Experimental builds are quit stable for recording 1080p RAW? Bcus if it is then i might as well just stick with it instead of still using the old usual Nightly build which only does 1080p.

masc

Quote from: ultimatemale on May 01, 2018, 09:18:59 PM
Thanks a lot dude

In your test would you say the new 4k Experimental builds are quit stable for recording 1080p RAW? Bcus if it is then i might as well just stick with it instead of still using the old usual Nightly build which only does 1080p.
Sorry, I can't tell you how safe it might be. But if I hear "experimental", I wouldn't think first it is "stable". You might have luck and it works brilliant (what often is the case), but maybe it doesn't ;)
5D3.113 | EOSM.202

ultimatemale

Quote from: masc on May 01, 2018, 09:33:25 PM
Sorry, I can't tell you how safe it might be. But if I hear "experimental", I wouldn't think first it is "stable". You might have luck and it works brilliant (what often is the case), but maybe it doesn't ;)

Thanks for the reply dude

Danne

@masc
Great :).
I'll test and implement. Talk soon again. Let's see if mr B will linux up my script as well ;)

Danne

Thanks masc for path routines. Works perfectly!

Latest download:
https://bitbucket.org/Dannephoto/mlv-app-danne/downloads/MLV%20App.app.zip

It would be nice if the master branch could update my HDR_MOV-command script to the latest version here like I did in my fork:
https://github.com/dannephoto/MLV-App/blob/master/platform/qt/bash_scripts/HDR_MOV.command

source
https://github.com/dannephoto/MLV-App

masc

5D3.113 | EOSM.202


Danne

@masc
Hi again. Been building another script derived from this:
https://www.magiclantern.fm/forum/index.php?topic=21089.msg193622#msg193622

Here is the script in question:

https://bitbucket.org/Dannephoto/mlv-app-danne/downloads/tif_clean.command

I can drag this script into session window and it installs fine into MacOS folder. Problem is when I select the newly installed script tif_clean.command it will instead execute HDR_MOV.command. Seems to be default. Tried restart and also removing HDR_MOV.command but not helping here unfortunately...

jpegmasterjesse

Very low priority issue here, BUT

I tried out QTv0.8alpha on Crouton XFCE4 on my Chromebook Acer R11.  The machine isn't cut out for this kind of work but I wanted to see if I could use it for MLV processing in a pinch.

Crouton on Chromebooks is already hacky, so I didn't have high hopes, but MLV App worked as expected except on export - it only gets to 50% and instead of putting out a movie it puts out PNG frames.

Again, I doubt anyone is seriously depending on a Chromebook for this kind of work but I thought I'd report it as Linux container support will (fingers crossed) soon work natively within ChromeOS.

bouncyball

Quote from: jpegmasterjesse on May 06, 2018, 06:39:19 AM
I tried out QTv0.8alpha on Crouton XFCE4 on my Chromebook Acer R11.  The machine isn't cut out for this kind of work but I wanted to see if I could use it for MLV processing in a pinch.
Why do you use v0.8 (released 227 days ago) when v0.15 (latest) is out? :) Link

Edit: Ahh.. yes because of the only linux release MLV.App.v0.8.alpha.x64.Linux.AppImage ;)

If on Crouton XFCE4 can be installed Linux version of QT 5.9 you could compile latest Git version on your chromebook.
   

Danne

@masc
Looks like this is executed every time even if another script is chosen:
In Scripting.cpp
    //enabling HDR processing, no questions asked. Yet.
    QProcess process;
    process.startDetached("/bin/bash", QStringList()<< "HDR_MOV.command");

Could "HDR_MOV.command" be turned into a variable?

Regards this issue:
https://www.magiclantern.fm/forum/index.php?topic=20025.msg201009#msg201009

jpegmasterjesse

Quote from: bouncyball on May 07, 2018, 02:50:20 PM
Why do you use v0.8 (released 227 days ago) when v0.15 (latest) is out? :) Link

Edit: Ahh.. yes because of the only linux release MLV.App.v0.8.alpha.x64.Linux.AppImage ;)

If on Crouton XFCE4 can be installed Linux version of QT 5.9 you could compile latest Git version on your chromebook.

I'm such a noob - any chance you could point me to a tutorial that could get me started?  I've never compiled nothing :D

bouncyball

Try this guide kindly written and translated to English by @escho

Danne

I should build a compiler tool for Mlv App for mac.

Danne

Tried out the white balance branch to see what could be done with color

Opened up a MLV file out of the box gives following(canon eos 100D)


Edit: I notice camera matrices for eos 100D isn´t added in camera_matrices.c? Code not used?


Official release


White balance branch


WB set to 2000(Official release)


WB set to 2000(white balance branch)


50mm1200s

The WB branch seems stuck, right? Ilia even commited the Andy600 suggestion, but backed off...

bouncyball

This is not the matrix issue. White balance branch more or less correctly calculates WB for 4500K-6500K range. For 5500K it's perfect :). Other temps are quite off. As I said this branch is highly experimental.

bouncyball