Please login or register. June 20, 2019, 03:51:48 PM

Author Topic: Open Source DCP Workflow  (Read 93046 times)

MaSTeRMaMay

  • Guest
Re: Open Source DCP Workflow
« Reply #15 on: December 07, 2011, 12:04:08 PM »
If conforming is done by MXF creation process, is audio pitch also adjusted?

Terrence Meiczinger

  • Administrator
  • Hero Member
  • *****
  • Posts: 561
    • View Profile
Re: Open Source DCP Workflow
« Reply #16 on: December 07, 2011, 02:21:49 PM »
Why would the pitch need to be adjusted? So long as the duration of your media stays the same, then the audio doesn't need to be changed. Whatever you do with the video is independent of the audio... unless you duration is being changed.

luca

  • Jr. Member
  • ***
  • Posts: 56
    • View Profile
Re: Open Source DCP Workflow
« Reply #17 on: December 07, 2011, 09:18:06 PM »
I fear there is a misunderstanding about the usual procedure people in Europe (and as I see also in Russia) use to bring 25 fps material into a DCP 24 fps. The pictures are taken 1 by 1 so at the end you have a longer duration by 25/24 and you have to compensate for this in your audio, otherwise you will be very soon out of sync (you should use a sample rate of 46080Hz, but this is not possible). So MaSTeRMamay is correct.

The misunderstanding very probably arises from the different TV standards in USA and Europe. In the old times with the PAL standard or variations upon it,  the standard way to bring a movie 24 fps into TV was a plain speed-up. That is two interlaced fields per picture. Audio spectral content was also shifted, 25/24 higher. We had no system comparable to 2:3 pull-down until quite recently something similar came up.

@MaSTeRMaMay

In an earlier post you used a normalization to -20 dBFS in sox to "conform" audio for DCP. This is double wrong, I think.
1. You apply the normalization to each channel separately. This will destroy the mix.
2. -20 dBFS is not the maximum allowed dynamic, it is the reference level. You know, this is like having a maximum peak at 0.1 in a range 0-1.
In a correctly calibrated theatre you will hear something, but not too much.
« Last Edit: December 07, 2011, 09:34:58 PM by luca »

Terrence Meiczinger

  • Administrator
  • Hero Member
  • *****
  • Posts: 561
    • View Profile
Re: Open Source DCP Workflow
« Reply #18 on: December 07, 2011, 10:57:03 PM »
I fear there is a misunderstanding about the usual procedure people in Europe (and as I see also in Russia) use to bring 25 fps material into a DCP 24 fps. The pictures are taken 1 by 1 so at the end you have a longer duration by 25/24 and you have to compensate for this in your audio, otherwise you will be very soon out of sync (you should use a sample rate of 46080Hz, but this is not possible). So MaSTeRMamay is correct.

Yes, that's why I said assuming your duration is not changed. I don't know much about video editing, but it would seem strange to simply go from 25fps to a 24fps DCP with 1:1 frames, end up with 4% longer duration, and then adjusting the audio to compensate. I can see doing that for non-linear editing purposes, but for a DCP conversion, it would seem better to recapture the video at 24fps from the 25fps material. I'm sure this is a pretty common issue and there are reasons for doing it that way.

MaSTeRMaMay

  • Guest
Re: Open Source DCP Workflow
« Reply #19 on: December 08, 2011, 12:55:05 AM »
Yup, I'm not that correct, I saw this today, testing... How it would be better to obtain audio normalization using commandline? I'd like to get almost everywhere at

Usually people just go from 25 to 24 fps, because recapturing the video means just recalculating all the frames, and that's not fair at all.

actpower

  • New Member
  • **
  • Posts: 39
    • View Profile
Re: Open Source DCP Workflow
« Reply #20 on: December 08, 2011, 04:41:03 AM »
About the 25 to 24 FPS conversion, I always do the same thing when 25 FPS material arrives here in Spain (that is very usual), that is reinterpret the video or sequence images from 25 to 24, to not add these jerky jumps, and later stretch the audio maintain the pitch. The resulting is a perfect image and audio mix, and nobody can notice the trick.

Discard ona frame from the 25fps to match 24fps is the ugly thing that you can make. And blend the frames cannot help. I hate these jumps from bad conversions from 25 to 24!

Terrence Meiczinger

  • Administrator
  • Hero Member
  • *****
  • Posts: 561
    • View Profile
Re: Open Source DCP Workflow
« Reply #21 on: December 08, 2011, 12:02:28 PM »
Haha, I guess I just can't handle the idea of changing the duration. If I created a 5 minute video, I want it to remain 5 minutes regardless of the frame rate. Having it become 5 minutes and 12 seconds seems so wrong to me. Also, if you add subtitles, wouldn't you need to re-work all those as well?

actpower

  • New Member
  • **
  • Posts: 39
    • View Profile
Re: Open Source DCP Workflow
« Reply #22 on: December 08, 2011, 12:27:49 PM »
Is correct, the lenght is about 4% more. But there is not other option, unless you want jerky playback. In fact, the DVDs edited in PAL countrys are 4% shorter because they do the same thing, speed it these 4% to avoid the jerky jumps...

I hope that the servers update to support 25FPS DCP "soon", in my test, some Doremy servers accept SMPTE and "illegal" Interop 25FPS DCP, but some others not...

Subtitles arenĀ“t a problem because there is tools to stretch it from 25 to 24 :-)


MaSTeRMaMay

  • Guest
Re: Open Source DCP Workflow
« Reply #23 on: December 09, 2011, 04:59:39 PM »
I would like to say, that there is no problem in that =) People usually do this. If you would like to get the whole picture, I would consider simply getting some bluray version (24p) and DVD one (PAL, 25 fps) and simply check their durations. Subtitles could not be the problem, simple script can handle framerate changes.

actpower

  • New Member
  • **
  • Posts: 39
    • View Profile
Re: Open Source DCP Workflow
« Reply #24 on: December 10, 2011, 05:41:29 AM »
http://www.michaeldvd.com.au/articles/PALSpeedUp/PALSpeedUp.asp

I hope to make 25FPS DCP soon. But the truth is that for now is not chance, to the trick is the good path :-D

mrcin

  • Guest
Re: Open Source DCP Workflow
« Reply #25 on: January 29, 2012, 03:29:52 PM »
Hello.

I've installed the kakadu demo from http://www.kakadusoftware.com, and copied kdu_compress to my opendcp directory (as was written here before) but when i'm trying to create one using code:

Code: [Select]
"D:\DCP\programy\OpenDCP1 zainstalowany\bin\opendcp_xml.exe" -r k:/out1/bat1.mxf k:/out1/bat1audio.mxf -r k:/out1/bat2.mxf k:/out1/bat2audio.mxf -r k:/out1/bat3.mxf k:/out1/bat3audio.mxf -d -i DCM -a mototest -t mototest -b mototest -k feature
i only get this: OpenDCP XML 0.0.23 (c) 2010-2011 Terrence...

Any sugestions what am i doing wrong?

Terrence Meiczinger

  • Administrator
  • Hero Member
  • *****
  • Posts: 561
    • View Profile
Re: Open Source DCP Workflow
« Reply #26 on: January 29, 2012, 04:28:11 PM »
Kakadu doesn't have anything to do with the XML part of the DCP process. If they XML files are being created, then all is good.

mrcin

  • Guest
Re: Open Source DCP Workflow
« Reply #27 on: January 29, 2012, 04:53:19 PM »
Well, they don't. Or i don't know where to look for them - where should those files be located after executing command? In directory with mxf files? Do i have to add some speciffic output command at the end of my command line? And finally is my command line correct? ;)



Terrence Meiczinger

  • Administrator
  • Hero Member
  • *****
  • Posts: 561
    • View Profile
Re: Open Source DCP Workflow
« Reply #28 on: January 30, 2012, 07:01:19 PM »
The XML files are created wherever you ran the command from. You should run the command from the directory where the MXF files are located.

mrcin

  • Guest
Re: Open Source DCP Workflow
« Reply #29 on: January 31, 2012, 07:14:22 PM »
Ah, i've moved .bat file to a directory where my mxf is, and it worked. Thank you!

dcinemaforum.com

Re: Open Source DCP Workflow
« Reply #29 on: January 31, 2012, 07:14:22 PM »