Feature Requests

[admin]

It's described in the user manuals, also discussed in threads here.

Bottom line: for determining Avisynth Crop() values, mod 2 is fine. For DGSource() cropping, currently nVidia supports mod 2 only for the right and bottom borders. If you try to select mod 2 for the top and left borders and export that to your script, you will get a warning popup. If you must have mod 2 on the top or left, then use Avisynth Crop().

[admin]

I have another iteration on the cropping design and would appreciate your testing before I slipstream it. The changes are:

* Cropping tool now shows the mod values for the crop values.

* Cropping tool now has an OK button. Exit via OK to retain and use the configured cropping. Exit via the window's X button to discard any changes.

* The 'Export to script' checkbox is eliminated. You now control whether the values get exported by including or not including the __crop__ macro in your template. It's sensible to always include it because it will expand to nothing if cropping is not configured.

* The DGI file now includes a new line called DISP following the SIZ line. This gives the display area, which may differ from the coded size, e.g., when coded height 1088 should be displayed as 1080.

* The DGI file format version is bumped to 22.

* Note: if a stream has a display area smaller than the coded size (e.g., 1080 versus 1088) and cropping is not exported or manually included in DGSource(), then DGSource() will use the display area.

Here are the 64-bit binaries and docs:

http://rationalqm.us/misc/CropRedesign.zip

[Guest]

Did a quick test.
Went item by item on your list and (no surprise) it worked as described

[admin]

Thanks for the testing, gonca!

[sparktank]

DGI file formats and semantics are described in the DGIndexNV user manual. Is there something missing?

Back to this:

I was thinking like a quick index page or something.
A table sheet where these values are quickly mentioned.
Say, if I were to make a PNG of such values for such codecs (avc/hevc).

For the most part, I don't have to look up much references to such things as most content I deal with is progressive.

But in the case of SD content from DVD, I dig around my PC for a cheat sheet with such info before I go gallivanting around with QTGMC.
I know TV productions can have hybrid formats, but I can't trust my own notes, especially when I don't archive them online anywhere.

A cheat sheet would handy, in addition to the full documentation.

[sparktank]

Addendum:

with regards to this:
viewtopic.php?f=8&t=677

COLORIMETRY 2 2 2
in regards to the document (freely available), it's not always easy to remember where to find the document on my PC for the document to narrow down the (possible) values I might encounter.

So something among the lines of a quick reference to these without the document handy would be appreciated.

[admin]

sparktank, I'll see what I can do to make you a happy camper. Thanks for the suggestions.

[Boulder]

I have two feature requests:

1) A resize calculator alongside the cropping tool. Currently I usually downscale UHD and HD sources to 1080p and 720p respectively and I need to open one m2ts file in the Avisynth script creator in MeGUI just to do the maths instead of getting all the necessary things out of one tool. Having the calculator in the indexer would make things much simpler. It wouldn't need to be anything too fancy.

2) Add a possibility to output the HDR related information in the Avisynth/Vapoursynth template to easily copy the information to the x265 command line. Currently I need to check things from the dgi file and translate it to the command line syntax.

Thanks in advance for even considering them

[renols]

Hi.

I have a request for a new command line option.

Would it be possible to get an option that allows subs to be demuxed as well. I can now choose video and audio but not subs.

To be honest I am trying to replace eac3to in my workflow, because I am experiencing heavy slowdowns using eac3to, where it can take 20+ minutes to demux a bluray, whereas dgindexnv can do the same job in like 5 minutes.

If the option is already there please excuse me. But I can't see it in the command line reference in the help file.

I think that by now, dgindexnv can pretty much do all that eac3to can do when it comes to demuxing, and a great program just continues to get better

renols

[admin]

Good point. I'll add this feature. Thanks for the suggestion.

[admin]

Hehe, it already works. Everything is demuxed with -a. I need to update the user manual. Let me know if it is not working for you.

[Guest]

Upgrade to cropping tool
Would it be possible to have an auto setting where the tool scans the video and picks the crop values?
Reason:
Some movies (one of the latter Batman movies comes to mind) are filmed with different cameras and different resolutions, i.e. one can be 1980 x 1040 while the other camera could be 1980 x 800 and the footage will be mixed
Usage:
Tool would sample video at set points (hard-coded or user select-able number of points) and produce crop values based on the "largest" resolution

If it is too much bother don't worry about it
P.S.
How is the swim team doing?

[Guest 2]

Upgrade to cropping tool
Would it be possible to have an auto setting where the tool scans the video

Similar to another topic I opened.

[admin]

I updated the user manual to mention that -a demuxes all audio and subtitle streams and replaced it in the 175 slipstream. Slipstreamed a slipstream, LOL.

On the cropping tool requests I am still thinking it over.

[Guest]

Thank you for the update

P.S.
I want to see it when you update a slipstream to a slipstream that was slipstreamed to update it

[admin]

Hehe. Maybe I'll just get rid of version numbers entirely and identify things with the date. That will drive the crackers crazy.

[hydra3333]

especially if you use american date format

[Guest 2]

especially if you use american date format

I'd rather prefer builds starting with integer PI, adding a decimal for every iteration.

[Guest]

Do you think it would be worthwhile to put x32 and x64 versions into seperate directories
i.e.
DGDecNV>>>x32 Binaries
................x64 Binaries
................other files

[admin]

I'll think it over. Thanks for the suggestion.

[Guest 2]

Perhaps I asked some months ago but it would be nice to have DGSource to output arbitrary resolution, using z.lib or other algorithm in CUDA.

I just sent a old 760 to RMA and they replaced it with a plain 1060, enabling me to decode HEVC in hardware. at least

Problem is the video card has 3 GB only and processing 4k with multithreaded OpenCL AviSynth plugins it's too demanding.

A 1920*1080 fast and accurate output from 4k source would be useful

EDIT: just noticed rw, rh parameters... which method did you implement? Bilinear, Lanczos?

[admin]

I use the NVDec post-processor. I don't know for sure what algorithm NVDec uses. You are welcome to speculate.

[Guest 2]

I have tried to dig into html manual but had no hint.

Is there any 3D support? I mean, I really don't need it.

This last sentence needs explanation: a friend of mine gave me a 3D BD for free, as he doesn't own a 3D television and neither I do.

I'd like to be able to extract the 2D only to watch it and I know almost anything about 3D BD format.

The structure is m2ts and ssif subfolder.

Using eac3to, it finds me the movie mpls and it list left and right eye,

What will happen if I feed DGIndexNV with the mpls? Will it handle it correctly or better to import directly m2ts?

[admin]

I don't know what will happen (probably nothing good). Try it and let us know.

The sanctioned way is to demux the ES and then use DGMVCSource() or FRIMSource().

[Guest 2]

Try it and let us know.

I demuxed left eye with eac3to and fed to DGIndexNV,

Thanks god 3D is disappearing.