Welcome to Doom9's Forum, THE in-place to be for everyone interested in DVD conversion.

Before you start posting please read the forum rules. By posting to this forum you agree to abide by the rules.

 

Go Back   Doom9's Forum > General > DVD2AVI / DGIndex

Reply
 
Thread Tools Search this Thread Display Modes
Old 9th December 2009, 21:59   #1  |  Link
kshawkeye
Registered User
 
kshawkeye's Avatar
 
Join Date: Dec 2009
Posts: 46
NTSC to 23.9, is it worth it?

Hey, I have a bunch of DVDs that I just now got around to Encoding, and I'm wondering if it's worth it to select "Forced Film" on DGIndex to convert the 29.97 NTSC .VOBs into 23.97 to increase the bitrate. I want to go for the best quality, and knocking off all those frames sure does make a difference, but is it really worth it to force down to 23?

Also, is it the easiest way to converting from 29.97 to 23.97 by selected "Forced Film" in DGIndex, or is there some better way in Gordian Knot?

Any help is appreciated! Thanks!
kshawkeye is offline   Reply With Quote
Old 9th December 2009, 23:08   #2  |  Link
osgZach
Registered User
 
Join Date: Feb 2009
Location: Waterloo, WI - USA
Posts: 650
It really depends on the source material. Some DVD's contain hybrid material that plays back at different frame rates. Although this is more common in Anime DVD's, I think I have seen it in some of my live action DVD's too (or surely there are cases where it exists?).

If DGindex on the other hand, is telling you the video is 100% Film, then I think Force Film should work just fine (I think that's what its intended for) as long as you can verify the standard 3:2 pulldown pattern. But if you get something that is a mix of Video/Film then the proper way to encode it while saving space (and preserving the correct frame rate), is to encode it using the 2-pass VFR method. (TIVTC package has examples of this)

Otherwise if you try to force it all to 24fps but some sections contain 30fps material (or vice versa) you will end up with stuttering.
osgZach is offline   Reply With Quote
Old 9th December 2009, 23:42   #3  |  Link
Guest
Guest
 
Join Date: Jan 2002
Posts: 21,923
If you can cleanly go to 23.976 it's desirable because you eliminate pulldown judder.
Guest is offline   Reply With Quote
Old 10th December 2009, 02:10   #4  |  Link
Blue_MiSfit
Derek Prestegard IRL
 
Blue_MiSfit's Avatar
 
Join Date: Nov 2003
Location: Los Angeles
Posts: 5,337
Indeed. It depends on the content. Documentaries and such are commonly hybrid, but hollywood movies are almost always 24p, and can almost always be cleanly recovered via Force Film mode.

If you're unsure how to proceed, post a sample of your source.

~MiSfit
Blue_MiSfit is offline   Reply With Quote
Old 10th December 2009, 04:48   #5  |  Link
kshawkeye
Registered User
 
kshawkeye's Avatar
 
Join Date: Dec 2009
Posts: 46
Hmm, under the "Video Type" it reads that it is NTSC, nothing to do with FILM and a %. Also, what iDCT Algorithm is best?
kshawkeye is offline   Reply With Quote
Old 10th December 2009, 05:11   #6  |  Link
Guest
Guest
 
Join Date: Jan 2002
Posts: 21,923
It may be hard telecined.

Read the manual for the answer to your question about IDCTs.
Guest is offline   Reply With Quote
Old 10th December 2009, 05:21   #7  |  Link
kshawkeye
Registered User
 
kshawkeye's Avatar
 
Join Date: Dec 2009
Posts: 46
So that means what? That is may be hard to telecined? That "Forced Film" is the best way to go?

Quote:

Specify which iDCT algorithm will be used by DGIndex and DGDecode.
+ 32-bit MMX (1 in INI file, D2V file and CLI)
+ 32-bit SSE MMX (2)
+ 32-bit SSE2 MMX (3)
+ 64-bit Floating Point (4)
+ IEEE-1180 Reference (5)
+ Skal SSE MMX (6)
+ Simple MMX (7)

Algorithms that are not supported by your processor are automatically deleted from the menu, so you may not see all of the above options.

The iDCT algorithm that you should use depends primarily on what CPU you have, and to a lesser degree on the desired iDCT accuracy. All of the available options are IEEE-1180 compliant. For more information on iDCTs please see the DGDecode manual.

Quality: IEEE-1180 Reference > 64-bit Floating Point > Remaining iDCTs.

Speed: SSE2MMX is the fastest. IEEE-1180 Reference is the slowest.

Note that the selected iDCT algorithm is placed in the D2V file. DGDecode can be configured to use this setting (default) or it can be overridden via the "idct" parameter.

This setting is also stored in the DGIndex.ini file and therefore is saved across DGIndex sessions. If you edit the DGIndex.ini file, or the D2V file, to specify an iDCT algorithm that is not supported by your processor, then DGIndex and DGDecode will demote your setting to the next available setting that is supported by your processor.
I found that before I posted my question about iDCTs, yet I couldn't make a decision about what mode to select, so there for I asked here for some help from the community on the standard.
kshawkeye is offline   Reply With Quote
Old 10th December 2009, 07:05   #8  |  Link
~Revolution~
Registered User
 
Join Date: Sep 2009
Posts: 89
Whats the difference between forced film and tfm().tdecimate() in the avs? maybe I'm missing something
~Revolution~ is offline   Reply With Quote
Old 10th December 2009, 09:57   #9  |  Link
burfadel
Registered User
 
Join Date: Aug 2006
Posts: 2,235
Most things are filmed in 23.976, and are telecined to 29.97 to match the TV standard. The TV standard of 29.97 for NTSC was originally to match the power frequency of the US, which is 60 Hz (2 fields of 29.97, aka interlaced = 59.94Hz). PAL is mostly 50Hz for the same reason, the power systems there are 50Hz! (there is a 60Hz PAL for countries that have a 60Hz power system).

For countries that have 50Hz PAL, the Film speed of 23.976 is sped-up by around 4 percent, which is not that perceptible when you watch it. Thats why the PAL versions of movies are shorter than NTSC versions, a 100 minute movie in NTSC is actually only 96 minutes in PAL (even though its the exact same content). For NTSC, 29.97/23.976 = 1.25 (it sounds wrong, but put it in a calculator you'll find its exact). So, what they do is have 5 frames in place of having 4 frames. The extra frame is made by mixing the top and bottom fields from two adjoining frames, and this is done twice not once which probably would make more sense in a way. The 4 frames are still present in the 5 frame group, but 2 frames are now mixed, so you have 3 original frames and 2 mixed frames (telecined). THis is why the term 3:2 pulldown is used on devices that remove the telecining via hardware. The downside to telecining is the mixed frames have fields which are from a different moment in time, as well as a new false time period created by having a new frame, which is why 29.97 shudders in movement which is particularly noticeable in horizontal pans.

Deinterlacing 29.97 without reoganising the mixed frames for compression is bad, as once it is deinterlaced it is much harder to get back to the original 23.976. Also, encoding at 29.97 is less efficient as you are not only encoding the same material but spread out over 5 not 4 frames, but the detail in 2 of those frames is actually different to the other 3 as you have mixed fields from different frames, which may have different moments of motion due to the different time of the original capture of the frames that are now mixed.

Taking all that in to consideration, inverse telecine (IVTC), aka 'Forced film' is definitely a worthwhile exercise on 29.97 material that was originally 23.976. I believe there is very little material that is actually filmed in 29.97. I'm not sure whether 'Forced Film' option in DGindex is a more suitable option than using an avisynth IVTC filter, maybe someone can clarify that. Setting 'Forced Film' in Dgindex though has the downside that it should only be selected for telecined material, which you just have to remember you selected that. Using an avisynth IVTC filter you can at least see it present (usually) and remember that it is set!

Last edited by burfadel; 10th December 2009 at 16:57. Reason: Originally misintentionally wrote 23.976/23.97 = 1.25
burfadel is offline   Reply With Quote
Old 10th December 2009, 13:14   #10  |  Link
kshawkeye
Registered User
 
kshawkeye's Avatar
 
Join Date: Dec 2009
Posts: 46
Wow, great help! You really know your stuff! I understand all of it except for:

Quote:
Deinterlacing 29.97 without reoganising the mixed frames for compression is bad, as once it is deinterlaced it is much harder to get back to the original 23.976.
So what should my aproach be? Is selecting "Forced Film" the right way to get the 29.97 back into 23.97?
kshawkeye is offline   Reply With Quote
Old 10th December 2009, 14:26   #11  |  Link
Guest
Guest
 
Join Date: Jan 2002
Posts: 21,923
Quote:
Originally Posted by kshawkeye View Post
So what should my aproach be? Is selecting "Forced Film" the right way to get the 29.97 back into 23.97?
Only if DGIndex reports it as Film 95% or greater, i.e., it is soft telecined.

If it is hard telecined, then use an IVTC in your script.

If it is not telecined, then you should leave it at 29.97.
Guest is offline   Reply With Quote
Old 10th December 2009, 14:28   #12  |  Link
kshawkeye
Registered User
 
kshawkeye's Avatar
 
Join Date: Dec 2009
Posts: 46
Okay, I'm not 100% sure how I can tell what it is (I get the whole FILM 95% but I still would like to be sure so I don't encode 200 TV episodes and end up being wrong), so whats the best way to save a 10 second clip? If I encode it how can you tell what the actual VOB is?

Last edited by kshawkeye; 10th December 2009 at 14:40. Reason: I sounded like a noob :p
kshawkeye is offline   Reply With Quote
Old 10th December 2009, 14:49   #13  |  Link
Guest
Guest
 
Join Date: Jan 2002
Posts: 21,923
Load the VOB in DGIndex. Set the range you want with the [ and ] buttons. Then do Save Project and Demux Video. Upload the resulting M2V file to mediafire.com and post the link.

Give us 50MB from a high motion scene. Pick a scene that you think is playing jerky (based on wht you said in your other thread).
Guest is offline   Reply With Quote
Old 10th December 2009, 15:08   #14  |  Link
kshawkeye
Registered User
 
kshawkeye's Avatar
 
Join Date: Dec 2009
Posts: 46
http://www.mediafire.com/?a4yjoxwhwd3 Should I use IVTC or "Forced Film" based on that 10 second preview?
__________________
"Earth is only part of life" -Kyle Schwarz

Last edited by kshawkeye; 10th December 2009 at 15:12.
kshawkeye is offline   Reply With Quote
Old 10th December 2009, 15:43   #15  |  Link
Guest
Guest
 
Join Date: Jan 2002
Posts: 21,923
It's NTSC which is 0% Film, so no Forced Film here.

Stepping through the frames, it is obviously 3:2 hard telecined, so use an IVTC in your script, such as:

telecide()
decimate()

or:

tfm()
tdecimate()
Guest is offline   Reply With Quote
Old 10th December 2009, 15:55   #16  |  Link
kshawkeye
Registered User
 
kshawkeye's Avatar
 
Join Date: Dec 2009
Posts: 46
It seems to auto add

Quote:
Telecide(order=1,guide=1).Decimate()
for me, will that work just fine or does it need to be

Quote:
telecide()
decimate()
__________________
"Earth is only part of life" -Kyle Schwarz
kshawkeye is offline   Reply With Quote
Old 10th December 2009, 17:06   #17  |  Link
Guest
Guest
 
Join Date: Jan 2002
Posts: 21,923
What do you mean by "it" when you say "It seems to auto add them"?

Read the users manual to see what those parameters are and then you will be able to know what you need to do.
Guest is offline   Reply With Quote
Old 10th December 2009, 17:08   #18  |  Link
kshawkeye
Registered User
 
kshawkeye's Avatar
 
Join Date: Dec 2009
Posts: 46
sorry about that, Gordian Knot
__________________
"Earth is only part of life" -Kyle Schwarz
kshawkeye is offline   Reply With Quote
Old 10th December 2009, 19:35   #19  |  Link
Guest
Guest
 
Join Date: Jan 2002
Posts: 21,923
If you don't understand what's going on then just take what GK gives you.
Guest is offline   Reply With Quote
Reply

Thread Tools Search this Thread
Search this Thread:

Advanced Search
Display Modes

Posting Rules
You may not post new threads
You may not post replies
You may not post attachments
You may not edit your posts

BB code is On
Smilies are On
[IMG] code is On
HTML code is Off

Forum Jump


All times are GMT +1. The time now is 03:36.


Powered by vBulletin® Version 3.8.11
Copyright ©2000 - 2018, vBulletin Solutions Inc.