How to select 29.97 framerate

[Henderson]

Whenever I rip a DVD with AGK it always turns out 23.976 fps, even though the DVD is 29.97 fps.

What setting am I missing to force it to retain the original fps of the DVD?

Thanks.

[CWR03]

AutoGK is likely doing its job properly. If you're converting a theatrical movie, the film was originally 23.976 fps. If it were to retain the DVD's 29.97 you would see some blended frames on playback.

[Henderson]

It's a DVD I recorded myself, NTSC system in the US at 29.97 fps. The recorder is a Lite-On DD-A100GX, if that gives any further info.

[CWR03]

After loading a file, hold Ctrl and press F9 for the "hidden menu." There is a selection for 29.97 fps.

[manono]

If you recorded it from a TV set, and if the program being recorded was shot on film, then 23.976fps is the correct framerate. AutoGK is usually right about these things and you had better know what you're doing before messing around in the Hidden Options.

If choosing 29.97fps incorrectly where the real framerate is 23.976fps, the result won't be blended frames but duplicate frames, resulting in jerky playback most obvious during certain kinds of movement. That's because of the deinterlacer used by AutoGK.

Post the log for one of the AutoGK runs that you think should have resulted in 29.97fps. Doesn't it play smoothly at 23.976fps? If it was really 29.97fps and AutoGK gave you the wrong framerate, it would be almost unwatchable because it would play very jerky.

[Henderson]

Yes, I'm recording off the TV. I usually discard the log files when done, but I did a short clip just to get the log file for it and it's below.

Under the hidden options, it seems that 23.976 always stays checked, unless I also check "Reduce FPS for 60/50fps HDTV progressive sources", which I didn't think I should be checking for any reason.

If I do check "Reduce FPS..." and select 29.97, then it will keep the 29.97 fps. Otherwise the default of 23.976 fps is the result. Whether it's 29.97 or 23.976 though, the results are never jerky at all, they both play smooth, in sync, and can't really see any visual difference at all.

I guess it's not that big of a deal, but was just curious why the fps is changed from the original, thinking keeping it the same as the original recording would give best results.

Log from a short file (original file 29.97 fps, output ended up 23.976) :

[9/29/2009 10:41:14 PM] AutoGK 2.55
[9/29/2009 10:41:14 PM] OS: WinXP (5.1.2600).2
[9/29/2009 10:41:14 PM] Job started.
[9/29/2009 10:41:14 PM] Input file: C:\Hauppauge\Recordings\_Cbl-4_20090929_175017.mpg
[9/29/2009 10:41:14 PM] Output file: C:\Hauppauge\Recordings\_Cbl-4_20090929_175017-unchecked.avi
[9/29/2009 10:41:14 PM] Output codec: XviD
[9/29/2009 10:41:14 PM] Audio 1: Audio Stream 0 MPEG
[9/29/2009 10:41:14 PM] Subtitles: none
[9/29/2009 10:41:14 PM] Format: AVI
[9/29/2009 10:41:14 PM] Target size: 50Mb
[9/29/2009 10:41:14 PM] Audio 1 settings: Auto
[9/29/2009 10:41:14 PM] Started encoding.
[9/29/2009 10:41:14 PM] Demuxing and indexing.
[9/29/2009 10:41:15 PM] Processing file: C:\Hauppauge\Recordings\_Cbl-4_20090929_175017.mpg
[9/29/2009 10:41:15 PM] Source resolution: 720x480
[9/29/2009 10:41:15 PM] Found NTSC source.
[9/29/2009 10:41:15 PM] Source aspect ratio: 4:3
[9/29/2009 10:41:15 PM] Analyzing source.
[9/29/2009 10:41:23 PM] Source has percentage of interlacing in motion areas: 23.73
[9/29/2009 10:41:23 PM] Source has percentage of telecined patterns: 60.45
[9/29/2009 10:41:23 PM] Source has percentage of progressive patterns: 39.55
[9/29/2009 10:41:23 PM] Source has percentage of interlaced patterns: 0.00
[9/29/2009 10:41:23 PM] Source is considered to be hybrid (mostly FILM).
[9/29/2009 10:41:23 PM] Looking for optimal hybrid thresholds.
[9/29/2009 10:41:27 PM] Source seems to be pure FILM.
[9/29/2009 10:41:27 PM] Output will contain 987 frames
[9/29/2009 10:41:27 PM] Encoding audio.
[9/29/2009 10:41:30 PM] Using VAQ in XviD
[9/29/2009 10:41:30 PM] Audio1 size: 537,216 bytes (0.51 Mb)
[9/29/2009 10:41:31 PM] Overhead: 39,424 bytes (0.04 Mb)
[9/29/2009 10:41:31 PM] Video size: 51,852,160 bytes (49.45 Mb)
[9/29/2009 10:41:31 PM] Running compressibility test.
[9/29/2009 10:41:31 PM] Writing the following script to C:\Hauppauge\Recordings\agk_tmp\_Cbl-4_20090929_175017-unchecked_comptest.avs
===========================================================
LoadPlugin("C:\Graphics\AutoGK\DGMPGDec\DGDecode.dll")
LoadPlugin("C:\Graphics\AutoGK\filters\autocrop.dll")
LoadPlugin("C:\Graphics\AutoGK\filters\decomb.dll")
LoadPlugin("C:\Graphics\AutoGK\filters\RemoveGrainSSE3.dll")

movie = mpeg2source("C:\Hauppauge\Recordings\agk_tmp\_Cbl-4_20090929_175017-unchecked.d2v")
cropclip = autocrop(movie,mode=0,wmultof=4,hmultof=4,samples=10,aspect=0,threshold=34,samplestartframe=0,leftadd=0,rightadd=0,topadd=0,bottomadd=0)
fixed_aspect = 0.888888888888889
c_width = width(cropclip)
c_height = round(height(cropclip) / fixed_aspect)
input_par = float(c_width)/float(c_height)
input_par = (input_par > 1.4) || (input_par < 1.25) ? input_par : (4.0/3.0)
out_width = 640
out_height = round(float(out_width) / input_par)
hmod = out_height - (floor(out_height / 16 ) * 16)
out_height = (hmod > 4) ? (out_height + (16 - hmod)) : (out_height - hmod)
new_aspect = (float(out_width) / float(out_height)) / fixed_aspect
Telecide(movie,guide=1,post=2).Decimate(mode=0)
autocrop(mode=0,wmultof=4,hmultof=4,samples=10,aspect=new_aspect,threshold=34,samplestartframe=0,leftadd=0,rightadd=0,topadd=0,bottomadd=0)
LanczosResize(out_width,out_height)
RemoveGrain(mode=2)
SelectRangeEvery(15,15)
===========================================================
[9/29/2009 10:41:53 PM] Duration was: 22 seconds
[9/29/2009 10:41:53 PM] Speed was: 44.14 fps.
[9/29/2009 10:41:53 PM] Compressibility percentage is: 614.17
[9/29/2009 10:41:53 PM] Using sharper matrix
[9/29/2009 10:41:53 PM] Switching b-frames off
[9/29/2009 10:41:53 PM] Chosen resolution is: 720x512 ( AR: 1.41 )
[9/29/2009 10:41:53 PM] Predicted comptest value is: 268.34%
[9/29/2009 10:41:53 PM] Running first pass.
[9/29/2009 10:41:53 PM] Writing the following script to C:\Hauppauge\Recordings\agk_tmp\_Cbl-4_20090929_175017-unchecked_movie.avs
===========================================================
LoadPlugin("C:\Graphics\AutoGK\DGMPGDec\DGDecode.dll")
LoadPlugin("C:\Graphics\AutoGK\filters\autocrop.dll")
LoadPlugin("C:\Graphics\AutoGK\filters\decomb.dll")
LoadPlugin("C:\Graphics\AutoGK\filters\RemoveGrainSSE3.dll")

movie = mpeg2source("C:\Hauppauge\Recordings\agk_tmp\_Cbl-4_20090929_175017-unchecked.d2v")
cropclip = autocrop(movie,mode=0,wmultof=4,hmultof=4,samples=10,aspect=0,threshold=34,samplestartframe=0,leftadd=0,rightadd=0,topadd=0,bottomadd=0)
fixed_aspect = 0.888888888888889
c_width = width(cropclip)
c_height = round(height(cropclip) / fixed_aspect)
input_par = float(c_width)/float(c_height)
input_par = (input_par > 1.4) || (input_par < 1.25) ? input_par : (4.0/3.0)
out_width = 720
out_height = round(float(out_width) / input_par)
hmod = out_height - (floor(out_height / 16 ) * 16)
out_height = (hmod > 4) ? (out_height + (16 - hmod)) : (out_height - hmod)
new_aspect = (float(out_width) / float(out_height)) / fixed_aspect
Telecide(movie,guide=1,post=2).Decimate(mode=0)
autocrop(mode=0,wmultof=4,hmultof=4,samples=10,aspect=new_aspect,threshold=34,samplestartframe=0,leftadd=0,rightadd=0,topadd=0,bottomadd=0)
LanczosResize(out_width,out_height)
RemoveGrain(mode=2)
===========================================================
[9/29/2009 10:42:13 PM] Duration was: 19 seconds
[9/29/2009 10:42:13 PM] Speed was: 49.82 fps.
[9/29/2009 10:42:13 PM] Expected quality of first pass size: 211.93%
[9/29/2009 10:42:13 PM] Trying to adjust settings.
[9/29/2009 10:42:13 PM] No adjustment is possible
[9/29/2009 10:42:13 PM] Warning: final AVI will likely be undersized.
[9/29/2009 10:42:13 PM] Running second pass.
[9/29/2009 10:42:41 PM] Duration was: 27 seconds
[9/29/2009 10:42:41 PM] Speed was: 35.39 fps.
[9/29/2009 10:42:41 PM] Job finished. Total time: 1 minute, 27 seconds

[manono]

[9/29/2009 10:41:27 PM] Source seems to be pure FILM.
It's supposed to be 23.976fps. To prove that your 29.97fps encode is at the wrong framerate, open it in VDubMod and scroll to a section with movement. Then start advancing a frame at a time. You should notice that every 5th frame is a duplicate. At 23.976fps every frame is unique.
...unless I also check "Reduce FPS for 60/50fps HDTV progressive sources", which I didn't think I should be checking for any reason.
Since you've now said you're capping Hi-Def videos, if capping from one of the networks that broadcasts at 720p, then they're broadcasting progressive 59.94fps and you'll want to check that box.
...but was just curious why the fps is changed from the original, thinking keeping it the same as the original recording would give best results.
Film, including the stuff you see in the movie theaters as well as most major TV shows these days, is shot at 24fps. However, all NTSC is 59.94, fields per second for DVD, standard def broadcast, and 1080i broadcast, or frames per second if broadcast as 720p. Since you apparently have a 29.97fps 1080i cap here, 3:2 pulldown (and a slight slowdown) is applied to convert 24fps to 29.97fps. This results in duplicate fields being broadcast. AutoGK, through an IVTC, removes those dupe fields to restore the video to its original and smoother playing 23.976fps.

[Henderson]

It's supposed to be 23.976fps. To prove that your 29.97fps encode is at the wrong framerate, open it in VDubMod and scroll to a section with movement. Then start advancing a frame at a time. You should notice that every 5th frame is a duplicate. At 23.976fps every frame is unique.
I did open up the original file recorded off the TV which GSpot and MediaInfo says is 29.97fps in VirtualDubMod and there are no duplicate frames at all. So since every file is unique that means it's really 23.976, right? When I open up the file that I forced to encode at 29.97, there's no duplicate frames there either. Not sure what that tells me though hehe.

Since you've now said you're capping Hi-Def videos, if capping from one of the networks that broadcasts at 720p, then they're broadcasting progressive 59.94fps and you'll want to check that box.
I only checked that option once to force encoding at 29.97fps for further testing, otherwise I can't change the fps. But I'm not receiving HD at all, don't even have an HD TV yet.

I can't complain about the results though in any case. Whether I let it choose and it turns out as 23.976, or force it to 29.97, the resulting video plays smooth, not jerky at all, and in sync... can't ask for anything better than that.

Thanks much for the explanations.

[manono]

Oh yeah, you're right - not Hi-Def. I apologize.

When I said to check the 29.97fps one for duplicate frames, I didn't mean the source capture, but the AVI that AutoGK made when you forced the 29.97fps framerate in the Hidden Options.

But you said that checking the 29.97fps capture also shows every frame to be unique. In every 5 frame sequence do you see 3 progressive 'clean' frames and 2 interlaced frames? If so, that's your proof that the 'real' framerate is 23.976fps. If you like, you can cut a small 10 second piece of the source, a piece with movement for upload to MediaFire so we can have a look.

[Henderson]

Yes, when I check the forced 29.97 file frame by frame, 3 are clear and then 2 are interlaced, throughout the file.

Amazing that AutoGK can compensate so well and even when I force it to a wrong framerate, the video is still smooth when played.

[CWR03]

It may look somewhat smooth with the blended frames, but if you're watching it on an LCD it may not be noticeable on playback because with the response time all frames are a little blended anyway.

As I said to begin with, AutoGK will usually give you a proper framerate file, otherwise it would be called "Force29.97GK."