Tdeint/eedi2 error - no of frames mismatch

[halsboss]

Hi, any suggestions why the code below yields this error message for my small test clip ? "The number of frames in edeint clip doesn't match that of the input clip!" ... a quick HD 1080i capture from Adelaide Australia channel 10 both before VideoReDo editing and after gives the same result.

# PAL HD 1920x1080i => 576i
MPEG2Source("G:\HDTV\ZZZZ.d2v",cpu=6)
AssumeFPS(25)
AssumeTFF()
LastW=LAST.width()
LastH=LAST.height()
bicubicresize(720,LastH)
# in http://forum.doom9.org/showthread.php?p=748199#post748199
## tcritical says to use for case 4 (TFF, keep top field and interpolate bottom field
##interp = separatefields().selecteven().EEDI2(field=1)
##tdeint(order=1,field=1,edeint=interp)
interp = separatefields().selecteven().EEDI2(field=1)
tdeint(mode=1,order=1,field=1,edeint=interp) # mode=0=same rate output mode=1=double rate output (bobbing) order=0=BFF order=1=TFF
bicubicresize(720,576)
separatefields()
selectevery(4,0,3)
weave()
Converttoyv12(interlaced=true)
SetPlanarLegacyAlignment(True)


echo on
"C:\SOFTWARE\DGindex\DGIndex.exe" -IF=[G:\HDTV\ZZZZ.ts] -OF=[G:\HDTV\ZZZZ] -IA=3 -OM=0 -EXIT
"C:\SOFTWARE\HC\HCenc_022.exe" -ini "G:\HDTV\ZZZZ-HC-CQ2.ini"


*INFILE G:\HDTV\ZZZZ-CQ2.avs
*OUTFILE G:\HDTV\ZZZZ.mpv
*INTERLACED
*TFF
*LOGFILE G:\HDTV\ZZZZ-HC-CQ2.log
*PROFILE BEST
*ASPECT 16:9
*BIAS 30
*BITRATE 9200
*MAXBITRATE 9400
*LASTIFRAME
*AUTOGOP 12
*PULLDOWN
*CLOSEDGOPS
*DC_PREC 10
*CQ_MAXBITRATE 2
Fox2 HQ matrix from http://forum.doom9.org/showthread.php?p=759321#post759321
*MATRIX MPEG



DGIndexProjectFile16
1
G:\HDTV\ZZZZ.ts

Stream_Type=2
MPEG2_Transport_PID=202,2a0,82
Transport_Packet_Size=188
MPEG_Type=2
iDCT_Algorithm=3
YUVRGB_Scale=1
Luminance_Filter=0,0
Clipping=0,0,0,0
Aspect_Ratio=16:9
Picture_Size=1920x1080
Field_Operation=2
Frame_Rate=25000 (25/1)
Location=0,0,0,f4e4

900 1 0 793360 0 0 0 32 32 92 b2 b2 a2 b2 b2 a2 b2 b2 a2
900 1 0 1797656 0 0 0 32 32 92 b2 b2 a2 b2 b2 a2 b2 b2 a2
900 1 0 2823760 0 0 0 32 32 92 b2 b2 a2 b2 b2 a2 b2 b2 a2
900 1 0 3830312 0 0 0 32 32 92 b2 b2 a2 b2 b2 a2 b2 b2 a2
900 1 0 4858108 0 0 0 32 32 92 b2 b2 a2 b2 b2 a2 b2 b2 a2
900 1 0 5874436 0 0 0 32 32 92 b2 b2 a2 b2 b2 a2 b2 b2 a2
900 1 0 6895276 0 0 0 32 32 92 b2 b2 a2 b2 b2 a2 b2 b2 a2
900 1 0 7883404 0 0 0 32 32 92 b2 b2 a2 b2 b2 a2 b2 b2 a2
900 1 0 8748768 0 0 0 32 32 92 b2 b2 a2 b2 b2 a2 b2 b2 a2
900 1 0 9760772 0 0 0 32 32 92 b2 b2 a2 b2 b2 a2 b2 b2 a2
900 1 0 10819400 0 0 0 32 32 92 b2 b2 a2 b2 b2 a2 b2 b2 a2
900 1 0 11821252 0 0 0 32 32 92 b2 b2 a2 b2 b2 a2 b2 b2 a2
900 1 0 12852056 0 0 0 32 32 92 b2 b2 a2 b2 b2 a2 b2 b2 a2
900 1 0 13862180 0 0 0 32 32 92 b2 b2 a2 b2 b2 a2 b2 b2 a2
900 1 0 14896556 0 0 0 32 32 92 b2 b2 a2 b2 b2 a2 b2 b2 a2
900 1 0 15905176 0 0 0 32 32 92 b2 b2 a2 b2 b2 a2 b2 b2 a2
900 1 0 16756816 0 0 0 32 32 92 b2 b2 a2 b2 b2 a2 b2 b2 a2
900 1 0 17771264 0 0 0 32 32 92 b2 b2 a2 b2 b2 a2 b2 b2 a2
900 1 0 18780636 0 0 0 32 32 92 b2 b2 a2 b2 b2 a2 b2 b2 a2
900 1 0 19805612 0 0 0 32 32 92 b2 b2 a2 b2 b2 a2 b2 b2 a2
900 1 0 20827956 0 0 0 32 32 92 b2 b2 a2 b2 b2 a2 b2 b2 a2
900 1 0 21827552 0 0 0 32 32 92 b2 b2 a2 b2 b2 a2 b2 b2 a2
900 1 0 22832976 0 0 0 32 32 92 b2 b2 a2 b2 b2 a2 b2 b2 a2
900 1 0 23853064 0 0 0 32 32 92 b2 b2 a2 b2 b2 a2 b2 b2 a2
900 1 0 24879920 0 0 0 32 32 92 b2 b2 a2 b2 b2 a2 b2 b2 a2
900 1 0 25915424 0 0 0 32 32 92 b2 b2 a2 b2 b2 a2 b2 b2 a2
900 1 0 26928932 0 0 0 32 32 92 b2 b2 a2 b2 b2 a2 b2 b2 a2
900 1 0 27949584 0 0 0 32 32 92 b2 b2 a2 b2 b2 a2 b2 b2 a2
900 1 0 28962904 0 0 0 32 32 92 b2 b2 a2 b2 b2 a2 b2 b2 a2
900 1 0 29965884 0 0 0 32 32 92 b2 b2 a2 b2 b2 a2 b2 b2 a2
900 1 0 30982964 0 0 0 32 32 92 b2 b2 a2 b2 b2 a2 b2 b2 a2
900 1 0 31984628 0 0 0 32 32 92 b2 b2 a2 b2 b2 a2 b2 b2 a2
900 1 0 32986856 0 0 0 32 32 92 b2 b2 a2 b2 b2 a2 b2 b2 a2
900 1 0 34001116 0 0 0 32 32 92 b2 b2 a2 b2 b2 a2 b2 b2 a2
900 1 0 35024400 0 0 0 32 32 92 b2 b2 a2 b2 b2 a2 b2 b2 a2
900 1 0 36042608 0 0 0 32 32 92 b2 b2 a2 b2 b2 a2 b2 b2 a2
900 1 0 37066832 0 0 0 32 32 92 b2 b2 a2 b2 b2 a2 b2 b2 a2
900 1 0 38080716 0 0 0 32 32 92 b2 b2 a2 b2 b2 a2 b2 b2 a2
900 1 0 39100240 0 0 0 32 32 92 b2 b2 a2 b2 b2 a2 b2 b2 a2
900 1 0 40139880 0 0 0 32 32 92 b2 b2 a2 b2 b2 a2 b2 b2 a2
900 1 0 41149252 0 0 0 32 32 92 b2 b2 a2 b2 b2 a2 b2 b2 a2
900 1 0 42171220 0 0 0 32 32 92 b2 b2 a2 b2 b2 a2 b2 b2 a2
900 1 0 43172696 0 0 0 32 32 92 b2 b2 a2 b2 b2 a2 b2 b2 a2
900 1 0 44184888 0 0 0 32 32 92 b2 b2 a2 b2 b2 a2 b2 b2 a2
900 1 0 45204976 0 0 0 32 32 92 b2 b2 a2 b2 b2 a2 b2 b2 a2
900 1 0 46235780 0 0 0 32 32 92 b2 b2 a2 b2 b2 a2 b2 b2 a2
900 1 0 47249476 0 0 0 32 32 92 b2 b2 a2 b2 b2 a2 b2 b2 a2
900 1 0 48270316 0 0 0 32 32 92 b2 b2 a2 b2 b2 a2 b2 b2 a2
900 1 0 49296608 0 0 0 32 32 92 b2 b2 a2 b2 b2 a2 b2 b2 a2
900 1 0 50305980 0 0 0 32 32 92 b2 b2 a2 b2 b2 a2 b2 b2 a2
900 1 0 51311404 0 0 0 32 32 92 b2 b2 a2 b2 b2 a2 b2 b2 a2
900 1 0 52328672 0 0 0 32 32 92 b2 b2 a2 b2 b2 a2 b2 b2 a2
900 1 0 53354776 0 0 0 32 32 92 b2 b2 a2 b2 b2 a2 b2 b2 a2
900 1 0 54358508 0 0 0 32 32 92 b2 b2 a2 b2 b2 a2 b2 b2 a2
900 1 0 55376528 0 0 0 32 32 92 b2 b2 a2 b2 b2 a2 b2 b2 a2
900 1 0 56379320 0 0 0 32 32 92 b2 b2 a2 b2 b2 a2 b2 b2 a2
900 1 0 57390196 0 0 0 32 32 92 b2 b2 a2 b2 b2 a2 b2 b2 a2
900 1 0 58403140 0 0 0 32 32 92 b2 b2 a2 b2 b2 a2 b2 b2 a2
900 1 0 59395404 0 0 0 32 32 92 b2 b2 a2 b2 b2 a2 b2 b2 a2
900 1 0 60426772 0 0 0 32 32 92 b2 b2 a2 b2 b2 a2 b2 b2 a2
900 1 0 61428248 0 0 0 32 32 92 b2 b2 a2 b2 b2 a2 b2 b2 a2
900 1 0 62433672 0 0 0 32 32 92 b2 b2 a2 b2 b2 a2 b2 b2 a2
900 1 0 63450752 0 0 0 32 32 92 b2 b2 a2 b2 b2 a2 b2 b2 a2
900 1 0 64466328 0 0 0 32 32 92 b2 b2 a2 b2 b2 a2 b2 b2 a2
900 1 0 65484160 0 0 0 32 32 92 b2 b2 a2 b2 b2 a2 b2 b2 a2
900 1 0 66505188 0 0 0 32 32 92 b2 b2 a2 b2 b2 a2 b2 b2 a2
900 1 0 67502528 0 0 0 32 32 92 b2 b2 a2 b2 b2 a2 b2 b2 a2
900 1 0 68508328 0 0 0 32 32 92 b2 b2 a2 b2 b2 a2 b2 b2 a2
900 1 0 69499652 0 0 0 32 32 92 b2 b2 a2 b2 b2 a2 b2 b2 a2
900 1 0 70500564 0 0 0 32 32 92 b2 b2 a2 b2 b2 a2 b2 b2 a2
900 1 0 71499408 0 0 0 32 32 92 b2 b2 a2 b2 b2 a2 b2 b2 a2
900 1 0 72522316 0 0 0 32 32 92 b2 b2 a2 b2 b2 a2 b2 b2 a2
900 1 0 73516648 0 0 0 32 32 92 b2 b2 a2 b2 b2 a2 b2 b2 a2
900 1 0 74524892 0 0 0 32 32 92 b2 b2 a2 b2 b2 a2 b2 b2 a2
900 1 0 75535016 0 0 0 32 32 92 b2 b2 a2 b2 b2 a2 b2 b2 a2
900 1 0 76558864 0 0 0 32 32 92 b2 b2 a2 b2 b2 a2 b2 b2 a2
900 1 0 77565040 0 0 0 32 32 92 b2 b2 a2 b2 b2 a2 b2 b2 a2
900 1 0 78589076 0 0 0 32 32 92 b2 b2 a2 b2 b2 a2 b2 b2 a2
900 1 0 79594876 0 0 0 32 32 92 b2 b2 a2 b2 b2 a2 b2 b2 a2
900 1 0 80607632 0 0 0 32 32 92 b2 b2 a2 b2 b2 a2 b2 b2 a2
900 1 0 81607980 0 0 0 32 32 92 b2 b2 a2 b2 b2 a2 b2 b2 a2
900 1 0 82600056 0 0 0 32 32 92 b2 b2 a2 b2 b2 a2 b2 b2 a2
900 1 0 83528212 0 0 0 32 32 92 b2 b2 a2 b2 b2 a2 b2 b2 a2
900 1 0 84473288 0 0 0 32 32 92 b2 b2 a2 b2 b2 a2 b2 b2 a2
900 1 0 85483412 0 0 0 32 32 92 b2 b2 a2 b2 b2 a2 b2 b2 a2
900 1 0 86481128 0 0 0 32 32 92 b2 b2 a2 b2 b2 a2 b2 b2 a2
900 1 0 87489560 0 0 0 32 32 92 b2 b2 a2 b2 b2 a2 b2 b2 a2
900 1 0 88491224 0 0 0 32 32 92 b2 b2 a2 b2 b2 a2 b2 b2 a2
900 1 0 89506800 0 0 0 32 32 92 b2 b2 a2 b2 b2 a2 b2 b2 a2
900 1 0 90505456 0 0 0 32 32 92 b2 b2 a2 b2 b2 a2 b2 b2 a2
900 1 0 91496780 0 0 0 32 32 92 b2 b2 a2 b2 b2 a2 b2 b2 a2
900 1 0 92495060 0 0 0 32 32 92 b2 b2 a2 b2 b2 a2 b2 b2 a2
900 1 0 93485256 0 0 0 32 32 92 b2 b2 a2 b2 b2 a2 b2 b2 a2
900 1 0 94478836 0 0 0 32 32 92 b2 b2 a2 b2 b2 a2 b2 b2 a2
900 1 0 95490652 0 0 0 32 32 92 b2 b2 a2 b2 b2 a2 b2 b2 a2
900 1 0 96484984 0 0 0 32 32 92 b2 b2 a2 b2 b2 a2 b2 b2 a2
900 1 0 97532520 0 0 0 32 32 92 b2 b2 a2 b2 b2 a2 b2 b2 a2
900 1 0 98495080 0 0 0 32 32 92 b2 b2 a2 b2 b2 a2 b2 b2 a2
900 1 0 99507836 0 0 0 32 32 92 b2 b2 a2 b2 b2 a2 b2 b2 a2
900 1 0 100523976 0 0 0 32 32 92 b2 b2 a2 b2 b2 a2 b2 b2 a2
900 1 0 101532784 0 0 0 32 32 92 b2 b2 a2 b2 b2 a2 b2 b2 a2
900 1 0 102533884 0 0 0 32 32 92 b2 b2 a2 b2 b2 a2 b2 b2 a2
900 1 0 103525772 0 0 0 32 32 92 b2 b2 a2 b2 b2 a2 b2 b2 a2
900 1 0 104361808 0 0 0 32 32 92 b2 b2 a2 b2 b2 a2 b2 b2 a2
900 1 0 105380016 0 0 0 32 32 92 b2 b2 a2 b2 b2 a2 b2 b2 a2
900 1 0 106381492 0 0 0 32 32 92 b2 b2 a2 b2 b2 a2 b2 b2 a2
900 1 0 107391428 0 0 0 32 32 92 b2 b2 a2 b2 b2 a2 b2 b2 a2
900 1 0 108404560 0 0 0 32 32 92 b2 b2 a2 b2 b2 a2 b2 b2 a2
900 1 0 109431980 0 0 0 32 32 92 b2 b2 a2 b2 b2 a2 b2 b2 a2
900 1 0 110434960 0 0 0 32 32 92 b2 b2 a2 b2 b2 a2 b2 b2 a2
900 1 0 111442264 0 0 0 32 32 92 b2 b2 a2 b2 b2 a2 b2 b2 a2
900 1 0 112447124 0 0 0 32 32 92 b2 b2 a2 b2 b2 a2 b2 b2 a2
900 1 0 113455932 0 0 0 32 32 92 b2 b2 a2 b2 b2 a2 b2 b2 a2
900 1 0 114460416 0 0 0 32 32 92 b2 b2 a2 b2 b2 a2 b2 b2 a2
900 1 0 115454936 0 0 0 32 32 92 b2 b2 a2 b2 b2 a2 b2 b2 a2
900 1 0 116471828 0 0 0 32 32 92 b2 b2 a2 b2 b2 a2 b2 b2 a2
900 1 0 117480448 0 0 0 32 32 92 b2 b2 a2 b2 b2 a2 b2 b2 a2
900 1 0 118482300 0 0 0 32 32 92 b2 b2 a2 b2 b2 a2 b2 b2 a2
900 1 0 119498628 0 0 0 32 32 92 b2 b2 a2 b2 b2 a2 b2 b2 a2
900 1 0 120508940 0 0 0 32 32 92 b2 b2 a2 b2 b2 a2 b2 b2 a2
900 1 0 121507596 0 0 0 32 32 92 b2 b2 a2 b2 b2 a2 b2 b2 a2
900 1 0 122512268 0 0 0 32 32 92 b2 b2 a2 b2 b2 a2 b2 b2 a2
900 1 0 123532544 0 0 0 32 32 92 b2 b2 a2 b2 b2 a2 b2 b2 a2
900 1 0 124516160 0 0 0 32 32 92 b2 b2 a2 b2 b2 a2 b2 b2 a2
900 1 0 125543580 0 0 0 32 32 92 b2 b2 a2 b2 b2 a2 b2 b2 a2
900 1 0 126581528 0 0 0 32 32 92 b2 b2 a2 b2 b2 a2 b2 b2 a2
900 1 0 127616656 0 0 0 32 32 92 b2 b2 a2 b2 b2 a2 ff

FINISHED 100.00% VIDEO

BTW, which HC matrix do you prefer/recommend for very HQ crisp footage ... let's say just for example, with predominantly fast sports (AFL football) movement and constant panning

[Guest]

You need mode=0 on tdeint().

I've never used eedi2() for double rate, but if you must leave mode=1, then try omitting the SelectEven() on the interp line. Otherwise, ask tritical how to do it for double rate.

[halsboss]

Oh. Thanks neuron2. http://forum.doom9.org/showthread.php?p=1110713#post1110713
indicated "bob it" to deinterlace then denoise and reinterlace and I thought that doubled the framerate in the deinterlacing. I guess I could go back to

TDeint(mode=1,order=1) # mode=0=deinterlace mode=1=doubleframerate, order=0=BFF order=1=TFF

in the middle, however I saw a lot of posts saying how good EEDI2 was and greedily wanted the best outcome :) Oh well.

Anyone have suggestions on a HC matrix ?

[halsboss]

fyi just some results from an AMD 3500+ 1Gb SATA 500Mb ...

(1) resize horizontally, deinterlace, resize vertically, reinterlace = 5.3 fps
# PAL HD 1920x1080i => 576i
MPEG2Source("G:\HDTV\ZZZZ.d2v",cpu=6)
AssumeFPS(25)
AssumeTFF()
LastW=LAST.width()
LastH=LAST.height()
bicubicresize(720,LastH)
tdeint(mode=1,order=1) # mode=0=same rate output mode=1=double rate output (bobbing) order=0=BFF order=1=TFF
bicubicresize(720,576)
separatefields()
selectevery(4,0,3)
weave()
Converttoyv12(interlaced=true)
SetPlanarLegacyAlignment(True)



(2) deinterlace, resize, reinterlace = 3.2 fps
# PAL HD 1920x1080i => 576i
MPEG2Source("G:\HDTV\ZZZZ.d2v",cpu=6)
AssumeFPS(25)
AssumeTFF()
LastW=LAST.width()
LastH=LAST.height()
tdeint(mode=1,order=1) # mode=0=same rate output mode=1=double rate output (bobbing) order=0=BFF order=1=TFF
bicubicresize(720,576)
separatefields()
selectevery(4,0,3)
weave()
Converttoyv12(interlaced=true)
SetPlanarLegacyAlignment(True)


Looks the same to my eye. Not sure if (2) would be better for sports footage with lots of motion though (greater area remains to motion-predict for deinterlacing), will have to source a clip and try ir.

EDIT: oops, partial overlap with http://forum.doom9.org/showthread.php?t=129933 ... now to figure out how to move a post...

[halsboss]

I notice that in http://forum.doom9.org/showthread.php?p=1114448#post1114448 scharfis_brain indicates how to do doubleframerate using yadifmod and nnedi instead. I wonder if this

# PAL HD 1920x1080i => 576i
MPEG2Source("G:\HDTV\ZZZZ.d2v",cpu=6)
AssumeFPS(25)
AssumeTFF()
LastW=LAST.width()
LastH=LAST.height()
bicubicresize(720,LastH)
tdeint(mode=1,order=1) # mode=0=same rate output mode=1=double rate output (bobbing) order=0=BFF order=1=TFF
bicubicresize(720,576)
separatefields()
selectevery(4,0,3)
weave()
Converttoyv12(interlaced=true)
SetPlanarLegacyAlignment(True)


may in theory yield a "cleaner" result of some kind if changed to this ?

# PAL HD 1920x1080i => 576i
MPEG2Source("G:\HDTV\ZZZZ.d2v",cpu=6)
AssumeFPS(25)
AssumeTFF()
LastW=LAST.width()
LastH=LAST.height()
bicubicresize(720,LastH)
# see http://forum.doom9.org/showthread.php?p=1114448#post1114448
# tcritical plugins http://bengal.missouri.edu/~kes25c/
# for yadifmod tff Order=1 ; for nnedi field=-2 means assumeTFF is important
yadifmod(mode=1, order=1, edeint=nnedi(field=-2) #bob deinterlace it without stairstepping to 50
bicubicresize(720,576)
separatefields()
selectevery(4,0,3)
weave()
Converttoyv12(interlaced=true)
SetPlanarLegacyAlignment(True)

[Guest]

What's your theory? You're losing the benefits of nnedi with your method.

[halsboss]

Er, I had a theory (obviously not well thought through) that using yadifmod+nnedi to deinterlace (double framerate) before downsizing to SD would produce crisper/cleaner resulting interlaced clip for sports footage than the TDEINT method...

[Guest]

Oh, sorry I read your post backwards. Your method uses yadif. :)

Try it out. Maybe you're right.

[halsboss]

Well, roughly speaking for my AMD3500+ 1Gb-mem SATA-disk box, (1-pass HC, btw)
yadifmod+nnedi+HC version = about 4 frames per second
tdeint+HC version = about 4 frames per second
ffmpeg version = about 15 frames per second effective, taking into account 2 passes

So for a 2.75 hour sports game, approx encoding times
yadifmod+nnedi+HC version = about 17 hours or 0.7 days
tdeint+HC version = about 17 hours or 0.7 days
ffmpeg version = about 4.6 hours or 0.2 days

Resulting file sizes fior the same 3.5 minute clip :-
yadifmod+nnedi+HC version = 202,230Kb
tdeint+HC version = 202,230Kb (the same !)
ffmpeg version = 194,496Kb
To my eye, the ffmpeg version appeared to be "sharper" than the yadifmod+nnedi+HC and Tdeint+HC versions which both appeared ever so slightly "foggy". No diff to my eye between either of the HC versions.

Note, if you comment out the 1st bicubicresize so that tdeint/yadifmod works on the whole HD frame, it slows from about 4fps to 0.3fps which is about 9.5 days to encode.

At those rates, I'd use
ffmpeg version for just about everything
tdeint+HC version for really really really precious stuff
yadifmod+nnedi+HC version for really really really precious stuff
Unless you wanted to filter the HD clip in some way with avisynth, in which case toss a coin over yadifmod+nnedi+HC and Tdeint+HC.

The ffmpeg script
@echo on
REM 2-pass 1080i to 576i
set fINPUT=G:\HDTV\input HD.mpg
set fOUTPUT=G:\HDTV\input HD.mpg.mpg
set f2passlogfile=G:\HDTV\input HD.mpg.log
set fSIZE=704x576
set fASPECT=16:9
set faveragebitrate=9000k
set fMAXBITRATE=9400k
set fMINBITRATE=3000k
set faudiofreq=48000
set faudiobitrate=192k
Rem Tff=1 Bff=0
set FieldFirst=1
"C:\SOFTWARE\ffmpeg\ffmpeg.exe" -y -i "%fINPUT%" -pass 1 -passlogfile "%f2passlogfile%" -target pal-dvd -b %faveragebitrate% -minrate %fMINBITRATE% -maxrate %fMAXBITRATE% -qmin 1 -qmax 31 -ilme -ildct -flags +ildct+ilme -top %FieldFirst% -dc 10 -s %fSIZE% -aspect %fASPECT% -an "%fOUTPUT%"
"C:\SOFTWARE\ffmpeg\ffmpeg.exe" -y -i "%fINPUT%" -pass 2 -passlogfile "%f2passlogfile%" -target pal-dvd -b %faveragebitrate% -minrate %fMINBITRATE% -maxrate %fMAXBITRATE% -qmin 1 -qmax 31 -ilme -ildct -flags +ildct+ilme -top %FieldFirst% -dc 10 -s %fSIZE% -aspect %fASPECT% -ab %faudiobitrate% -ar %faudiofreq% -ac 2 -acodec ac3 "%fOUTPUT%"
pause
exit

[halsboss]

Then again, the ffmpeg user mailing list confirms what I subsequently found out... there's bitrate spikes in the output file which go outside the DVD spec.

Back to the ever-trusty HC !