Scenarist Error on Import of .m2v File

[bluespot]

Anyone ever seen this error before?

Error D:\!VOBs\AMPie2\ampie2.m2v
Error : Number of fields (38) in GOP(# 2) or between
sequence_header_code should be equal or smaller than 36.

Info 0 file(s) accepted, 1 file(s) rejected

This .m2v file was prepared as follows:
DVD2AVI .d2v project forced film from NTSC source => VFAPI => CCE => pulldown to 29.97 FPS => Scenarist

CCE was encoded as per Doom9's guide.
The is actually the second time that I've had this problem. The first time, I just re-encoded in CCE with all the same settings, except for one. I changed the number of passes from 5 to 3 in the interest of saving time. After I pulled down the new encode, it imported into scenarist without error.

I'll set the latest movie back down from 5 to 3 passes and reencode tonight. If it imports, then I guess it'd be pretty safe to say that the 5 passes in CCE caused the problem, but still, why? And any other ideas in the mean time? I could be totally wrong here...

Thanks all
bluespot

[bluespot]

OK, reencoding with 3 passes had no effect, Scenarist still refuses to accept the pulled down CCE output. Same error. Now I'm at a complete loss :confused: any suggestions???

bluespot

[Joseph2]

I have not seen this error, but I got quite a similar one encoding with TMPEG. I was playing with GOP settings, and accidentally didn't set a SEQ header every GOP. This made my video file unacceptable for Scenarist. When I checked again a SEQ header every GOP the problem dissapeared. I don't know if this will help you.

[bluespot]

Joseph, thanks for the tip! But I already have SEQ Header set for every 1 GOP in CCE, so I don't think that is the problem.

With this latest movie, I just deleted the .vaf and .mpv files produced by CCE, and hit encode again, without changing anything else. This time, the movie imported without complaint into Scenarist. Could this be a bug in CCE? I'm using version 2.62

Also, is there anyone who could explain what this error means? Why does the number of fields in a GOP need to be less than 36? The file that Scenarist finds unacceptable looks perfect out of DVD2AVI, so even better yet, is there any way I could fool scenarist into accepting this videostream?

Thanks again

[auenf]

ok, your GOP size is too big, its 38fields rather than 36 fields.

in CCE reduce the GOP size by 1 P frame, that should do it

now heres why it happens:

with the 3:2 pulldown, you are turning 24fps into 29.97fps, which means an extra 12 fields/second, but you are keeping the original gop size.

so if you specify 12frames per gop max, you are actually getting 18frames per gop after 3:2.

Enf...

[UTec]

It's essential to delete the .vaf file produced by your previous encode if you get this error BEFORE re-encoding with the new GOP settings. Otherwise if you don't, CCE ignores the new GOP (or any other settings) and simply uses the old settings of your current .vaf

I had the same problem and learned this at my own expense... an extra 6 hours of CPU time waiting for the thing to be done.

Oh and thanks auenf. Learned something important here :)

[mpucoder]

um - math is off a little, 3:2 pulldown adds 25%, not 50%. But that is to the displayed number of fields.
The actual limit is 30 encoded fields (PAL) or 36 encoded fields (NTSC) per GOP, pulldown does not enter into this.

The error message may be terse, but it explains it, the number of fields in one GOP should be equal or smaller than 36.

[auenf]

Originally posted by mpucoder
um - math is off a little, 3:2 pulldown adds 25%, not 50%. But that is to the displayed number of fields.
The actual limit is 30 encoded fields (PAL) or 36 encoded fields (NTSC) per GOP, pulldown does not enter into this.

The error message may be terse, but it explains it, the number of fields in one GOP should be equal or smaller than 36.

i was probably half asleep when i wrote that anyway, and im a PAL man anyway, i dont need to know the exact calculations ;)

Enf...

[BassPig]

I have been having this problem with SOME renders from Adobe Premiere CS3. I had to render a 90-minute video for Scenarist authoring, and got the error. However, I don't think it's the settings for the encoder because I could render out a short 2-minute segment with the same encoder settings and Scenarist would accept that. Somehow the size of the video seems to have some bearing.
I'm using a 3/12 GOP setting in all cases (23.976fps video) and only the long 90-min video fails with the "Error : Number of fields (38) in GOP(# 1) or between sequence_header_code should be equal or smaller than 36." error.

I'm wondering if there is some way to predict how Scenarist is going to react with various MPEGs with the same encoder settings?

[n0mag!c]

I'm wondering if there is some way to predict how Scenarist is going to react with various MPEGs with the same encoder settings?
This is not Scenarist issue, this is totally encoder issue. Scenarist just calculates encoded frames while importing. You must wondering why and when your encoder produce 38 fields if you're always using 3/12 GOP setting.
And if you got what you got, I guess the solution is to always use 3/11 GOP setting instead.

P.S. If I got it right 12 is number of P picture in GOP? Or I got it wrong and 12 is GOP length in frames?

[BassPig]

3 is the number of "M frames" and 12 is the number of "N frames" according to the Adobe Media Encoder menu. I have choices of 3,6,9,12,15,18,21 and so on up to 196 GOP lengths. (Why anyone would go past 18 GOP length I have no idea.)
So I could do 3/12 or 3/9 (M/N frames). Why the encoder changes the length of the GOP to an illegal value on long renders is a matter I'm curious about. 'Think reducing the value to 3/9 will fix the long renders?

[n0mag!c]

I have choices of 3,6,9,12,15,18,21 and so on up to 196 GOP lengths.
So I could do 3/12 or 3/9 (M/N frames). Why the encoder changes the length of the GOP to an illegal value on long renders is a matter I'm curious about. 'Think reducing the value to 3/9 will fix the long renders?
It may help or may not help 'cause this bug of encoder may produce again too many frames in GOP as we don't know how exactly this bug is behaving. :)

[BassPig]

I repeated the render, with the 3/12 GOP setting I had before, and this time the resulting file was valid and passed Scenarist's import checking. Something random going on with the renders. Go figure...