Strange pulldown pattern

[Alex-Kid]

Two nights ago I captured an old movie (from a HD signal) that seemed to be taken from an also old telecined source. I has pulldown applied but I couldn't find the appropriate pattern.

I come here asking for help on this, so I leave a TS sample file for analysis.

http://www.mediafire.com/?l5eqfovue55q26p

Saludos

By ALEX-KID

[Guest]

A talking head is almost useless as a sample. Please post a sample with a lot of movement, pans, scene changes, etc. Also, post the original TS capture, so we can see if there is any soft pulldown. And why did you post a re-encoded sample? Just cut the original TS with DGSplit.

[Alex-Kid]

I'm sorry, my mistake (don't know what I was thinking :confused:)

Now I uploaded a TS sample.

[Guest]

That's much better. :)

It's not pulldown. It's field blending, presumably from a PAL->NTSC conversion. You can try unblending with srestore(). I'm not that experienced with unblending, so hopefully others will chime in.

Oh, I'm curious about your capture unit. Is it a Hauppauge HD-PVR 1212 or Colossus? I'm thinking the latter, because the audio is AAC. I just installed a Colossus and find it way more stable than the USB unit.

[Alex-Kid]

Thanks for you reply, neuron. I'll try when I get some time.

About my capture unit, it's a Maintronics STB ISDB-T1801 (http://articulo.mercadolibre.cl/MLC-35359357-receptor-de-television-digital-terrestre-isdb-t-norma-chile-_JM). I don't know what kind of chip uses, but it's compatible with ISDB-Tb, MPEG-2 and H.264 video with AAC and AC3 audio.

[Sharc]

You may try

tdeint(mode=1) #bob deinterlace
srestore()
tdecimate() #remove duplicate frames
assumefps(25)