trouble with jagged lines

[mito0]

i'm having problems with some of the movies i'm encoding.
most turn out fine, but some of them have jagged lines no matter what i do.
i used to use tfm(order=1,pp=0).tdecimate() along with QTGMC(Preset="Slow", InputType=1) to smooth out jagged lines with great success, but now it doesn't seem to work anymore. i'm hoping someone here has some tricks up their sleeve.
i've uploaded a small clip that has jagged lines on the car door.
can some kind soul help me clear up this mess?

[manono]

i used to use tfm(order=1,pp=0).tdecimate()...

It's already progressive. Why would you try and field match? TDecimate alone is enough to get it back to 23.976fps.

I take it you didn't capture this yourself?

It looks to me like it was blown up from a standard def source and I can't satisfactorily anti-alias it. One solution might be to downsize it, get rid of the jaggies followed by upsizing it again:

BilinearResize(960,540)
TDecimate()
QTGMC(Preset="Slow", InputType=1)
nnedi3_rpow2(rfactor=2,cshift="lanczosresize",fwid th=1920,fheight=1080)

[mito0]

i captured this from turner classic movies' hd channel, which is 1080i.
megui detects the clip as interlaced.
like i said, i've captured movies that looked like this in the past, but the (pp=0) command was enough to clean them up.
i've also used Telecide(guide=0,post=2,vthresh=45) successfully in the past, but lately, it hasn't worked for me.
as for the resized theory, i don't know the technique used by tcm to achieve 1080i, so that's a possibility.

[manono]

i captured this from turner classic movies' hd channel, which is 1080i.

Which says nothing about the movies themselves. Okay, you did capture it yourself. Good.

megui detects the clip as interlaced.

No, it detects that it was encoded as interlaced, not nearly the same thing. Use your eyes. If you don't see any interlacing then it's not interlaced.

...but the (pp=0) command was enough to clean them up.

If the sources are progressive already then you're doing nothing but wasting computer cycles. (pp=0) did nothing.

i've also used Telecide(guide=0,post=2,vthresh=45) successfully in the past, but lately, it hasn't worked for me.

Once again, Telecide isn't needed here, no matter what settings you use. And you can do serious damage to your sources by letting Telecide post-process movies that are already progressive. And movies, by definition, are progressive. I don't know if TCM uses any field-blended crap or broadcasts any hard telecine. Those will be interlaced.

[mito0]

Yes, I see what you're saying now.
Thanks for your help.