Sleestack said:
That's a great explanation.
How would I test this with my sources on my two 1080p panels?
To my mind, your best bet would be to try and find the HQV benchmark DVD. HQV is a chip manufacturer that produces extremely good scalars, and also provides a test DVD which contains a battery of tests from noise reduction and jaggie destruction to cadence detection.
I believe Audioholics uses this test for many of their TV reviews (see bottom of
this page for instance). I think however, this DVD costs a ridiculous amount, although I have not personally confirmed this. I'm sure Clint or one of the core staff can tell you more about it.
I should warn you though, there are those who doubt the validity of this test, because HQV does manufacture it's own chips afterall, and one has to wonder if this test is truely impartial or is it geared more towards showcasing their own chip. Having said that, this test is widely respected in the industry.
A cheaper solution might be to use the AVIA test disc, but again, I have no first hand experience with this product either
Personally, if you think you're already getting an incredible picture, don't go hunting for trouble! I am sure with the kind of high-end gear you have, your PQ must be off the charts!
Wouldn't 2:3 be a ratio, mathematically, if it was written 2/3, i.e. 0.666...?
In Engineering, 2:3 represents a scale (though not one in common use). Isn't this what's meant in video parlance; a scale factor?
When viewed that way, you're absolutely right. In fact, to be honest it doesn't matter whether you call it 2-3 or 3-2. A 3-2 pull down is actually a 2-3 pull down shifted by 1 frame. It seems to only be a matter of reference.
But to understand that you need to understand a bit about the 2-3 pull down process. Just to recap, the POINT of 2-3 pull down is to convert 24fps film to ~30 FPS NTSC video.
The first step of the 2-3 pull down is to slow the 24fps film down to 23.blah FPS.
23.blah/~30 = 4/5.
So ad 23.blah fps, for every 4 frames of film, you will have 5 NTSC vid frames @ ~30 fps. So basically 2-3 is adding an extra frame. How does it do this? It takes advanatage of interlacing!
So each frame of video actually contains two fields. 1 field for the odd scan lines, and the other field for the evens. 2-3 doubles up some of the FRAMES across FIELDS to achieve this. So let's look at the following 4 FRAMES in film:
ABCD
We split these frames into FIELDS but using the 2-3 cadence:
AABBBCCDDD
The BC boundary is actually ONE FRAME and so is the CD boundary! That's how we end up with 5 frames (i.e: AA, BB, BC, CD, DD).
So this brings us back to my original point. If you start your frame of reference at frame A, then it's 2-3. If you start it at B it's 3-2.
Clear as mud? Great.
That's exactly how clear it will be if your TV's 2-3 does not work properly. U'll end up with ghost images from the previous frame at the BC and CD boundary iif the damn thing screws up.
Hopefully my limited understand of this hasn't served to confuse you even more!