OK am in danger of getting totally confused.
I'm editing HDV footage ( 1440x1080i ) in FCP by importing it as Pro res at the same dimensions, 1440x1080.
Having edited the footage I send it to Qt at the same dimensions 1440x1080 so I can impoprt it into Compressor to make mpeg2 files for Bluray..
In QT properties, QT reports that the 'actual' size of the exported Prores QT file is 1888x1062.
Now I understand that this difference in dimensions is simply because QT is using square pixels to display the actual rectangular pixels of HDV and that QT discards 'redundant' pixels to reduce the dimensions from what they should be, 1920x1080, to 1888x1062.
OK so this could get really confusing but I think I'm happy in understanding it but here's where it gets even more confusing.
I import the 1440x1080 Prores QT file into compressor to make an MPEG Elementary Stream and when I preview this in QT it tells me that the dimensions have changed again to 1440x810 pixels??
What else is going on here?
What's the point of having Full HD at 1080 vertical lines of resolution if the Bluray file is discarding 210 lines of resolution by default. Is compressor discarding these by default. I don't believe I came across an option to reduce my 1080 resolution to 810 whn using compressor. I'll do it all again just to scrutinize it but I'd be interested in some experienced users comments.
Peter

Hi Peter - I am not sure about Blu-ray but was surprised when I came across the same thing for 1024x576 delivery (720x576 anamorphic widescreen PAL) when encoding broadcast Mpegs, you end up with something far less than that in the mpeg stream (720x404 I think), which makes sense in terms of a 16:9 square pixel solution.

You'll notice that in your situation, you are also ending up with the same thing, 1440 wide by 810 high is 16:9 in square pixels.

I imagine that you would need to convert your footage to a non anamorphic square pixel aspect ratio first (so it is actually 1920 x 1080, square pixel) then do your MPEG Compression to end up with a Full HD image. This is one of the complications of the footage not being Full Raster to begin with I imagine.

Non-square pixels are absolutely the DEVIL. I convert my footage to square pixels FIRST THING, so there is none of this garbage about pixel aspect ratios, stretching, display versus capture size, timeline settings, etc.

Make the footage 1920x1080 square pixels, and you're all set to go all the way through your workflow.

Craig, you got it in one. Never dawned on me, that's exactly what's happening. Many thanks to both of you for the response, you need eyes in the back of your head with this stuff.
Peter