"Vegas makes very good 50p/60p from 50i/60i when you set deinterlace to interpolate."

Sure. But what he reports here is that it seems to be not possible to pair two 1080 50i streams within a 1080 50p project setting - and allign the streams accoding to the 50p in the project setting. I wonder if that is true really - but I have not tested it.

I have not tested that because I use two TD10 also when I shoot with two cameras. The two TD10 have lance, and have been selected from 4 TD10 units with the aim to have two cameras that stay in sync for at least 30-45 minutes (the so called pairing actions of cameras). And they are started with a lanc controller. So, from that side you have no requirement to use a 50p project. Sure one could shoot in 1080 50p and pair these streams - but the performance requirements are higher then compared if one uses two 1080 50i streams (and my new system can handle that for sure, but it was weak on my older system).

As said, to maintain the s3D effect a high degree of synchronisation is required. If your cameras have not been paired in the beginning, then it will be hard to know how well they will act. And it is important to understand that this is something that cannot be fixed in the post, since we are talking here about the synchron readout from the sensors of both cams - after starting synchron.

For slow moving targets it will not harm your s3D effect for sure. For things like a waterfall - well I do not know how your cameras act. Could be that for some takes it is fine, but for some takes it is an issue. Depends how synchron the units start to record.

So sorry, without testing it I do not know if there is a solution in Vegas. Because to my opinion it should have been possible to import 50i footage in a 50p project and pair the two events in the timeline to s3D.

[I]I still think that it is a bug. If you put 50i clips into a 50p project, then Vegas should not be concerned with fields, except to convert them to full frames.[/I]

I think Vegas has to be concern which field is which. I am not sure of the inner workings of a video file but knowing whether a frame is upper or lower must make a difference otherwise we wouldn't have an option to chose the field order in the properties. I think it especially a role when converting interlaced to progressive as you should combine the "correct" fields into one frame for a lower progressive frame rate and interpolate the missing part from the correct fields.

If you take a multicam interlaced project and the clips are not aligned according to their field, you could have two upper or lower frames at a cut and the subsequent clip would be played in reverse field order. I think Vegas treats a s3D track much like a multicam track and aligns the clips with corresponding fields.

Well, it's all just my opinion but does make sense to me.

A field is half the resolution of a frame. In order to create "full frames" it has to interpolate the two fields. And field order matters. Fields are 1/60th (ntsc) of a second different in time.