Have tried getting away from "linear" motion?

Just tried both of those. Perhaps what I forgot to mention is that I'm rendering this as 25p. Rendering as 50i it looks as smooth as a babies bottom.
Reason for the 25p render is I don't want any interlace artifacts as this is going up on plasma / LCD displays. I can get it to look just fine rendering as 50p but that doesn't do much good as it's got to be delivered on PAL DVDs.
Adding Linear Blur helps but that still doesn't look quite right.
Other odd thing is this, if I get the graphic to move across the frame in 1 second and render at 50i it looks fine. Move the keyframe so it now takes 2 seconds and render as 25p it still looks jerky. I'd have thought one would offset the other but no such luck.
Bob.

I can get almost acceptable results by first rendering at 50fps (double PAL in Vegas). Then bring that into a new project, apply a few frames of MB and a little Linear Blur with some key frames to control it. Being able to render at say 200fps would probably get better results but looks like Vegas only allows certain frame rates, why is that?
Surely it cannot be that hard to do what a film camera can do!

Just so anyone trying to help understands what I want. Some graphic flies across the frame in a blur and comes to a quick stop. The fast motion should generate a pretty smooth blur that trails off as the object slows down and stops.

Gosh, maybe I should practice what I preach, if you want it to look like film, shoot film. And to think I could have bought an animation system (minus camera) for scrap metal prices some time ago.
Bob.

Maybe try a slight gaussian blur on each frame before the motion blur?

Farss, remember that motion is a directional vector. But when you pan a camera, you not only adding a motion vector but are changing the lens effects and reflected light from the images in the field of view. There are many ways to produce the this effect, though, I don't remember off hand a quick solution for vegas that produces commercial qty end product.

Bob - your wanting a filmic experience so "why not shoot film" was the wake up call for me! You've started to analyze WHAT you want, I think the next approach may GIVE what you want - that was because I started to think along filmic lines and develop the layers, digitally . .ok

2 tracks?

Track 2: The movement of the Text: Motion Blur

Track 1: The "refocusing" or coming clear of the text

This way you've got control over movement and inter movement key-framing - Track Motion here?

. . thinking about it why not 2 events and a controlled Transition envelope to create that refocusing "dissolve"?

I also thought about the render high 50, re-import 25.

Any good? Could be the canine's proverbials?

Grazie

Grazie et al,
motion blur in Vegas simply doesn't cut it, here's why.
Lets assume we're shooting film at 25 fps, just to keep the maths simple.
A film camera has a 180deg shutter, that means the shutter is 1/50th of a second. Now let's assume an object crossed the field of view in that same time, what will the camera record? Well if it's lucky enough, nothing, if it sneaks past while the shutter is closed. However if it starts to cross the field of view just as the shutter opens then we get a streak right across the frame. It's a little bit more complex than even this because the shutter I think doesn't even start to expose the whole frame instantly nor does it block the whole frame all at once so so parts of the frame are exposed slightly more than the rest, so our streak may have comet tails etc. Lets just ignore that issue though.
Now what happens in Vegas using track motion.
Well in the above example if you tried to simulate that you'd never see the object fly past. If in frame 1 it's off the LH side of the frame and in frame 2 it's off the RH side of the frame then what's going to capture it? Bear in mind I'm asssuming we're working in 25p here, in 50i we'll get one field with it in the middle of the frame, hardly a filmic view of the object!
Now Vegas does attempt to address this problem by providing Supersampling. That computes the position of objects moved using track motion at intermediate positions between frames, creates intermediate frames and composites them together. It's a good try but for fast motion hardly adequate. 20 frames of Supersampling gives 20 snapshots across the field of view in the above example, again it sure isn't a streak. Technically you'd need at least 720 slices between the two frames to come close, I'd guess to generate sub pixel rendering increase that by an order of magnitude. Now that's one huge amount of calculations just for two frames, I'd bet the render times for a minute of video with that level of accuracy would be hours.
Now I do have one way to try getting the right result. Run the project at say 1/100th the final speed. Create 100 copies of the track and offser each one by one frame, set the composite level of each to say 1%. Render that out (if it ever does finish!) and then speed the whole thing up 100 times.
Bob.

Bob, around this time one tends to start thinking that you really need a different title dynamic! As an aside, most film cameras use disk shutters so the exposure duration is the same for all areas of the frame, but of course the frame is progressively opened and progressively shut. Some cameras used an oscillating blade shutter, so in that instance there was a variation in exposure across the frame (usually less at the top -- the sky area). Not all had 180 deg shutters -- the Bolex was 165 deg. So now that useless information can be included in your modelling!

Serena