Frankly, I didn't expect the full V2 functionality in Vegas 10 - why would ProDAD enter such a deal? They certainly want us to spend more on the full version.

I also have a question to those using image stabilization (in general - not necessarily Mercalli; I post it here though as this is the hottest thread on stabilization at the moment)...

With pristine resolution of the EX1's 1920x1080 raster, I'd never choose to zoom in in order to achieve the most effective stabilization - just too high a price to pay in terms of resolution loss. Also, using variable (optimum) zoom and filling the border with color is out of question. For me, the only viable method is fixed border; after stabilizing a clip I change the project from 1080p to 720p (the 1080 frame gets squeezed), and then use Track Motion to expand the squeezed video just enough for the border to become invisible. I then render out to 720p. OK, I don't get full HD, but at least I keep the sharpness and resolution in terms of "dpi" (or pixels per inch, should I say)...

I've been wondering why none of the several stabilizers we've discussed on this forum so far has the option of a more precise "fixed border" definition, like e.g. stabilizing into the fixed 1280x720 window for full HD material, or into the fixed 720x576(480) for both 1080 and 720 material. It would guarantee the resultant frames' "dpi" doesn't change when rendering stabilized HD into one of the lower resolution formats (720p, or SD), with no zoom in/out through entire workflow.

Sure it can be done manually, but some fixed border presets would be welcome....

What do others think?

Piotr

"... I'd never choose to zoom in in order to achieve the most effective stabilization - just too high a price to pay in terms of resolution loss."

I understand what you're saying. Doesn't it depend on how shaky the video is as to how far in the stabilizer has to zoom?

How do you deliver most of your videos, e.g., Web, DVDs, or Blu-ray?

Doesn't it depend on how shaky the video is as to how far in the stabilizer has to zoom?

It sure does, Jay - but it usually so happens (well - at least with me:)), that when a clip really needs to be stabilized, it usually requires a lot of zooming in, or quite a large border....

How do you deliver most of your videos, e.g., Web, DVDs, or Blu-ray?

Most of my commercial projects are classical music live performances, and I always edit them in HD all the way through. With the BluRay version waiting for enough market demand to justify printing BDs, for the time being I deliver them in down-rezzed DVD versions.

But this is irrelevant, as with these multi-camera projects we always use tripods of course, so no stabilizing is necessary.

When I do need stabilizing sometimes is with my short forms and other - rather amateurish - ventures, as I sometimes shoot hand-held. If it so happens that a shaky take or two are valuable enough, I'm ready to sacrifice the full HD final format - stabilize them as good as only possible, and render as 720p, or even SD. Whenever I need to make such decisions, I can feel free to apply extra effects like panning, scanning and zooming on other HD takes that do not require stabilization.

But that's just me, of course.

Piotr

Piotr, I'm in total agreement with everything you've said. However, I've not used much in the way of stabilization, I'm just thinking out loud, as it were.

Has anyone determined just how much one can zoom in on a 1920x1080 image before seeing an unacceptable reduction in image quality? The reason I ask is I can imagine there might situations where reduced shake, with a slightly softer image, would be more desirable than an image that looked like it was shot during an earthquake.

Like I said, it's just a thought--deciding between the lesser of two evils.

From the few tests I've done with Mercalli V2 today, I am really quite impressed. I may just forego the Vegas upgrade and buy Mercalli V2.

Hi,

Can anybody decode this statement (from ProDAD newsletter) :

"Mercalli V2 can even stabilize and correct multiple video clips right on the timeline with advanced scene-detection technology." ?

I have been discussing with ProDAD issue, how to use V1 to stabilize long paths with intentionally different kinds of movements (first stage pan, secong stage tilt, thirt stage parallel movement ... in order to achieve smoth transitions beetween stages (it is impossible to produce it if you split event, visible jump is an effect)). We have finished discussion with conclusion, that I have to wait for V2 (V2 will have some new features, connected with scene recognition....) In last newsletter I have found info about scene recognition, but can't understant what does it mean and how to use it. Can anybody help ?

Best regards,