DGates: "Geez, I can't believe we're still discussing this. " Why?

Why does it surprise you? Have I missed something? Hmmmmm .. . now what could that be? That I want to understand? That I wish to hear exactly where others are coming from? To have a chance to allow others to explain and expound on their ideas for Vegas, workflow and how it fits within the Market Mix?

I've learnt stax from DSE, Bob, our original poster, and the PROS here? Without this thread I wouldn't have known about what Douglas knows? Now THAT has to be quality info? Surely? It's been great to hear what Z is trying to do to make his biz more comfortable - I find that valuable? Always willing to learn. I've found that sometimes I need to explore to understand. Kinda guy I am.

The other trick that I've used back in my wobble cam days was to simply slow things down to around 90%, I'd completely forgotten this until Victor reminded me of how much he used to like to overcrank film, not that his shots would have had any wobble but it can make the actual motion of things / talent look smooth.

One other thing worth mentioning is that image stabilization is a pretty damn complex business to get it really right. It's probably not too bad at the long end of the lens but a wide shot can pose some interesting issues. The image is no longer a simple 2D affair, the optics have introduced aspherical distortion which shifts as the camera moves, this also needs to be compensated for and that means first camera motion in 3 vectors needs to derived, the image flattened to remove the distortion and then after the image is shifted it needs to be curved back again to match the look. Heaven forbid if you've zoomed in in the middle of the shot!
Doing all this to a HD frame means an awesome number of calcs.

And why are we still talking about this?

I'm with Grazie on this, to learn.

That's alright then. I still have a load of random DV footage that I plan on using eventually. It could come in handy there.

Also, how about the Huffey codec? Will it work with that? The file sizes are slightly smaller than uncompressed.

Actually I haven't used Deshaker or Dynapel at all because of the focus changes for real world camera shake. A slowdown in speed really helps. What is really needed is an interpolater that adds frames from one "in focus" frame to the next "In focus" frame "linearly". The problem is how does the application determine which frame is in focus and which frame is not. Manually, maybe to install markers like chapter markers in DVD-A at all of the in focus frames, a tedious job. Lets see, run it through the "Auto focus" of the Z1 to determine the in focus frames and add markers maybe, then add the interpolated frames to eliminate the camera jerks.
JJK

Sony aquiring Boris, that would be good

Wish they had aquired paint shop pro instead of Corel

Then the portfolio would have been great IMHO :-)

farss

It is complex and with HDV the system overhead would be a lot

If it were a plugin it would have to rely on the host application i.e. Vegas to provide memory management which is far better managed at the operating system level. This can cause the overall application stack to be unreliable and nobody wants that

So what's the answer to suit everyone?

I think there are several

Working outside of Vegas is another bit of management I have to do which I would rather Vegas did in software.

I tend to work by putting everything on the timeline as my storyboard and then working on individual elements from there. It's like using Vegas as the portal to everything I do which suits the way i work.

I would like a better plugin interface for apps that provide video stabilizers and other functionality that requires Vegas to pass all timeline information to the plugin, it's like internal frameserving which makes everything work together

The possibilities are endless for all sorts of functionality and choice, ACID, Cinescore, Smartsound, Dynapal, Boris to namne but a few

The frameserving plugin architecture would be OK for small stabalizer jobs but what about a bigger jobs that might make the application stack unreliable?

I would like Vegas to integrate with other apps better so we could all choose the one that suits us best.

A simple way of doing this would be the way Vegas handles takes, provide options to configure different files formats to have a specific application handle the creation of a new take

I would configure Photoshop for still images, Soundforge for audio and Boris for video

When I right click on the event and edit the take it would open the source media in the application of my choice allowing me to work on that part and the take would automatically update the way Vegas does when it looses focus to other applications i.e. it pulls in the updated media if the filename is the same.

It might be possible to do this with some fancy scripting and a customised toolbar, maybe.

I think we all need to do things with Vegas for different reasons and simple enhancement to the aplication might provide the plugin feel for things such as stabalization and probably a whole lot more

:-)

IMHO

Paul

Can there be such a thing as a two-pass effect? Don't think so...at least not like a deshaking thingie would need.

And why on earth, if the above is even possible, would anyone want to add all that extra time to each and every render?

Do it once, outside, and bring the footage in, deshook.

Sounds like a good idea. Actually, it would be perfect for some time lapse footage I have where I've just removed a few out of focus frames, causing things to slightly speed up.

Hi GenJerDan,

The whole idea of the takes approach is that it would be done outside. The app you had configured for that media type and the Vegas portal would bring all of it together in one place, you do it within the take with the external applications functionality, you render within the external app.

Two pass is a deshaker terminology from what i can see, a plugin architecure that passes timeline information also allows the plugin to access previous and the next frames for comparison which is the same as the two pass approach (almost)

This is more about different approaches to workflow and what we define as a plugin

We all have to render somewhere, it's just a case of where or why not let the individual make the choice :-)

Paul

Returning the deshook footage as a take would really be the only viable option. Bear in mind we could be talking hours of render per minute of footage here, particularly in HD. Add to that the amount of time to setup the deshaker.
Even when I go to Sound Forge to Edit As A Copy there's a considerable overhead while Vegas renders a new copy of the audio file and as I exit SF, it has to render it back and build the waveform files. This isn't too bad as SF and audio isn't a major memory hog.
A video deshaker app would be a very different beast, it'd sure like all the memory it can get and all the CPU cycles as well. Having Vegas itself still lurking in the background mightn't be such a good idea anyway.

Well, the previous and next thing is more like previous and next 30 frames or so. I don't think just one on either side would do the trick. You'd probably end up with tiny shakes all over the place as it tried to stabilize things in the short term rather than over time.

Of course, those frames could be buffered to hell and back to allow for on-the-fly deshaking...but I'd imagine the time penalties would be horrible. And then you do something silly like cut or CC or pan the event and it has to toss out any stored info it had and do it all over again.

But if it's just a "plugin" to access the feature from within Vegas, rather than as a realtime effect...

...I don't think I'd like that, either. Why tie up the NLE when you you can have the deshaking going on over there on something truly external while you continue to work on something else?

I think we are on the same page with different scenarios

A little bit of stabalizing - use a plugin with access to the timeline

A bit more - use a take, a bit of overhead but the workflow benefit outways this

A big shake - do it outside of Vegas, it's the only way

So all it takes is for Vegas to provide timeline information to plugins and to develop the "take" metaphor and everyone can do it which ever way suits them best

I think thats a good approach but i am open to ideas, I keep changing my mind.........

I used to be indeciisive but I think I'm ok now :-)

Paul

Ps. It's not just stabalization that would benefit, it's anything with motion/time as a factor and this has come up before, a coupleof enhancements would create a whole lot of possibilities

I'd settle for an Auto Pan thingie of some sort.

I just finished up stabilizing some softball footage shot by someone who really really really needs to get a tripod for Christmas.

I had to sit there and reposition home plate at the bottom of the cropped frame so it was the same throughout - keyframes every 3 - 5 frames or so. Oy. There's still some movement, but at least it doesn't make you seasick anymore.

Something that could lock an area or point...that'd be wonderful, even if it was only in one direction.

This "question" has matured into a most interesting thread indeed.

It has touched on so much more than the title was promising earlier.

It has gone past the normal "catch-all" of WHY something can't be done and allowed for some neat "out-of-the-box" thinking along the way - and away from how difficult it might be, to a potential "could-be".

Paul, great ideas. I do hope that Sony is listening.
It might be possible to do this with some fancy scripting and a customised toolbar, maybe. - Even I can see this. How about if the Event in question was "laid-off" in its OWN veg to be reintroduced AS a nested TAKE! How about that - a NESTED-Take!?!

I think we all need to do things with Vegas for different reasons and simple enhancement to the application might provide the plugin feel for things such as stabilization and probably a whole lot more - Absolutely.

There's a world of difference between a plugin and sending something off to another app such as SF.

Video plugs in Vegas can be applied at the Event, Track, Bus and Media levels. That imposes some pretty heavy rendering in Vegas even before they get to your chosen plugin if you're coming from the Bus, all tracks have to resolved as well as all composites and then the frames passed to that app / plug.
Doesn't sound like a good idea to me at all, we've got enough users complaining about lack of RT playback as it is.

So OK, we go down the other path, render an event level copy to a new file and open another app to work on it and bring that back as a take. Don't see a problem with this other than it'd be pretty simple to script and it's pretty simple to do manually anyway.
Select clip / region.
Render to new file with Render Looped Region Only.
Open 3rd party app and work on new clip, render out.
Open Vegas and add rendered output from above as take.
Job done in a few mouse clicks, compared to the amount of work you'd be doing in a full blown image stabiliser a totally trivial saving. I've only once tried to stabilise something, a few minutes of footage and I worked on those few minutes for the best part of a day and even then it was a lost cause. In the end it was quicker and easier to keyframe it in Vegas. Problem was the tracking points kept moving out of the frame. Better software is smart enough to acquire new ones but that stuff isn't cheap and you'd better hope the ones it chooses are OK.

As for the nesting idea, sorry don't like it. The amount of time these calcs take can be huge, think minutes per frame. Remember Vegas needs to build a proxy for the nested file, how long you want to wait for that to happen.

I'd suggest downloading a demo copy of Combustion or DF, allocate a few days to work through the sample projects and you'll then get a handle on another world of video,a world where amazing image manipulation can be performed that's very realistic. Thing is it's a very different headspace to traditional editing although the 2D compositing apps are creeping towards what NLE's do.

Aside from all this I can think of a few more pressing enhancements that need to be added to Vegas and with some urgency, image stabilization wouldn't be high on my list of priorities, 10 bit sure is though.

Bob.

Bob, as always, you speak so much plain good sense. I guess I would never come up with these unsatisfactory ideas if I knew what goes on under the bonnet/hood of Vegas? We are truly very lucky to have you about to bring your candid view on things digital.

Thank you!

. .and yes, at present when I want some Steadying, it is exactly that dance that I do.

I suppose as PCs get even faster and cheaper, Sony Vegas is again best placed to adapt to the "better" faster processors. So maybe then a non-engineer like myself, will have their ideas adopted. Nothing like considering the impossible to make the future exciting.

Hi Bob,

I agree with everything you say, perhaps it's just different people have different approaches to similar problems which on the face of it sound the same but a little bit of difference can change the approach to solve the problem

In your scenario i would have used boris red to stabalize either with the motion tracker with keyframes to manage the out fo frame movement or the optical stabalizer

At a few minutes it sounds like i would have done outside of Vegas, if it were a few seconds I would be lazy and want to use red or anohter tool such as steadyhand as a plugin and let it take longer to render whilst I was getting some sleep.

But maybe neither approach would have suited your situation but keyframing it in Vegas for a whole day would make my eyes bleed :-)

Would 10 bit make renders longer anyway, not sure what impact on performance and realtime preview those extra 2 bits would have, sounds like you could do with it. I think i am still ok with 8 bits for now.

Different needs, different approaches? A plugin interface that passes the timeline to the plugin would suit some of the people some of the time for stabalizing and other stuff.

Maybe 10 bit and an improved plugin interface along with developing the take metaphor would be key features for Vegas 8 that would help with a whole load of things for a lot of different users

Maybe further down the Vegas timeline so to speak

Some interesting stuff here though, in summary

Always use a tripod if possible, prevention is better than cure

Jumping out of planes with tripods is not recommended, at 100 mph it's hard to get that bubble in the right place and the tripod creates problems on exit, descent and landing

The earth moves for Bob (lucky man)

sometimes thinking outside of the box and editing your way out of a shake can actually be the best option

Paul

As suggested earlier, use a tripod or learn how to shoot handheld. Shallow breathing, wide angle only, brace your arm into your chest and lean against something solid. Use a third party app to stabilize in post on those "rare" occasions. I find Fusion to be the best. Combustion, AE and RED are also pretty good.

GG

Well all bickering aside, if Sony doesn't want to address these concerns with Sony plugins then they should open everything up so that other companies can do the things Sony doesn't have time to do or is not intrested in developing, End of story. Doesn't seem to be good policy to keep saying no we have more important issues and then not opening code to allow others to address the concerns of at least a good percentage of Vegas Users. I am sometimes hired to edit features together using someone elses footage not as perfect as I would like and have to jump through hoops to make it pretty enough for broadcast. These issues may not reflect the needs of the elite but I would geuss that is less then 10% of the user base. I say it is time to pull back the curtain or at least adress these issues pronto. Until then we are pushed to other software to fulfill our needs. I love Vegas and have since version 2 but come on already.

L.T.

"a good percentage of Vegas Users"

A small percentage of the total users are actually here on the forum.

Yeah and most of these are the professional users. If they are asking for it then you can bet that there are many others who don't even know about this forum. Besides that is no valid excuse. In my opinion.

L.T.

I strongly encourage all of you who are interested in better plugin integration to fill out a Product Suggestion form under the Support drop-down menu on this web site. Although Sony moderators read these forum postings, I suspect they are generally Tech Support individuals. In contrast, suggestions typically go product managers in the Marketing department. They are the ones who usually specify the content of the upcoming releases.