The post many messages above is probably the script (or at least a portion of it). By comparing the data to data entry points for DeShaker you can work out which parameter does what.
You may find it easier to experiment by using VirtualDub with DeShaker in stand alone mode, because then you can do direct entries without going through a script written by somebody else.
The parameters to play with are in pass 2. The first is "edge compression", with default being "none"; I haven't played with this but among options given you can zoom the image to fill to the edge. The next choice is to use "previous frames to fill in borders" where you can nominate the number of previous and future frames to be taken into consideration; the default is 30. Currently I'm cutting material taken from a power boat and I've found best results by not filling the border so I get black outside the steadied frame. You then can choose whether you want soft borders (to the steadied frame) and the number of pixels for the soft edge.
You can deduce that if the camera effect intended is locked down (ie not waving about or tracking) then reconstructing edges from previous and future frames will do very well. If the camera was "smoothly" tracking across a reasonably uniform scene, then again filling the edges isn't too difficult. But something very different in the future frames is likely to start getting integrated into the edge reconstruction, and something very different in the previous frames is likely to leave its presence.
So you might reduce the number of previous and future frames taken into consideration. Where you do get small edges containing distortion you might assume that they will lopped off by the normal TV overscan. Otherwise you need to use crop/pan to enlarge the image (and perhaps reposition it) to get rid of them.

>>>"Deshaker is delayed by 30 seconds"<<<

Yes, that's correct. You must realign the audio with the video, which will be 30 frames out. In stand-alone mode you will get 30 black frames (plus words) at start and end of clips (read John's tutorial). To my knowledge DeShaker will only process avi files.

OK, ive been following this thread for a while now, and i honestly do not see any benefit to running this within vegas,consideirng the vast range of paramaters which can be set within the app (Vdub/Deshaker plug).

Like any other effects app, rendering that portion as huffy, then importing to VDub works for me... and as it processes the 2 steps, i continue editing on Vegas.
Once its done those 2 steps, i then render back out to huffy and reimport the footage to my ongoing non stop edit

Might seem tedious, but look at the big picture..
you have a series of clips with a variety of different motions right. Now VDub only works with ONE of the shaker logs at a time.. BUT if ur frameserving to the plug, your still stuck with one.. now what if the footage on clip number 7 is vastly different to clip number one?

See where im getting at? Even if the script is designed to analyse the clips each time the filter is used for any number of events, that is MORE time in itself which youre wasting during render.

So why not edit while VDub works in the BG?
Its no biggy, Huffy YUV is also superfast to edit with and renders MUCH faster than SOny YUV, u have the power to tweak the filter and the point is to keep ediing while the hardcore deshaking is done

As for the frame offset, if in PAL use 25fps

Once u establish the frame, its easy to realign back into the
timeline..

Yes having a script is great and im not down noting it, but what Vegas needs is this same script to be run through as a"Satish-like" reworked interface for Vegas, which writes up NEW shakerlogs for each clip it passes.
In deshakers case, there are 3 passes. the first to analyse, the second to correct and the 3rd to render
Having a customsed shaker log is paramount to the plugins effectiveness, and i just dont see how NOT having these parameters avaialble to you, let alone a customised shakerlog file would actually work effectively. Yes you could prolly get away with alot with only just one logfile, but to me, that is very limtiing

DJ - After reading your response, I can see what you're getting at.

I think you may be right - just take the clip and render out in Vdub/Deshaker while working in Vegas.

Cliff Etzel
bluprojekt

>>>>As for the frame offset, if in PAL use 25fps<<<

I'm in PAL-land and DeShaker still gives me 30 frames offset. One should emphasize that if you don't want to lose frames from the clip you need to stitch 30 frames to the beginning or end of the clip (depending). Doesn't matter what the added material is, but usually I just copy the first 30 frames and paste it onto the head of the clip. Where the clip has pre-run, not necessary.

EDIT: I agree with DJPadre in that I use my second (i.e. older) machine on the network to do all my deshaking while I'm editing on the main machine. I don't see any advantage in running deshaking within Vegas because I already know if a useful clip needs fixing. This particular thread was initiated with a different need, which was batch deshaking of footage shot from a helicopter. In this the deshaking parameters are constant and it's time-saving to be able to set up processing to run overnight rather than having to initiate processing of each clip (even if in background). After ignoring John Meyer's batch processing script until his 7 days had elapsed, I realised that I could have batch processed material in my current job. But now it's all done.

While having the manual operation of the Deshaker filter is a plus about 80% of the options are useless. I flogged out this filter for days using different settings and what seemed like about 10,000 combinations of Deshaker settings with no appreciable results. The johnmeyer defaults work just fine for me.


JJK

I have tried downloading this but there is a message saying it was created with an unlicensed winzip, why is that?

I've played with Deshaker some time ago and was amazed at the results. Just seemed too tedious for me to run all the programs separately. Having in within Vegas is a great idea.

Does anyone know if this can be used in Vegas Movie Studio (Platinum 8)?

Also, since the the download link expired a while ago, does anyone have it mirrored or be willing to send it?

Thanks!

Does anyone know if this can be used in Vegas Movie Studio (Platinum 8)?

No it cannot. The "Movie Studio" versions of Vegas cannot run scripts. You'd need to upgrade to the full version of Vegas for that.

...yep... link expired...

Guys , I'm trying the deshaker in Vegas Pro 8.1 but when i launch the script it says:

Despite planning for all possible contingencies Deshaker script had the following error:
Error: Value cannot be null.

What am I doing wrong?
Thanks

Hi,

I'm really interested in using the 'deshaker' - anyone have a download link or info where to get all the various components.
Vicmilt??

Ta

http://www.sonycreativesoftware.com/forums/ShowMessage.asp?ForumID=4&MessageID=598847jrazz put a link within this thread . . . [/link]

Grazie

Robbie75:
You have to reset the six log entries in the script. Check the previous threads for this procedure.
JJK

johnmeyer ~ thank you! This is an incredible piece of work and the price is right ;)

Question:
Is there any way to have the deshaker output AVI files compressed? I have a large project with many shaky clips and need to process much video. The output files are quite large (uncompressed AVI).

I did install the Panisonic codec ~ but cannot see it in VirtualDub or figure out how to access it as part of my settings ~ or if it can even be set as the default output codec.

Any help would be greatly appreciated.