Simon Ubsdell
6
COW Coins
Forum Replies Created
-
Looks like they’re really stepping it up a gear with this one – I defy anyone to call it anything other than “pro”.
Simon Ubsdell
http://www.tokyo-uk.com -
[julienne rathore] “how do I track so that the cloned area follows the movement of the camera?”
You would need to match move the bezier shape to something close to its source – you shouldn’t need to adjust the offset layer if you’ve picked the right bit.
Simon Ubsdell
http://www.tokyo-uk.com -
Looking at your source image, I’d say that you definitely won’t get a good result from your blurring technique but by cloning patches of the ground over the cable and tracking those to any movement you should be able to make it disappear quite convincingly.
The fact that the ground is so messy and broken up is very much on your side.
Don’t try and do the whole cable run with one single patch however as that is going to be almost impossible – treat it in manageable sections, even though this seems like it’s more time-consuming.
Here’s how to do it in Motion.
Duplicate the source layer and apply a bezier mask to the new layer to reveal the layer underneath. (Maybe add a temporary quick obvious colour effect to the lower layer to see what you are doing – e.g. make it blue or something.) Then adjust the position of the lower layer until you are showing through to something that is not cable – you only need to move it a bit so you get something that looks like it belongs there.
Turn the resulting two layer into a group, duplicate the original layer (unmodified) and put it above the new group – carry on as above to make each new patch. In other words, apply a new mask to the top layer and adjust the group underneath which now contains the result of your previous fix.
Hope that makes sense.
Simon Ubsdell
http://www.tokyo-uk.com -
This sounds as though it could be a very problematic shot to fix thpugh without seeing your surce material it is hard to be sure.
A suggestion would be to try “cloning” a clean bit of floor rather than using your blurring idea. Keep your bezier mask and your tracking but try offsetting the position of the masked layer so it gives you the nearest possible bit of clean floor without cable. Not tsyaing it will work, but it’s worht a try and will be less noticebale.
I’d say if your motion tracking is working then you’re much better off with that than trying to keyframe it. Keyframing, even if you’re really good at it, will almost always result in noticeably irregular behaviour which is instantly distracting.
Simon Ubsdell
http://www.tokyo-uk.com -
[Chris Harlan] “The only thing I would add is that we all tend to see Corporate behavior as highly premeditated, when in fact a number of warring impulses can all be firing at the same time and in different directions.”
Agreed. I think what we’ve been seeing here is more than anything a glimpse of Apple in “headless chicken” mode rather than executing some co-ordinated evil plan.
Somehow, the “evil plan” theory feels a bit less alarming than the lack of one – at least there would be a plan of some kind …
Simon Ubsdell
http://www.tokyo-uk.com -
[Chris Harlan] “His observations strike me as shallow”
I wasn’t actually suggesting that his was the most in-depth knowledge that could be had. I just think his comments are a relevant pointer to the to me obvious fact that FCPX began to take shape around the time of iMovie 08 – and hence the plausible conclusion that iMovie 08 was the “proof of concept” for FCPX.
So many people still gloss over this, mostly through complete ignorance of iMovie and it history, but also because it doesn’t fit the notion that FCPX is something uniquely special (one way or another, either for good or ill).
At some point, someone at Apple (let’s for the sake of argument call him “Steve”) said: let’s not waste our time rebuilding the old Final Cut, let’s instead build a new “pro app” out of this really cool consumer app that we’ve built to replace the old iMovie.
Yes, there are complications to that story, not least the inevitable demise of QuickTime and the rise of AV Foundation, but I find it hard to believe that any other story is remotely as plausible.
One of the other key indicators is that iMovie 08 (onwards) was way, way too sophisticated in fundamental conception for its target market at the time.
Simon Ubsdell
http://www.tokyo-uk.com -
[David Lawrence] “Totally believable that Ubillos came back in sometime in 2010 with an order to simplify and build on his iMovie work. To me, this scenario makes the most sense and explains the baffling nature of the release. This implies the fork was sometime in 2010.”
It does seem as though this answers some of the questions but there is one piece of evidence that doesn’t fit:
https://sachin.posterous.com/why-apple-built-final-cut-pro-x
Sachin Agarwal claims here that he:
“worked on Final Cut Pro from 2002 to 2008. It was an amazing experience. The Final Cut Pro X project was just getting started when I left Apple.”
In other words it’s been in gestation since around 2007 …
And that timing I think supports the hypothesis that iMovie 08 (and beyond) was always meant to be the “proof of concept” for FCPX.
At least that’s my version.
Simon Ubsdell
http://www.tokyo-uk.com -
Thanks for the link – that’s actually the one I referenced in my original post. Doesn’t seem that they are aware of the particular (anbd very significant) issues that I raised however.
Unless they are rather coyly admitting that you can’t make transform changes within Photoshop that will show up correctly in FCPX.
Frankly, if your requirements for Photoshop are a basic as Apple are seeming to imply then you’re probably better off compositing your graphics in FCPX anyway.
The ability to handle transforms is hardly high end stuff!!!!!!!
Simon Ubsdell
http://www.tokyo-uk.com -
I hope xtoPro – unlike xto7 – will deal correctly with:
a) multitrack sources (come through as 4 times as many tracks in xto7 irrespective of how many tracks were used in the edit);
b) detached audio – xto7 has had and probably still does have issues with losing track of this;
c) zero start timecode sources (i.e non-timecde-stamped) – xto7 still has residual issues with this.
The last thing we need is another FCPX related product that has been rushed to market only semi-functional – just like FCPX itself.
Simon Ubsdell
http://www.tokyo-uk.com -
I’d also suggest that you [ublish the “roundness” parameter of your mask shape as it’s got sharp corners. It might be better to replace it with an ellipse rather than a bezier shape.
Simon Ubsdell
http://www.tokyo-uk.com