Motion Control?

[Jim Arthurs]

Hi all... pretty off-topic except that this would be MUCH better achieved on the RED, as opposed to the detail-starved HVX...

How'd I do this effect?

http://ftp.datausa.com/imageshoppe/outgoing/PAN_TILT/TAKES_BLENDED.mov

Background: this is a PROOF OF CONCEPT test for a client that wants a series of shots with dozens of clones of the same person in frame, all with a freely panning/tilting camera, able to follow the actors at will.

Don't judge the quick roto or any of those issues, actual shoot will be on a set with green screen and not outside with changing light and clouds moving across frame... just notice the fact that there are multiple takes combined into one.

Again, ideas on how this was achieved?

Thanks for looking...

[Curran Giddens]

Curt's Mirus MoCo head can do this. I think he posted a demo video similar to this using a pickup truck driving around some sand dunes.

[Jim Arthurs]

You bet it can! Wonderful product that... and I can't wait to see it at and meet Curt at NAB...

But how did I do it without Curt's pan/tilt rig? :)

[Curran Giddens]

I guess with lots of skill you could manually repeat the same camera movement.

[shaftbond]

not only do i want to know how you did that, i want to know what happens right after the clip ends when the two clones collide... rip in the space-time continuum? :)

[HD Hildebrand]

Great job. Do you have a breakdown of your gear and process?

[Jim Arthurs]

I guess with lots of skill you could manually repeat the same camera movement.
Interesting possibility... but notice that the timing is PERFECT over six takes in addition to the master take. Can't do that by hand....

not only do i want to know how you did that, i want to know what happens right after the clip ends when the two clones collide... rip in the space-time continuum?
Another clue... even after shooting I could alter the sync so they won't collide... in fact, that last clone is the second clone of the shot, from the SAME TAKE...

[Richard Andrewski]

Beautiful. I give up. Tell us...

[Tom Lowe]

Damn, that is awesome!

[Dan Blanchett]

I'm no expert on VFX. But maybe you had the camera on something that rotates at a constant speed (like a turntable or whatever). Then you shot a little wider for all six takes, and then created the tilt movement in post by zooming in a bit and repositioning. You could also fake the apparent varying of pan speed in post too.

Looks pretty cool however you did it.

[Jim Arthurs]

Okay, next hint... only hardware was the HVX, and my crappy old Bogen 501. I even accidentally kicked a leg a bit after the first take, but kept on going in the hope that it wouldn't show up.

Thinkbug, that idea is good and workable, but not what I did. But I will say that all takes were done at the same FOV.

Anyone want to see all the raw takes? Might be some clues there...

[Tom Lowe]

hey don't be coy jim, just tell us!

[Jim Arthurs]

Nah, it's more fun this way... at least for me. :)

I'm really fairly proud of this technique, as it breaks some barriers to these sorts of shots.

Here are the raw takes... the center one is the "master" take, which I shot first. All the others have completely different camera movement, even panning backwards to the direction of the master...

http://ftp.datausa.com/imageshoppe/outgoing/PAN_TILT/SOURCE_TAKES.mov

[Tom Lowe]

hahha.. sometime i feel the same way about my timelapse stuff. it took me so much damn time and treasure to figure out how to do it, im reluctant to tell people!

[Michael Schrengohst]

Did you make image sequences out of each take and stitch them together using After Effects 7 Pro tracking and then a clean up in Photoshop?

[Dan Blanchett]

Pretty damn cool, Jim.

[Jim Arthurs]

Well if you didn't do the pan by hand (by your own admission) for each of six takes, and only used sticks and a camera, then I give up.
I did do each take by hand panning and tilting the head... with no attempt to "follow" or repeat the master take. Though the process would work even better if I did. I purposefully tried to break it by doing far more extreme things than the final will call for... like the 360 degree master pan shot...

Did you make image sequences out of each take and stitch them together using After Effects 7 Pro tracking and then a clean up in Photoshop?
I did do image sequences, which I pretty much do with any shot going through multiple software, but used AE only to combine and roto, not track.

I call the technique "Pan/Tilt Re-assignment". Sexy name, eh? I want a few more people weighing in before the reveal...

[Nick Shaw]

I'm guessing you did something like stabilising the pan and tilt out of each take, to effectively place each take onto the inside of a sphere around the axis of the tripod head. Then so long as the man never breaks the edge of frame, he can be rotoed out and appear in the master take as it passes that part of the pan. That's how I'd do it.

However it's achieved, it's a nice effect.

[Jim Arthurs]

Nick, we have a winner!

1.) undistort the footage to image sequences.
2.) track each take in 3D tracking software for nodal/tripod
3.) in 3D program, camera front project each sequence onto the inside of a sphere with the 3D camera playing out the tracking motion... this effectively removes all the camera motion.
4.) repeat for each take...
5.) Have a master camera play out the master takes motion to re-record each projected take.
6.) comp together.

This is VERY flexible and allows you to slip sync like crazy, something you couldn't do with motion control (well, without using this technique on top of the mocontrol, that is). You can even replace the original camera motion with a different path... filling in missing areas of the image with a quick 360 degree swivel take of the environment to pull from...

You can violate frame with secondary takes and simply reframe tighter to fix the errors... with RED 4K this will work a treat and still give plenty rez for 2K finals...

[Dan Blanchett]

Brilliant.

Could this technique be as effective if the clones interacted with each other (i.e. fighting)? I guess the choreography would need to be well planned out.

[CVB]

Jim, You are a man of many talents. Mind if I ask how long it took to roto and comp that shot? Maybe we should cancel our NAB booth ;)

[Jaime Vallés]

Very impressive, Jim! It sort of makes the whole "multiple Agent Smiths" effect in Matrix Reloaded seem easy.

[Jim Arthurs]

Curt, no way... too many things MOCO can do for real... can't do time-lapse with this process, or precise repeats with tight framing... this is just an option for some situations... I put half a day into the roto, but it would need several days to trim up and be right. My technique vrs MOCO on this setup would require the same clean up though, and have the same issue of getting the takes done between clouds and shifting sun... Really look forward to meeting you in person at NAB!

ThinkBug... looks like we'll get our cameras the same week... also, sure, you could do a fight with this process... and many other things like the "guy about to get hit by the car who jumps out of the way in the nick of time" sort of stuff...

[CVB]

Yeah, moco definitely has its place. It makes things like that shot much faster.. no tracking involved but theres still plenty of roto if you aren't using a green screen. Have you seen this Kylie Minogue video (http://www.youtube.com/watch?v=kaiw3Qqx1Rk)? They used a milo for the shoot but even knowing how moco works it can still be confusing. Basically they did one shot where the crane loops back and the end of the shot plays right back into the beginning. They use the ladders and a light pole for convenient cut points. Very impressive use of moco though.

[Jim Arthurs]

Yes, that's a great piece... soft spot in my heart for Kylie, bless her... anyway, thanks to everyone for checking it out. I really look forward to seeing you all at NAB!

I hope to have time to push this technique a bit more, just have a hard time getting willing on-camera subjects... the dogs run away from me, and the wife vanishes the moment she sees the tripod come out...

[Seth Larney]

Hi Jim,

Very impressive, I love the concept behind this technique. I own a production company in Sydney Aus, and we are always looking at ways to do high production value shots with innovative techniques. Well done.

Seth.
________
R32 (http://www.cyclechaos.com/wiki/BMW_R32)

[Chris Gearhart]

Jim, your work always impresses. Good innovation indeed.

[Michael Schrengohst]

Your "actor" will start wanting royalities. Does he get paid 5 times?
I remember you saying SAG wanted their fees for the virtual skydivers!

[Jeff Kilgroe]

Jim, that's a neat little trick there. I'll have to jot that one down for later. ;)

[shaftbond]

love it man. what 3d software did you use?

[Jim Arthurs]

That's Lightwave...

I'll post a God's POV movie of one of the virtual cameras "painting" a take onto the world sphere... pretty freaky looking seeing the background locked in place and the FOV of that camera basically a "window" that moves across the sphere revealing more background as it follows the talent around...

[Jim Arthurs]

Here's that example I promised... helps show off what's exactly happening...

http://ftp.datausa.com/imageshoppe/outgoing/PAN_TILT/LW_DEMO.mov

No reason a compositing app with good 3D tools couldn't handle this part, I just used what I know best to prototype it out...

Regards,

[shaftbond]

very creative! thanks for posting all the "in process" shots. you are a scholar and a gentlemen

[Mike Harrington]

Very clever.......
did you roto out shadows as well?

what did you use for tracking

[Jim Arthurs]

I just used very basic roto shapes for the talent and his shadow. I was caught between doing enough roto that people wouldn't be distracted by it and miss the "really cool" thing of mixing dissimilar takes, yet I didn't want to finish it out because it was not a test of roto ability or talent. It COULD be made perfect, with two or three man days of work, but I'll save that dedication to the actual job.

SynthEyes was used for the tracking on each take, in "tripod" mode. In my work in general, I use tripod tracking instead of 2D compositing app tracking 75% of the time because you have much more precise results when a hundred or so trackers are averaged out compared to 1 or 2 track points...

[Alex Boothby]

Fantastic Jim. I've played around with a much less precise version of this technique on Inferno, but in that ap you loose a ton of res when projection mapping onto a sphere. Plus the res caps out pretty quick and texture errors occur. I think you're onto something pretty cool here. I was once going to do an entire music video with various performances projected onto a sphere, with a 4 minute crazy camera move (background was going to be black - haha.) Anyway, kudos to you. Ah... the wonders of nodal point pan / tilt, eh?

Alex Boothby

[Jim Arthurs]

Thanks Alex... it really works smoothly, and the front projections kept their resolution quite nicely... I didn't have to blur the untouched master take for a good match. I was even able to dial in the exact amount of motion blur, which varied depending on if the orignal element was moving with or against the desired final direction. Bottom line, I can't wait to try it on RED footage!

The devil is in the details; undistorting the source footage correctly, exact knowldege of the camera FOV and various methods to re-align each take to overlay the master take.

I love nodal! Of course my camera isn't on true nodal, but that is only an issue with the inital SynthEyes track and I think in this case it wouldn't impact the outcome with a little forethought and planning...

I will do more after NAB as soon as I have time...

[Alex Boothby]

Great stuff. I bet you could use the same technique to generate a clean plate, which you could use to difference key the talent... er, farmer. This would save a few hours of roto for accurate mattes (too bad the damn clouds don't take direction well...)

[Mike Harrington]

the old icarus tracker could extract moving objects that didn't go with the overall optical flow.....basically would key your talent right out...as long as it was a nodal pan..

i think the foundry has something similar, that pulls a matte using optical flow...
http://www.thefoundry.co.uk/pkg_plugins.aspx?ui=44213E51-76C0-4926-9D5D-3A8059449900

[Alex Boothby]

Icarus was great. Could reconstruct a primative 3D obj from a scene after tracking it. Very handy for 3D matte painting, set extension, etc. Miss that Icarus....

[_BK]

This is very similar to a technique we use quite often when doing VFX for productions that can’t afford motion control. We just call it "poor man's motion control" and you don't have to be locked on a tripod for it to work.

First, If all the sequences you are matching are shot in the same environment, this is quite simple to do. You just need to get the camera op to duplicate the moves as closely as possible, then track each move individually in 3D space. Import the multiple solved moves into a 3D program and line them all up so the environments match up exactly. Create a master camera move from the sequence you like best and render out each sequence with that same camera. Then its just a matter of bringing the sequences into a compositing program and doing a bit of roto.

If the shot environments dont match up.. (ie. a background plate and a actor on a green screen), you just need to setup 4-6 tracking markers that are in the exact same positions in both shots. This way you still have reference points match up the 2+ 3D tracked environments.

3D tracking software is getting so good, Its almost faster to just use the "poor man's" approach than dealing with the setup times of traditional motion control onset.

[DeadEyesSmiling]

I was looking for inexpensive motion control rigs or alternatives and was pointed to this thread and that is a very impressive result! Would you be able to use a similar technique for a dolly move? Or perhaps a rotation shot but facing in rather than out?

-DES

[Jim Arthurs]

From the testing I've been doing, you are very limited the moment you move the camera in any way other than a standard pan/tilt/bank (yes, you can bank if near nodal).

Limiting to just pan/tilt/bank you can have almost total freedom in retiming your takes and independent camera motion for each take.

Regards,

[allano]

"3D tracking software is getting so good, Its almost faster to just use the "poor man's" approach than dealing with the setup times of traditional motion control onset"

I dont know who you have been working with but my moco setup is as fast as a normal camera setup - sometimes even faster , so its really up to the prod planning to decide where to use the extra hours involved with multilayer work.Off course it takes extra time to do several takes whether you do it with a "free" camera or a moco camera and the post use time no matter what.

I also think its a clever method "the poor mans moco" , and it can be used sometimes and sometimes you choose MoCo.Its a matter of evaluating the assignment and choose the proper tools with regards to end result , budget , work conditions etc.

But please dont say that its the end off moco - thats silly

Like saying that :"Finally we have made a video chip that gives the same looks as film emulsion" - would be the same as saying "See now we have the Mona Lisa painting on a LCD screen......."

I like people who share their inventions - respect for that Jim - If anyone has moco question please feel free :-)

[Curran Giddens]

I just put in my reservation for Curt's Mirus MoCo head. $10,750 with a $750 deposit. I think the retail price will be more like $15k. The special NAB price offer ends tomorrow! Just thought I'd let you guys know....

I will actually be using Curt's Mirus head (with MoCo dolly/jib) + wireless remote, repeatable FIZ on almost every shot. As long as the actor stays within the same path for each take, I won't have to use a camera op., ac, etc. for all but the initial setup.

[allano]

As long as the actor stays within the same path for each take, I won't have to use a camera op., ac, etc. for all but the initial setup.

Congrats on your order on the Miro - I also want one :-)

Most actors are good at remembering movements and marks etc. but dont expect them to be in MoCo mode - often it takes the concentration from the acting.

It is often the layer with the most dominant move you use to dictate the camera move.

With your new head you can remember many takes and use that for several versions of a move - but remember it takes time if you have to do every layer in several versions.I always recommend to do the move until your are happy - it requires a decison on set by someone but it is a fast way to work.

[Jeff Kilgroe]

I've reserved the Mirus head as well... The only thing that's really missing is a method to roll the camera, but Curt said they plan to have a roll add-on at some point. It's starting to look like the Mirus is going to be my primary studio head for RED and I'll be buying a more compact fluid head w/sticks (Possibly the OConnor 1030, but probably a Vinten V11) for everything else. But I guess the Mirus will be out of the studio a lot too... Lots of outdoor timelapse and moco stuff I want to shoot.

Curran sounds like he's trying to be a one-man army! I'd love to see that setup in action. :) We don't need no stinkin camera operators!

[CVB]

Jeff and Curran ... I appreciate your confidence in Mirus and in Us. It takes a couple guys like you to realize the potential of this product and blaze a trail. I can't wait to hand over some beta units to you guys in a couple months and see your response. Its funny because the reason that my Brothers and I developed Mirus was because we wanted to have the flexibility to do many things at once just as Curran suggested. Moco basically gives you time to think things thru and avoid having to choreograph dozens of things at the same instant in time.

[Curran Giddens]

Mirus will be my primary studio head too. I'm not even buying sticks or a mattebox until I see what View Factor comes up with since I don't have any immediate plans for using this stuff yet. Hopefully sometime next year I will have some of my own projects to work on. Maybe the MoCo dolly/jib will be ready about the same time I will be ready to use it. Right now I am just in the process of buying the equipment and starting my company. Not to mention working my ass off to pay for all this.

I can imagine just me and a non-professional actress (make that a gorgeous supermodel) in the studio. I can do a couple takes with camera moves and then pick the best one. Next step is remote FIZ until it is perfect. Finally, I can watch the the monitor and record audio while my "actress" does the take until I'm satisfied.

Good thing I don't plan on using any professional actors. The unions would never allow this sort of thing.

[Jeff Kilgroe]

Jeff and Curran ... I appreciate your confidence in Mirus and in Us.

I'm all for taking a few risks now and then and having a little pioneering spirit. Besides, it looks like you have a great product at a truly great price. Getting about 1/3rd off the price by placing a pre-order seemed too good to pass up and it almost puts the Mirus in the same price league as many of the larger, conventional fluid heads. For such a MoCo solution, this is mind-blowing. I've followed your contributions to these forums and have seen that Jim and the RED team hold you in high regard. I trust I am in good hands and look forward to using the Mirus.... Which definitely has potential to be an indispensable tool in my arsenal.

- Jeff A.K.A. Mirus reservation holder #11. :biggrin:

[allano]

I'm all for taking a few risks now and then and having a little pioneering spirit. :

Thank you - it have been a long process to get all the conservative DoPs , Operators , and Directors used to the "new" technique , but us pioneers will keep up the good spirit and develop machinery for you all so you can do what you imagine.

I remember first time I had to do multilayer playback in sync with the cameramove with audio in the ears of the actor doing multiple characters - you should have seen his face when he heard himself talking to himself - when he did the 7th part he was pretty much confused - and he was not even a union actor - funny stuff....

I just started having having 3D CG objects on set and its the same story - people have to get used to it - but end of day they love it.

PS : I have difficulties in getting in touch with the Miro man - are you there ?
I want to pre-order a Miro but did not get reply on my mail to you.Are you aiming at Europe also ?

[CVB]

Allan, We are definitely aiming at Europe. We are going to have CE and FCC testing completed in about three months. We will be shipping Mirus units around Mid-October. I'll double check our email and see if I can find your letter - We've had a spam problem that has forced us to upgrade our email filtering so its probably in there somewhere but I'll PM you with my email address just in case.

[Jim Arthurs]

Here's another run at the technique I originally made this thread about... the great part is setting up the exact timing between the two plates AFTER they are shot.

This was trival compared to the first example, but differs in that it's a practical real world useage example that we can all relate to. Again, ignore the roto quality... these are only tests...

http://ftp.datausa.com/imageshoppe/outgoing/PAN_TILT/NEAR_MISS.mov

[CVB]

Thats awesome Jim. Even with moco its hard to time that stuff sometimes. I didn't really notice the roto in realtime.. theres only a few frames where his legs get skinny but it looks fantastic.

[Jim Arthurs]

Thanks Curt, it was a great pleasure meeting you in person at NAB. You guys have some fantastic things coming out... all RED users need to support you and your projects in any way possible!

I've got one more of these quick tests left to process and share... inspired by you reminding me of the Kylie Minogue video... it involves passing off the camera from following one take to the next and next, then loops seemlessly back on itself. Well, I at least hope so... :)

[Michael Hastings]

Can you extend the preorder discount? It looks like you have a cool product, but I literally just noticed this whole thread and the near miss movie 5 minutes ago. I am definitely on a budget (and currently refinancing the house to do the RED purchase) and I have never done any motion control, but after looking at the near miss video and realizing the potential, I'm already thinking this would be really cool to use, but I need to read through the other information, and since the preorder form says it expires on April 30 I wonder if you would consider extending to May 30 or longer since there are probably a lot of us just finding out about it and since my RED is delayed (I'm 216 and was thinking I would have it in June but now it looks like longer) maybe I can convince you that it would make sense for you to extend the deadline..


Allan, We are definitely aiming at Europe. We are going to have CE and FCC testing completed in about three months. We will be shipping Mirus units around Mid-October. I'll double check our email and see if I can find your letter - We've had a spam problem that has forced us to upgrade our email filtering so its probably in there somewhere but I'll PM you with my email address just in case.

[CVB]

Aqua, just to clarify... the near miss video is a cg shot that Jim did without moco. He's got an awesome process where he can move a camera on a tripod without any synchronization whatsoever. In post he maps those shots to a sphere to create an effect similar to a motion control shot. That said... the head we are making can do the same thing and can do it much faster, the amount of post effects work is minimal (usually just a split screen or tracking roto). We have taken a few more reservations past the deadline already so I see no reason why we couldn't extend the deadline for the faithful reduser crowd - May 30th it is.. keeping the the price artificially low isn't a good way to make sure our company is here for the long haul, if we could get the raw components at a better price I'd love to keep the price where it is, unfortunately not all companies are like Red is :)

[Jaime Vallés]

That's really fantastic, Jim! An incredible feat, considering that it didn't require ILM or thousands of dollars to stage that seemingly dangerous "stunt". The applications of your technique are almost endless.

Very impressive!

[Andrew Benz]

...We have taken a few more reservations past the deadline already so I see no reason why we couldn't extend the deadline for the faithful reduser crowd - May 30th it is.. keeping the the price artificially low isn't a good way to make sure our company is here for the long haul, if we could get the raw components at a better price I'd love to keep the price where it is, unfortunately not all companies are like Red is :)

Damn, thanks alot Aqua...( I was wondering the same myself :biggrin: ). Curt , I am in... thank you very much! When are we looking at having to pay you in full? I am just trying to work all of it out without any loans.

Cheers,

Andrew

[CVB]

Damn, thanks alot Aqua...( I was wondering the same myself :biggrin: ). Curt , I am in... thank you very much! When are we looking at having to pay you in full? I am just trying to work all of it out without any loans.

Cheers,

Andrew

It's actually not that bad. There's fully refundable deposit and the remainder is due at shipping time in October. Check out our website for the reservation form and instructions.

[Andrew Benz]

Hi Curt,

Hehe... I was having a little fun w/ Aqua... I have been all over your website and I have as much faith in your (and your brothers) talent, skill, and design acumen as I do in RED. We have converging goals and of course you have my financial support (not much-but hey...), Your cost conscious/great design ethos doesn't hurt you either ;-). I am thankful to get the opportunity to buy great tools at affordable prices and extend my craft/creativity.

Just downloaded the reservation form, I will fax it off thursday or friday.


Cheers,

Andrew

[CVB]

Awesome.. look forward to having you as a customer. Make sure to tell your friends ;)

[allano]

I love the dirty way that this effect is achieved - it makes it easier for more people to make stuff like this - and I must tell you that I really is deligthed that you share - not all people are willing too share "secrets" like this....

But the MoCo force inside me , urge me to make a little PR about why MoCo is needed sometimes - just for the record and for those who have not been there yet (and also for convincing myself that there IS a future for MoCo after all)

In the real life shooting with a large crew , many people use the monitors to get the notion of what the heck is going on.On these monitors you can see in sync with the move all layers mixed and keyed if neccesary ; Something is important for the DoP other stuff is important for the Director , even the actor can have good use of seeing the Surroundings / 3D environment he is acting in , instead of all the green/blue stuff

e.g. its important that actors have the rigth eyeline if you want to "convince" the audience .The rigth eyeline is not always to look at the "other" object , but may be a tennisball moving on a C-stand , and this eyeline can be checked because you have the sync from the move , and the same move , in all layers. (When you use sound it is off course evident that the sound needs to follow the move in same timing )

Even though not many will take notice , the eyeline is a little off in the end (in the "near miss" that is - BTW a good title to a docu on my life, miss in the meaning Miss) , a little detail that is easy to over see , especially if its not possible ......not important in a single shoot maybe but imagine a feature film stuffed with effects where all actors look the wrong way all the time

BUT

I see many timig issues though that is better achieved with a "free" camera , but hey , different shoots different gear.

Can you use the xyz data from the head to make your trick even faster ?

Or is it easier to just track it ? (If the pan is very long dont you need multiple trackers etc ?)



babble babble off for work
best wishes from Copenhagen

[Jim Arthurs]

Hi Allan and all... thanks for your thoughts and comments. Again, I must stress that this is a test, and such things as roto and eyelines weren't important, what was important is the fact the multiple plates are moving in sync with each other independent of their original motion and without any form of mechanical sync during shooting. Heck, my "talent" wasn't prepped more than a minute or two for the shot, and I shot quickly with a single take for the car and person.

I agree with all of your comments about motion control. I LOVE motion control. Always have since I was a teenager building little stop motion dolly tracks out of wood and hardware store aluminium angles and rails. I remember reading a CineMagic article in the early 80's saying MOCO would NEVER be affordable for the indy... and look what is happening! I really want a small multi-axis rig for stop motion and miniature work, and I see Curt's work taking us in that direction quickly.

As an FX supervisor, I believe that any time you get another technique that overlaps an existing one it gives you OPTIONS. More choices is a good thing. Some shots this method is quicker, some shots MOCO is quicker. My method fails the moment you add a travel axis into the mix. Or try to do very low frame rates or stop motion, or need your multiple takes to have absolutely identical camera paths and retain your full frame raster for all takes, or...

One clarification for Curt... the roto and comp work I'm doing is EXACTLY the same, whether MOCO or tracked... in this way they are absolutely identical. The guy passing in front of the car would need to be roto'ed whether MOCO or retargeted motion, same with adjusting for changing light conditions, etc. I have no more (or less) roto and comp work with this than the same setup would have with MOCO.

The actual client driven project that I used this technique on I won't be able to share for several months. That's why I'm showing off some of the work product that got me to this point and continuing to shoot new tests to improve and refine the method for future work. It's fun, and I share in that spirit. If there's something someone can take from this, and use in a project, then all the better.

[Michael Hastings]

Thanks, Curt I have been manufacturing underwater housings and lights for 25 years so understand the pricing/business issues and appreciate the extra time.

Also thanks for the explanation of Jim's technique. I think my basic idea is that your system provides a pretty low cost way to get some exciting effects into a low budget feature.

I have a feature idea I have kicked around for a few years, and one of the components was that our ability to bring a lot of underwater work to it at relatively low cost would be our primary "effect". But I also I have kept my eye out for anything else that could be leveraged to give a bigger budget feel to a low budget feature - and I see your system as a real possibility.


Aqua, just to clarify... the near miss video is a cg shot that Jim did without moco. He's got an awesome process where he can move a camera on a tripod without any synchronization whatsoever. In post he maps those shots to a sphere to create an effect similar to a motion control shot. That said... the head we are making can do the same thing and can do it much faster, the amount of post effects work is minimal (usually just a split screen or tracking roto). We have taken a few more reservations past the deadline already so I see no reason why we couldn't extend the deadline for the faithful reduser crowd - May 30th it is.. keeping the the price artificially low isn't a good way to make sure our company is here for the long haul, if we could get the raw components at a better price I'd love to keep the price where it is, unfortunately not all companies are like Red is :)