Forum Replies Created
Mike SevignyFebruary 27, 2014 at 4:15 pm
What you’re describing are practical effects. Bald cap and makeup will go a long way. If he turns his head too fast or DOF gets in the way you’ll quickly find you have a nightmare on your hands in post.
In terms of tutorials, you’re looking for face/head tracking tutorial so you can attach adjustment layers and paint to his face and head. I would be thinking track marks. Use an eyeliner pencil to put dots on his face.. but these will have to be painted out. You may actually have to matchmove his head in some cases which really ramps up the workload.
Depending on your skill level this may be out of reach. If I was supervising I would opt for makeup and only be hired for cleanup.
Not an easy task unfortunately..
Mike SevignyFebruary 26, 2014 at 9:41 pm
The effect you’re looking for is ‘video to ascii’ to give you somewhere to start. Here’s a tutorial I found that seems to have the expression you’re looking for.
Pretty neat… bookmark.
Mike SevignyFebruary 26, 2014 at 9:32 pm
For the stabilizing it sounds like you need something like
NewBlue Stabilizer or Lock & Load X to batch everything right in the NLE. Check out their tutorial section and see if it’ll do what you need.
Coremelt – Lock & Load X
Mocha should have worked in after effects though. In case you still plan to run all your clips through After Effects you can do the following to batch your renders.
1. Import all your clips and drag them to the 3rd icon from the left at the bottom of your project panel (Create a new composition)
2. In the window that pops up you can choose to put all the clips into one composition and sequence them (render them out as one long video), or create a new comp for each clip (maintain the current file structure).
3. Once you’re ready to render.. Select one of the compositions in the project panel and go to Composition>Add to render queue
4. In the new panel click on the down arrow next to the yellow word ‘Lossless’, go to ‘Make Template’. In the popup window setup the render settings you want (edit button), name it and then make it the default video template.
5. In the render queue panel click on the yellow text to the right of ‘Output To:’ and select a common folder where all the files will be rendered to.
6. Now you can select all of the other Compositions at once in the project panel and add them to the Render Queue.
Mike SevignyFebruary 26, 2014 at 4:52 pm
I’m not sure this will look perfect when your done with it no matter what technique you use. the motion blur will be a real pain. Here’s what I would try if I was faced with it.
1. Track the shot in mocha and attach the movement to a null object
2. Create a solid and parent it to your track, then roto the character on the solid to use it as a matte.
3. Create an adjustment layer, parent it to the null, apply wire/rig remover effect and adjust it to get rid of your grid
Some additional work to bring it all together will likely be necessary. Hopefully this is a good starting point.
Mike SevignyFebruary 25, 2014 at 10:58 pm
If you’re just exporting by going to File>Export then I would recommend using the Render Queue.
When you are ready to render:
1. Go to Composition > Add to render queue
2. Click on the yellow word ‘Lossless’ in the render queue window
3. Select your desired video compression (SWF)
4. Check ‘audio output’ at the bottom of that window and pick your settings. (when done, hit ‘ok’)
5. click on the yellow text to the right of ‘Output to’ in the render queue.. and pick where you want to output.
6. Click render (Button is to the right on the render queue window)
Hope this helps,
Mike SevignyFebruary 25, 2014 at 10:42 pm
The CPUs is not the only factor for speed when processing video. Make sure you’ve got plenty of ram and more importantly, a good video card. I’ve been building computers all my life but I haven’t had much success building a solid machine for editing/vfx that lasted more then a year or two so I can’t help you when it comes to specific hardware. What I would suggest is modelling your build around something from https://www.boxxtech.com/
I picked up a customized 4920 Xtreme station a year ago and I will be buying my workhorses from them in the future. The prices don’t seem competitive on the surface but you will not be disappointing.
You may want to post your question in a hardware forum.
Mike SevignyFebruary 25, 2014 at 10:26 pm
In your timeline window (at the top) deselect ‘Draft 3D’. I would turn down the settings before I do that.
Mike SevignyFebruary 25, 2014 at 4:29 pm
This intro looks great. You’re right when you say that the type doesn’t read well in Element. Perhaps keep them as 2D overlays but only have them come in at the end when the flash/pan happens.
I would argue that the ‘1320 video’ doesn’t really read either because it’s so close in color to the background. Is there a way to add a stroke effect around the text or the frame? I’m not that good with Element 3D so I don’t know what specifically you need to tweak for that.
Alternatively you could make the type with after effect’s default extruded 3D text and material options. That would likely give you the control you’re looking for.
Really nice work,
Mike SevignyFebruary 25, 2014 at 2:43 pm
I will have to check that out. The C4D workflow is looking more and more attractive.. I used to work in Max when I was inhouse but those are pricy apps. I’m getting the C4D demo right now.
Thanks for replying,
Mike SevignyFebruary 25, 2014 at 2:21 pm
good question Matt.
I’ve been experimenting with this technique for some time now. It’s been recommended to me to use projection man with cinema 4D but I’ve yet to check that out. The workflow between C4D and AE is tightening every version.
I’ve also been trying to speed up the process of the scene building in AE. I did not include that in the tutorial because it required many additional steps (to a tutorial).
Faster scene build: When you take the picture, take a small video with parallax as well. Complete your camera track in your favorite matchmove software (AE now does 3D camera track) and then use the point cloud to quickly position your objects accurately. This requires that you have control over the shoot when the image is taken, which is not usually the case.
I don’t think there’s any software that will know where everything is in 3D space based on the still.
Thanks for the feedack,