Outpost Alpha - Full Trailer
https://youtu.be/dg3Cbtl00wo

https://redd.it/1rz40dm
@vfxbackup

Which smoke exhale looks most real?

Doing a "smoking candle" trend and trying to make the exhale look 100% real so it actually confuses people.

The smoke variations are labeled in this video: Smoke 1-6

Smoke 1: Final (can't edit).
Smoke 2, 3, 6: Haven't been modified yet. I have full control over turbulence, shading, and speed.
Smoke 2 & 3: Opacity is set high for the screen recording. Lighting: These still need a lighting pass. These are different speed but same. 2 is slower and 3 is faster.
Smoke 2-6 all can be edited further

Looking for feedback on:

1. Which option (1-6) has the most natural physics and "swirl"?
2. Any specific direction on color, shape, or volume to hit 100% realism?
3. Tips on mouth integration so it doesn't look like a digital overlay?
4. What would you change, like, or hate about these?

(High-quality renders coming once completed).

https://redd.it/1rz5a5l
@vfxbackup

how are you guys handling the "can you remove the background from this video" requests from clients

is it just me or has every client on earth suddenly decided they need video background removal? i swear this was a once in a while request and now its like half my inbox



the typical scenario goes like this. client shoots an interview or a talking head in their office. no green screen, no lighting thought, just an iphone on a tripod pointed at someone sitting at their desk with a bookshelf and a dying plant behind them. then they email me asking if i can "just remove the background and put our logo behind them instead." and they need it by tomorrow



last month alone i had:



a real estate agent who wanted his weekly market update videos composited over property photos. he shoots them in his car. IN HIS CAR. he wanted the car interior removed. seven videos a month



a startup that does all hands meetings on zoom and wants their CEO isolated with a transparent background so they can drop him over a branded slide deck for external clips. these come in batches of 4-5 per week



a wedding videographer friend who wanted the couple cut out from a reception clip to put over a different background for the save the date video. moving people, dim lighting, terrible footage



an ecommerce brand that needs product demo videos with transparent backgrounds for their website. the presenter holds up products and talks about them against a white wall



every one of these is a different situation with different footage quality. the real estate guy's car footage is basically impossible. the zoom recordings are easy. the wedding stuff is a nightmare. the product demos are hit or miss depending on how the lighting is



what ive been doing



for the zoom/talking head stuff where the person is mostly still and the background is simple, ive been using online removal tools. Unscreen Pro for the quick turnaround stuff, DaVinci Resolve's Magic Mask when i need more control. honestly for a person sitting at a desk talking these work fine and take like 2 minutes instead of 30 minutes of rotoscoping



for anything with movement or complex backgrounds i still do it properly in Resolve or AE. theres no shortcut for the wedding footage or the guy in his car. those just take time



the part that kills me is pricing these jobs. the easy ones take 5 minutes now but clients dont know that. the hard ones take hours. and clients think they should all cost the same because "its the same thing, just remove the background right?"



what i want to know



how are other freelancers handling this? are you charging per clip regardless of difficulty? do you have a standard rate for bg removal? do you just tell clients no when the footage is garbage?



also for the people who do a lot of these, whats your actual workflow? green screen when possible and tools for the rest? or are you doing everything manually still

https://redd.it/1rz96th
@vfxbackup

Houdini swirl vellum quick process

https://redd.it/1rzm0g2
@vfxbackup

Shooting at various interiors and looking to comp the outside of the windows to look like the view from a train. How do we shoot and achieve this?

https://redd.it/1rzxi98
@vfxbackup

PAID VFX Compositor Needed for Indie Short Film (Greenscreen, 45 Shots)

Hi,

We’re looking for a VFX compositor to collaborate on an independent short film sequence (entitled the “Star World”).

\~45 shots, \~4 minutes total
Greenscreen footage with an existing environment already started
Goal is a more polished, photoreal integration (lighting, plants, star field, transitions)
Some shots remain on greenscreen; a separate artist is handling a dragonfly animation (not part of this scope).

Previous work was done in Fusion, but we’re open to Fusion or Nuke. Project files, tracking, and assets from the previous artist are available.

Timeline: mid-May delivery
Budget: \~$11,000 for the sequence

Looking for someone with strong greenscreen compositing and environment integration experience.

Feel free to comment or DM with your reel (especially relevant compositing work). Happy to talk further and share reference shots and footage via DM as well.

Thanks!

https://redd.it/1rzzimg
@vfxbackup

Ben Affleck's new AI Startup think they can reduce VFX cost by 50%

I ended up finding the patent submitted for their company and found some claims that does not align with the message he's been spreading about his AI use.

Notably :

"AI service providers can generate substantial revenue by charging a fraction of the cost savings as their fee, thereby creating a win-win situation for both the film industry and the technology sector. The technology also presents an opportunity to replace the cost associated with background artists, second or splinter unit aerials, inserts, and reshoots, achieving a day one 20% reduction in schedule and physical production and a 50% reduction in visual effects (VFX) cost."

And hes been on record saying that he doesn't believe in prompting to replace filmmaking but his patent clearly emphasize that use :

"By doing so, the Filmmaker models may be trained to understand and apply a wide range of cinematic techniques and language, thus enabling the models to generate video content that adheres to professional filmmaking standards, as parameterized by user inputs (i.e., prompts)."

That specific patent : https://patents.google.com/patent/US12438995B1/en?inventor=Benjamin+Geza+Affleck-Boldt

all the other patents they've filed : https://patents.google.com/?inventor=Benjamin+Geza+Affleck-Boldt

https://redd.it/1s61kny
@vfxbackup

After Effects - Converting footage to particles

https://redd.it/1s606c1
@vfxbackup

How was Naomi Scott's "Speechless" song in Aladdin (2019) done?

This is my luddite view of what happens:

She starts by walking with the guards, I imagine this is actual footage. Then she spins around and the guards vanish, so clearly CG here.

She proceeds to walk down stairs, I'm guessing looking at the surface of the balcony this is some sort of textured version of the physical set/steps she is walking down.

Then what really interests me is how the camera that is following her in reverse throughout transitions into a seamless 720 around her. Again from my noob perspective I thought how this was done, could this be robotics, but then I thought there is surely no way that She or her team would permit a robot swiveling around her, so I came to the conclusion this is surely CG, but how was this 720 done?

She then continues to sing and walk, various other assets seamlessly vanish like 2 more guards beyond the double doors followed by another tight 720 around her. She tears down the cloth and it seamlessly then floats in the air as she walks away. A few more appearing and disappearing assets to complete the tune.

Bearing in mind this entire sequence is visually one take, no cuts.

How was this done? Just interested to know.

https://redd.it/1s6ihao
@vfxbackup

Portfolio

https://redd.it/1s62osg
@vfxbackup

Best way to make realistic Rorschach inkblots

I’m trying to make photorealistic inkblots digitally (stills, not animation). Something that actually feels like ink got dropped on paper, the paper got folded, pressed, then opened.

What would you use and how would you go about doing this? This site has examples of what I'm trying to produce: https://www.rorschach.org/

https://redd.it/1s6jq2x
@vfxbackup

Made some changes to my scene. Would love to hear if there is anything else i can do :)
https://www.reddit.com/gallery/1s5xgqg

https://redd.it/1s5xqex
@vfxbackup

What does it mean to be a 3D Technical Generalist?

What kind of skills and knowledge would a person need to have to be a 3D technical generalist?

I really love making art and I was a traditional art student, but I am thinking of changing my area of study from 2d traditional art to 3d art. I did a little research and found out what a 3D technical generalist was and it sounds like a goal I should strive for, as a start.

At the moment I am taking a course on GameDev about Blender to start wrapping my head around how Blender and 3d art work in general as a start. Any advice on how to start studying or researching would be very helpful.

Thank you.

https://redd.it/1s6f8t7
@vfxbackup

Portfolio

https://redd.it/1s62qhy
@vfxbackup

: 5 years into VFX (Texture Artist)---- stick it out or pivot?

24M, ~5 years in VFX, currently working as a texture artist at a well-known studio (in India).
I’ll keep it real I’m feeling stuck.

Been at my current studio for 3+ years. Pay growth is meh, work hasn’t really leveled up, and most of the “good” shots go to seniors or other sites. What we get is usually leftovers or production work.
I do my job well, get good feedback, no issues with performance but none of that seems to translate into actual growth.

Biggest issue: the work doesn’t feel like art anymore. It feels like assembly line work inside a fixed pipeline. Do task ---submit ---- repeat.
On top of that, the industry overall just feels… unstable. Everyone I talk to like my coworkers, friends in other studios same vibe: uncertainty, layoffs, low confidence in the future.


Tried looking for other jobs LinkedIn is brutal. Barely any relevant openings, tons of applicants, and mostly no replies. Feels like hiring is heavily referral-based or internal.


I still work on my portfolio (interested in environments/lookdev), but honestly, motivation takes a hit when the bigger picture looks like this.


So I’m at a crossroads:
Do I double down, improve my portfolio, and try for bigger studios like Weta FX, Industrial Light & Magic, Cinesite, or Eyeline Studios (even though I’m not based in those countries)?
Or do I start planning an exit while I still have a stable job?


Switching isn’t easy either can’t fully commit to something new while working full-time, and I don’t want to half-ass both and end up worse off.
Also feels like specialist roles (model/texture) are getting saturated, and everyone wants unrealistic “generalists who know everything.”


If you were in my position, what would you do?
Stay and push harder, or start pivoting?

https://redd.it/1s6zvqd
@vfxbackup

We recreated a Fast & Furious scene with my friends.

https://redd.it/1s71pvn
@vfxbackup

Text to 3D tools in production pipelines, is anyone actually using these

Genuine question for other artists and tech people. Has anyone integrated AI 3D generation into a real production pipeline? Not personal experiments but something that actually ships.

I've been evaluating tools for our studio, 15 person team working on a stylized action project. Art team is overloaded and we need to speed up environment asset production without hiring.

Tested Meshy, Tripo, and a couple others over the past few weeks. Generated about 50 test assets across all tools.

Raw output is usable for distant LODs and background clutter but not production ready for anything mid distance or closer. Topology is universally bad across all tools, every model needs retopo. Meshy had the most consistent PBR textures, workable base after a material pass. Style consistency is the biggest unsolved problem though, outputs vary even with identical prompts.

We built a Houdini script to auto retopo and it helps but adds pipeline complexity.

My current take is AI generation works as a "first draft" tool that replaces the initial blockout phase. It does not replace the modeling team. Just shifts their work from creation to refinement.

Estimated time savings if we integrate properly: maybe 20-30% on environment props specifically. Characters and hero assets are still fully manual.

https://redd.it/1s70ugc
@vfxbackup

3D Gaussian splat, in vfx
https://www.youtube.com/watch?v=X8yRlA7jqEQ

https://redd.it/1s7063u
@vfxbackup