At long last it’s here! A podcast dedicated to the real-time visual effects community.
Real-Time Visual Effects is a field in high demand. Worlds are transformed with the magic of RTVFX. This podcast explores the industry at large by being a platform for people to learn and grow within the community. Be a wizard and make magic.
Because the way SEO and approval processes work for RSS feeds, there can be a delay (sometimes weeks) for it to appear. I’m working with other entities to get it uploaded. If you have a specific app you are looking at, let me know and I’ll reach out!
Thank you! I added it to my PocketCast and also listened to both episodes. Really cool! Do you have a twitter-account for the podcast, would love to share it!
And thanks for mentioning my little UnrealSimon-Channel <3 But I’m afraid my video is about an older version of slate and the UI changed quiet a bit I think.
p.s. I really like that you added links to the guests websites into the description. That helps a lot to check out the stuff quickly
No Twitter account yet! Maybe that will change as the podcast grows.
You’re absolutely right! Tools update so quickly. It’s hard to keep up. Who owns Slate?
Awesome podcast! I’ve listened to all episodes. Please keep it up.
The intro is really inspiring.
I’d like to hear more about optimization techniques and practical stuff in general.
Thanks!
Just listening to EP3 with @Niels. Super interesting! I’m always said when the episode is already over. For my taste it could be 3x as long
Are you pre-recording them in some kind of season and then release them one after another or are you recording one episode, do the cut, release and then make a new date with a new guest?
I have one question which I would love to hear an answer for from all the guest: how long takes a vfx for them from start to finish. I’m not talking about the visual aspects but from first concept to the final release-ready version in the game. For me it takes usually a very long time as I discover often that some systems in the engine are not ready yet to e.g. trigger the effect at the right time and then those system must be developed, tested, bugfixed and while iteration over the effect more and more “gaps” turn up. It can also be graphical bugs like “why does my spline-mesh has no motion blur while everything else has?”. This sums up to many days production time per effect. I wonder if it’s the same for everyone
Podcasting is highly demanding in both time and money. There is so much planning behind the scenes so to answer your question, I have ideally 5 episodes queue’d at a time to ensure there’s a weekly schedule. If I were to get sick, my audience would still expect a new release. There are a number of guests that are in the pipeline-- some that I’ve already interviewed, some scheduled for interviews, some on my wish-list of interviews. At the present time, it’s 1 ongoing series but I might segment them based on ‘seasons’ at some point.
While guest interviews are great, there are new episodes planned in the future that expand on this. Stay tuned!
As mentioned in the episodes, if you have a specific question for a guest, reach out to them directly! I’ve included their contact information in the show notes.
Cool new episodes! I liked the inbetween with all the thoughts and ideas regarding the current crisis. Also I like the new one and I’m glad to hear that Anthony confirms my own thought that it’s almost impossible to tell “how long does it take?” for most effects without knowing a lot of the context.
Btw.: Did you think about adding an answer to this thread each time you release a new episode? So that people can easily see if there’s a new episode (but they don’t use a podcatcher yet)?
I’m looking to pay for a dedicated domain name in the future so I’ll be sure to update these links appropriately. Right now I’m just trying to keep costs to a minimum.
Yes, it can get complicated quickly when it comes to getting an own domain and also webspace. But I prefer it to being dependent on a 3rd party provider who may decide from one day to another to kick your project or go out of business.
Very cool new episode! Enjoyed listen to you guys.
Finally checked out the podcast Travis. Awesome job! only a few episodes in so far but really enjoying it. Also really like the length of the episodes. Makes them easy to digest.
Awesome new Episode! I think the strength of Podcasting is, that it’s not so restricted in time like a show in radio or TV. So if there are amazing guests with a lot of cool stuff to talk about, I don’t mind even if the Episode is 3h long.
Thank you so much for mentioning my effect @imbueFX - what an honor <3 By the way, as you are so much in love with mountain biking: Friends of mine released their mountian biking game a while ago which you might like: https://lonelymountains.com/
About the topic with “cards are stacked against you” and “we’re all competitors”: I think it’s true but not as harsh as it sounds. When I was in games school, of course we all where 3D artists and competed against eachother, when it came to applications at the end of the studies BUT we all developed our own styles and preferences. With that, we all went different ways which led us to different companies.
During studies I fell in love with Fantasy Environments and I found a company who did a fantasy rpg. The guys who loved to create weapons and cars went to Crytek … so everyone found their place.
Not sure if it’s understandable what I mean.
Thanks @TravisMcCallumto bringing us this cool podcast! I’m super happy to hear that you play with the thought bringing me in as well By the way: I would volunteer to interview YOU! I’ve the feeling that we know a lot about your guests but not too much about you
Funny as I was thinking the other day if a mountain biking game would have a big enough appeal. The game looks fun, similar to something like Trials Evolution. I will most certainly be buying it!
So I didn’t feel like I gave a good enough answer about why deep learning interests me and its relation to FX. So here’s my mulligan:
I used the Nvidia upscaling demo as an example, where you can feed the neural nets super sampled renderings. What if we did this for fluid sims in real time? Let’s say we can render really low res, crude voxel data that doesn’t kill perf (it could something in screen space like velocities, vectors, etc.) That data then runs in segments in the screen and compares that to x amount of pre-rendered, film quality fluid sim renderings to composite visuals. I mean, I’m not smart enough to understand it, but it sounds game-changing.
I was following Steven Knippings Applied Houdini Course about Volumes and actually they have something like this. There is an NVidia Plugin which assumes from a quiet lower res rendering an improved version directly in Houdini. The only problem right now is, that it focuses only on 1 frame and doesn’t “know” about the others so I guess it would flicker a lot if you use it for an image sequence. But in the future, this will most likely be a common workflow - or we render everything directly in real-time with Embergen
By the way: I would volunteer to interview YOU!