3D-Pangel

On 5/28/2023 at 5:20 AM, hyyde said:

 Mouse responds to movement after 10 sec. not ideal but i don't mind that , as long as no BSODs will happen. 

Thanks for suggesting this MASH. 

Any latency in the mouse or the keyboard is not acceptable to me....let alone 10 seconds of latency.

Is it wired or wireless?  My one big knock against Logitech wireless mice and keyboards is that their unified receivers can be a bit buggy particularly on their M720 mouse and K850 keyboard that can support up to three different computers.  Using their unified receivers, I had to keep a wired keyboard and mouse around just to get myself out of jams should everything hang.  They have been working to improve their software stability, but I remain doubtful.  No issues with using a Bluetooth connection though but they can only support one out of the three PC's with that connection type.  If anyone knows of a more a VERY ROBUST wireless mouse/keyboard combination that can support more than one PC, please let me know.

With all that said, you have more patience than I do if you can live with 10 second latency.  Speaking from experience, that type of issue would drive me crazy super quick.

Dave 

Apart from having a nostalgic fondness for 30-year-old 3D software (I come from the tueSpace days), there are other indications of how age is affecting your work that you need to watch out for.....case in point:

Wow....it is cold in here?  Where is my sweater!!!!

Dave

That VFX breakdown provided a treasure trove of reference images. 

Great stuff.

I understand that while Picard had mixed reviews, the final episode in Season 3 made it all worth it.  Not enough of a payback there (IMHO) to warrant such an investment from me in terms of time and money to catch this series on Paramount+ streaming...but happy that fan service was ultimately done.

Dave

I agree and also disagree at the same time.  It was a big task and a gamble to create Harold Ramis digitally because if it was off, it would pull you out of the story at what was supposed to be a very emotional point at the climax of the movie.  Now that would be the same risk as hiring an actor and using make-up to do the same thing - the difference being that you can only go so far with an actor but you can go all the way (given enough time and budget) to do it digitally.  Given the importance of having Egon appear to his children at the end of the movie, I think it was a smart move to try and do it 100% perfect digitally.  It sold me and supported the movie the way it was meant to from a story standpoint.

Story points aside, the choice between using actors or digital doubles can be hit or miss either way.  Case in point, the recreation of Peter Cushing as Grand Moff Tarkin at the end of Revenge of the Sith.  Long before the days of perfecting digital doubles, they used an actor in make-up, and it was just a bit off:

But as he only had one small scene and no close-ups, they could get away with it but still when I saw it my immediate reaction was "the make-up is slightly off".  Now some would argue that his digital recreation in Rogue One was also a bit off but (IMHO) it was a far better recreation and faced the additional challenges of having dialogue and the acting that goes with giving that dialogue that Egon did not have in Ghostbusters.  Remember that up until Rogue One, most digital doubles did not have that much dialogue or were not required to carry important story points forward.  They just walked on....mostly silent (like in Blade Runner 2049 or Logan) and that was it.

Just my 2 cents.

Dave

I am 62.  I have no use for any social media outside of the Core4D.

Now....get off my lawn!!!!  

Dave

....hit play.....

Cool!!!!

Dave

....what time is it?  

On 5/11/2023 at 2:21 AM, EricNS said:

 

It's not a rumour, it is clearly indicated in the "How it works":

 

 

"An input image is submitted and passed through our AI algorithm for reconstruction.

The output mesh passes from our Quality Control pipeline to make sure that flaws are eliminated. That process aims to eliminate most of the unwanted technical characteristics of the generated models (i.e. tris, n-gons).

A Quality Control team member will check and improve the output where necessary for the 3D model to match our quality standards. Yes, our algorithm is not always perfect, and we want to ensure the output quality is always up to a high standard.

The output is returned to the user for further editing and the addition of textures."

Honestly, it is all crap unless they show you the mesh for the simple reason that someone has to still texture the model.  Even with human intervention, I would want to see the quality of the finished product (that is the mesh) simply because I don't think it is physically and financially possible to find, hire and retain enough skilled modelers to take the mind-numbing job of fixing bad geometry with enough quality at a fast enough rate to make money.

You just can't outsource Cerbera to a machine.  Just not possible.

Dave

I thought I would provide an update to this extremely old thread.  The whole project of modeling the Death Star Equitorial trench got a bit sidetracked by a C4D's shiny new feature: pyro.  That was hard to resist.  So, I got a little crazy and went off and dived into something a bit way above my skill level and took on replicating the City Destroyers emerging from a fiery cloud from 1990's Independence Day.  You may have seen that thread.

Well....with fluids out of my system (for the moment....), I went back to my Death Star WIP.  Like the Independence Day scene, doing hard surface modeling on essentially something the size of a small moon was also a bit crazy.  Definitely punching above my weight class again.

Well...here is the update WIP:

Still lots more to do and many adjustments to be made, but I think it is coming along better than I personally thought possible (trust me...there were many times I wanted to give up).

Dave

On 5/7/2023 at 7:08 AM, Cerbera said:

Some signs are just...wrong...

 

 

CBR

I would agree....

Well....I wish I could provide the same instructional benefit to you as you have provided to me but the only areas of expertise where I would presume to be able to offer advice would be in electronics manufacturing.  Relative to software, then definitely Microsoft Excel and a better than average knowledge of other MS tools I use at work like Smartheets,  Powerpoint, Word, etc.  I also have a pretty good knowledge base into the history of VFX up through the early 2000's buy then the rapid growth of CG techniques just became too fast to keep up with.  

But that's it!  Which is probably all pretty useless to a CG professional such as yourself.  So thus, my enduring appreciation.

Dave

12 hours ago, Cerbera said:

 

I'm glad you find them helpful ! I always worry that my thoughts on these things tend to emerge quite chaotically, and I usually forget at least 3 things I meant to say in every one, so nice to hear people find the main points are getting through !

 

CBR

It is evident from those short tutorials that you are extremely busy and were making those tutorials from what limited free time you could squeeze from your schedule (BTW: Now it is Sunday )

But....they do make a huge difference to me and I am sure to others.  They don't need to be long because as I said you packed more useful information into those short tutorials than I have seen in over half the tutorials out there.

So I am sincerely appreciative for whatever you can provide in the time that you have.

Dave

There was more information packed into that 9 minutes  than I have seen in full hour tutorials.

These need to become a permanent part of Core4D.  Can they be added to the Videos section please?

Thanks,

Dave

Gold....pure gold.  

Amazing how something becomes obvious ONLY AFTER it is shown to you.  All I can say is "Duh!!!" at that point and slap my forehead in disgust.

Thank you and I look forward to more!!!!

Dave

On 5/4/2023 at 7:10 AM, Cerbera said:

Questions aspiring modellers may want to ask...

 

1. Do we need to worry about the complex poles remaining, both before and after the bevels ?

2. If not, why, and if so, why ?

3. If we do wish they weren't there, what can be done / is already being done about them ?

 

CBR

The humble padawan will throw himself at the mercy of the Master and attempt to answer

1. Do we need to worry about the complex poles remaining, both before and after the bevels ?

Complex poles can lead to problems but in some cases are unavoidable.   So, the questions that need to be considered is whether or not the complex poles preserve good edge flow?  Are any of those poles made up of polygons that are not quads?    In short, quads and edge flows are the goals to be achieved at the base polygon level before you apply bevel modifiers.  If you can achieve that goal but create some complex poles in the process, then that is a good trade-off.

2. If not, why, and if so, why ?

See answer to 1.

3. If we do wish they weren't there, what can be done / is already being done about them ?

Well, that is where your genius comes in.  You need to visualize the result in your head and think in 3D what is the solution.  Then you have to figure out the most efficient way to get their using the modeling tools you have.  Being able to visualize abstract problems in 3D is a function of genius thinking (according to Mensa).  Not everyone can do it.  But as you are also a mighty fine composer then it is pretty well established how well your brain is working.  For us mere mortals, that is sometimes not so easy. In some cases, I have had to throw out the model and start all over again when I get into those situations.

But a question for you:

I don't quite follow your explanation of the first step: 

My approach would have been to start with a cylinder as you did, then save a selection of all interior edges.  Select all the faces and run the "Poke Polygon" command.  Deselect everything and re-select your tag of saved interior edges and hit dissolve.  

Now, while quicker, I will admit that your solution looks better in that your diamonds have edges that are equal lengths on all sides which will never happen with the poke polygon command.

So how did you do it?

Plus running the loop cut with quantize subdivision fixed at 1 (to give you 50%) and bidirectional set to off does work but then, as you get to the end, the cutting goes all wonky 

How did you get around this?  Did you have to start making specific selections and restrict the cut to only those selections or is there some other setting I am missing?  

And I still have no idea how you got to here while still preserving the curvature of the glass:

Yes...now you know why I am amazed...especially as it only took you 15 minutes to both visualize the solution and to execute it ((it took me over 90 minutes just to figure out most of what you did)!!!!!!!

Dave

I missed the fact there were ngons in that model.  You are right that as the SDS was very forgiving, their impact was not clearly evident in the final render.

But I am even more amazed.  Here is why:

Gaining an understanding of as many of these modeling techniques as possible is gold to me!!!

Dave

Wow!  Just wow...on so many levels.  Perfect topology,  Curved surfaces,  Undercuts.  Honestly....just wow!

Dave

P.S. I would like to thank Jay's parents for the use of his baby picture in this post.....

....just kidding.

...that is actually Igor.

Very nice.  Thank you.

Please note that this generator is also good for making cheese!!!!  

Dave

My hope is that he does make an amazing update and charges for it.  But....given his rant, I think he annoyed potential customers more than engender sympathy (though I do sympathize with him).

I have version 1.3.0 from 2019 which I hope is the latest version....but probably not given it was 4 years ago.

If anyone is looking for a good sci-fi generator, then check out:

https://www.artstation.com/marketplace/p/vD0jG/procedural-sci-fi-material-generator

It works with Substance player which used to work within C4D (not sure if it is still there in 2023.2).  But if not, get the free Substance player to generate your maps.  For $15, pretty good deal for all that it does.

Dave

Okay....I thought I was done.,...

The link though for the "Modelers" was too good not to include:

Meme - A 3D model collection by dubboi_49 - Sketchfab

Okay...now I am done.

I just came across this one:

My personal favorites

And another oldie but goodie

...and one that I just came up with....

Dave

1 hour ago, HappyPolygon said:

 

That's actually a feature. It is mentioned that they simulated how tires compress and deform (1:38)

Yes...I know.   Poor attempt at humor I guess.

Dave

There has to be a downside to using Unreal Engine?  I mean, even Houdini has a downside (the interface takes some getting used to).  I originally thought it was how Unreal 5.0's render engine handled the leaves on a tree or any thin object as they would completely disappear if you moved the camera far enough away.  Also there was the persistent flickering on rendering smooth surfaces (that is why all the demos when Unreal came out were of rocky canyons or stone temples.  But this 5.2 demo shows that all those flaws have been worked out.

I have heard that the GUI is a bit difficult to work with so if anyone can speak to the learning curve, then that would be helpful.  I will admit it takes me longer to learn how to accomplish things with nodes.  Too many years of building routines using written code I guess.

But that was an impressive demo and begs the question that if photo real can also mimic real work physics in real time, why learn anything else?

Of course, I did find one flaw:

That tire needs some air. 

Dave

11 hours ago, HappyPolygon said:

It's been 3 days now that I watch NAB 2023

Every single presenter showing how easy and fast Soft Bodies are:

- ... and we make a Soft Body tag that looks like a balloon on the object we want to inflate and we press play !

*the object just drops*

- Ooops, forgot to make a floor.

*object passes through plain*

- And this happens because we need to set a collider ...

me:

How many actual presenters made this same mistake - as you said "ever single presenter" working with soft bodies?

Honestly, adding few tags does not make soft bodies more difficult than the presenter's intent to convey their ease of use.

I would attribute it to the simple jitters that come with making public presentations.

....or maybe they all got flustered from being in the presence of Rick Barrett??  

...err...maybe not.

Dave

2 hours ago, HappyPolygon said:

 

 

Recently I found out that Poser still exists ...

POVray is still alive, I still have no idea how people use it

TopMod is still up (interesting modelling tools)

Astonished to find out Wings3D is still maintained

Completely forgot about Poser, POVray and Wings3D.  Never heard of TopMod.  For some reason, I always thought Poser was a stripped down version of DAZ Studio but those two programs never crossed paths.   Poser though has passed through a quite a number of hands:

1995 - Fractal Design Corp

1998 - Metacreations (when they were buying everything)

1999 - Curios Labs

2006 - e-Frontier

2008 - Smith Micro Software

2019 - Bondware Inc

I guess they make the news that Lightwave switched owners seem pretty "old-hat".

One other piece of software that comes to mind, though not strictly 3D, was Kai's Power Tools.  Remember them?  Or Elastic Reality for morphing (sold to Avid and then branded with Softimage where it probably got discontinued).  Also remember Renderworld?  They were the first software that could realistically portray the "calm" surface of an ocean and was used in Waterworld.  That software cost $5000 at the time in 1995 (but then again, so didn't Lightwave).  Now you can do that and more for free with Hot4D.

Ahh...those were the days....err...not really.

Dave