Image not available

584x326

1725427177986634.jpg

๐Ÿงต unpopular /3/ opinions

Anonymous No. 997298

The truth is, AgX is good, its damn good, but any tonemapper or colorspace isn't going to make or break you when it comes to what really counts which is animation.

The truth is, there are avenues out there that AI art has been totally unable to crack.

The truth is, maya is just as necessary as houdini in VFX.

Anonymous No. 997326

the first agx > srgb / rec709 transform was (and remains) too desaturated and flat for its own good, hence the updates imo.
the image was so lifeless that users, even experienced ones, would be pushed into breaking their renders to see acceptable results in the ipr.
working in ACES during lookdev was always a better experience and ACES' shortcomings could always be compensated in the grade - if anything, you had to massage the image less than you did with an agx render.

Anonymous No. 997328

>>997298
What avenues has AI cracked? Every time I'm presented with something "amazing" AI can do it looks like dogshit. OR the style is extremely repetitive and obvious. Even if you like it your stuff will look exactly like Vikrum Patel's Fiver profile.

Anonymous No. 997329

>>997298
> The truth is, maya is just as necessary as houdini in VFX.
Are there some strong features of Houdini that can't be easily enough replicated in Blender's nodes + procedural texture software like Substance Designer or Material Maker.

Anonymous No. 997331

>>997329
lol

Anonymous No. 997332

>>997331
Honest question. Is it safe to assume it's in principle just the same as Blender nodes but better?

Anonymous No. 997343

>>997332
Nta but superficially, yes; every node in geonodes is 'inspired' by a similar node in Houdini. The real answer is that Houdini is not like the any of the other DCC's, but it takes a few years of using it to really understand why. I encourage everyone to try Houdini, it is mind expanding software.

Anonymous No. 997350

>>997298
The real truth, is that animation will never reach the next level until deformation is volume based, rather than surface based.

Anonymous No. 997356

>>997350
the truth of the matter is that volume based deformation will never reach its apex until increasingly dynamic yet energy conserving integrators are introduced

Anonymous No. 997357

>>997350
>>997356
the real true truth here is that AI will outpace any developments made here, and any actual development here will be used to drive the AI.

Anonymous No. 997362

>>997357
ai looks terrible, kiddo. Also tons of people have gotten in trouble because of it and how closely it is monitored

Anonymous No. 997448

>>997298
Cris is actually kinda cool.

Anonymous No. 997468

>>997328
Ai has done shit, it's Ai faggots that blew everything out of proportion.
Businesses cannot use it, you can not legally protect it.
Ai it's a toy for faggots that can't grab a pen and start doing anything for themselves

Anonymous No. 997470

>>997448
That's not an unpopular opinion

Anonymous No. 997474

>>997470
unpopular opinion : if you can't draw in 2D on paper without a computer with pencil or pen you wont be able to sculpt in 3d well.

unpopular opinion : if you can't render color in 2D on paper with paint, colored pencil, or markers without a computer (no photoshop or ipad) you wont be able to render and texture in 3d well.

Anonymous No. 997475

>>997474
That's the normie opinion, reality is the opposite

Anonymous No. 997476

>>997475
>reality is the opposite
if you dont understand form, you wont be able to draw it with light and shade in pencil / pen and you wont be able to sculpt well, either in physical sculpting or 3d sculpting.

if you dont understand color, you wont be able to shade on paper / canvas or pixel. Its really that simple.

Anonymous No. 997477

>>997356
What are integrators?

Anonymous No. 997478

>>997477
something in sims that attempt to both simulate and fix instabilities and change the sim behavior because of this.

Anonymous No. 997479

>>997476
bait

Anonymous No. 997481

>>997478
Uh huh... And are they also used for non-sim stuff? Because when I imagine volume based deformation, I don't imagine it done via sim.

Anonymous No. 997492

>>997298
>AgX is good
Khronos is better

Anonymous No. 997510

>>997474
Those are not unpopular options, those are shit takes.
That only works if you make low poly and Handpainted texturing

Anonymous No. 997511

>>997474
>learning how to make a flat thing 3d is easier than making a 3d thing flat
Bait so retarded I had to bite bros...

Anonymous No. 997541

>>997492
khronos gives wildly different results. Vastly different.

Anonymous No. 997844

>>997328
You dont notice the good ai unless someone tells you its ai retard. Its all pixels behind a screen. You only notice the bad ones cause successful ai image/videos arent obvious.

Anonymous No. 997846

>>997362
The few twitter posts you read arent anywhere close to the thousands using it successfully. Stay filtered, makes my job easier.

Anonymous No. 997849

>>997844
>You dont notice the good ai
I do, except there is no good ai

Anonymous No. 998542

Yeah, totally agree.

Anonymous No. 999067

>>997326
ACES is just hollywood slop, it has now become political.

Anonymous No. 999070

blender is fucking awful

Anonymous No. 999080

>>997492
why?

Anonymous No. 999086

>>999080
look at the comparison image on the official blender 4.2 release download page comparing khronos / agx / standard. It's khronos in a landslide.