|
AMD processors are definitely great options, my last upgrade was to a Threadripper 1950x in March 2019. I remember after installing feeling quite impressed with the speed, its score on passmark is 26,543 which is only a little more than half as powerful as the Ryzen 5950x at 46,392. BonoMan posted:I just ordered a system from Puget so I'm super excited. 5950x, 3080 and 64 gigs ram. Are they competitively priced? I've been building my computers for two decades now but sometimes I just get tired of having to organize all the parts I want, checking compatibility etc. But I can't stand dealing with the Dell/HP route where you then have to do a clean reinstall to get rid of all their proprietary garbage.
|
#
?
Dec 14, 2020 00:17
|
|
|
|
| # ? May 31, 2024 09:49 |
|
|
BonoMan posted:I just ordered a system from Puget so I'm super excited. 5950x, 3080 and 64 gigs ram. 🔥
|
|
#
?
Dec 14, 2020 00:23
|
|
|
Listerine posted:AMD processors are definitely great options, my last upgrade was to a Threadripper 1950x in March 2019. I remember after installing feeling quite impressed with the speed, its score on passmark is 26,543 which is only a little more than half as powerful as the Ryzen 5950x at 46,392. I feel like they're competitively priced. I've built my machines all in the past as well... For my whole post production department as well! And when it came time for me to upgrade this year (I did upgrade my graphics card, but my base machine is from 2013 eek) I decided I had had enough of doing it myself for the time being. I always use puget for their research and wanted to patronize them for it. I love their articles/videos and just how specific they get for my use cases.Also they have access to some supply lines I don't. So, especially for this year, I felt like the upcharge was worth it for *not* having to spend my time trying to find a 3080 and 5950x. Plus... A free beanie! Edit: I should note this is for work so it's easier for me to justify the upcharge since I'm not spending it. But seriously puget deserves it!
|
|
#
?
Dec 14, 2020 01:55
|
|
|
BonoMan posted:I feel like they're competitively priced. I've built my machines all in the past as well... For my whole post production department as well! Curious if you actually have numbers on what percentage you're paying over buying the parts separately? Certainly agree that it's probably money well spent though compared to the trouble trying to track down a 3080 on your own...
|
|
#
?
Dec 14, 2020 02:09
|
|
|
Listerine posted:Curious if you actually have numbers on what percentage you're paying over buying the parts separately? Certainly agree that it's probably money well spent though compared to the trouble trying to track down a 3080 on your own... And at this point the 5950x is just as hard to get too! It's still a 6 week wait to get both of them (or "up to 6 weeks"... Could be sooner). I did do a parts vs purchase comparison a couple of months ago but ended up getting a different config. I would have saved roughly $750 if I remember correctly. And if it was my own money and not through work I might do the parts route. Even for this purchase I still chose to buy a few parts myself. The monitors and M.2 ssds. Saved me a couple hundos.
|
|
#
?
Dec 14, 2020 02:14
|
|
|
Are M2 SSDs noticeably faster than regular SSDs? I only just found out they exist. I guess I'm wondering most about whether it would help with playback of cached files in Houdini.
|
|
#
?
Dec 14, 2020 08:50
|
|
|
This is my new m2 vs few years old SATA both shamesung 970s   can I feel this difference? barely
|
|
#
?
Dec 14, 2020 09:53
|
|
|
500 posted:Are M2 SSDs noticeably faster than regular SSDs? I only just found out they exist. I guess I'm wondering most about whether it would help with playback of cached files in Houdini. As echinopsis points out: The Pci-e ones can easily hit 2000-3500MB/s, while the PCI-E v4 ones already have some models capable of 7000MB/s reads. (Honestly, just like SSDs, they're mostly limited by the speed cap imposed on them, since the standard only uses 4 PCI-E lanes per pcie m.2) (PCI-E specifically! there are also SATA m.2 drives. It's easy to tell the difference though, anything with read/write speeds >600MBps is a pci-e one.) And they can sustain those speeds pretty well, provided they're doing sequential reads. (Cached files would be a great case for that.) M.2 compatability can be a bit annoying on some mobos and etc, so it's wise to check up on things ahead of time (if your mobo has support/slots). Once PCIE_Gen5 rolls around we'll see them bounce up to 10,000, then 14,000 MB/s pretty quick, but that's a couple years off. 2022 or later.
|
|
#
?
Dec 15, 2020 01:15
|
|
|
SubNat posted:
Yeah for instance on the mobo in my Puget system - it only has 3 main PCiE slots and then 2 m.2 slots (and some standard SATA III slots but whatev). If you use the bottom m.2 slot it also disables the bottom PCIE slot leaving you with only two (1 really since the GPU uses one). Kind of annoying as I wanted a 10gbe card and a mini monitor. oh well.
|
|
#
?
Dec 15, 2020 13:36
|
|
|
Man, what a long year. I haven't posted anything in a while, so just going to image dump a bunch of personal work I've been loving with. My old studio went under over the summer, so most of these are just practice to keep everything fresh while I figure out my next move; lots of material experimentation and vray (I skipped v4!) exploration. Using a lot of maxscript (and now, tyflow I guess) to approach things when I can. I've been doing some vague/light process blogging (over here if anyone's interested), too, which has been a boon to keeping productive, even though it feels absolutely...self-indulgent? Whatever, I'm having fun and staying sane; highly recommended!  material studies; the peppermints are all native max maps   https://vimeo.com/458009528 experimenting with tyflow to simulate the raindrops   mood & material study, tyflow cobwebs   mostly modeling study; got really inspired by an interior in Arch Digest Handiklap fucked around with this message at 15:37 on Dec 15, 2020 |
|
#
?
Dec 15, 2020 15:03
|
|
|
I want to start learning procedural textures and modeling. I've worked with Grasshopper in Rhino, so I am familiar with the concept, but want to move more towards visualization than the manufacturing world I've been in. Would Blender or Houdini Indie/Apprentice be the best place to start? This is going to be all self taught so I'm looking for whatever has more tutorials and resources that explain why you make a choice instead of very specific workflows. e. Was over thinking it and am going to go with the seat time approach. I'll sit down with Houdini free and see how it goes. Industry standard, lots of information out there, etc. etc. ant mouth fucked around with this message at 06:00 on Dec 16, 2020 |
|
#
?
Dec 15, 2020 20:52
|
|
|
Re: Hardware. For a while, I was buying used HP Z8x0 class workstations, heck one of my gaming rigs is the same spec HP Z800 24 core machine I had back at Dreamworks in 2012, just with SSD's and a 1080 GTX. There's no way I'd go intel now, I'm due for an upgrade and I'm bouncing between a thread ripper or a Ryzen 9 setup, mostly leaning towards threadripper and the increased cost because of scalping/shortages on the Ryzen 9/Geforce 30x0 RTX side. I'd like to build my own for a change but I'll probably get prebuilt, I wish HP had a Z series AMD machine, right now I"m looking at the Lenovo P620 which is their AMD-based Thinkstation line. Dell has a Ryzen 9 box with 3080, but its an Alienware machine that has 64 GB of max ram, I'd want 128 minimum. I heard good things about Puget, I'll check them out. I'll have to check Boxx again, I got a Boxx Apexx workstation provided by Work for work-related stuff and it was an I9 extreme edition which beat the pants off my old dual xeon by a factor of 2-ish, and it was a nice quiet system. Traded it in over the holidays so I'll have to see what IT comes up for me in January. /edit Yikes Boxx isn't even remotely competitive with Lenovo or Puget on the Threadripper side of things. Big K of Justice fucked around with this message at 05:41 on Dec 17, 2020 |
|
#
?
Dec 17, 2020 05:30
|
|
|
Handiklap posted:Man, what a long year. I haven't posted anything in a while, so just going to image dump a bunch of personal work I've been loving with. These are really nice! Sorry to hear about the studio going under - hope you're able to find something even better soon. I took a long break from personal work after finishing that 40hr+ image and I was trying to decide what to make my next project. I'm just going to focus on learning some technical things - going to spend a few months doing small 2-3 hour tyflow experiments. i know there are literally a million other people also doing this exact same thing but I don't care. Pushing for twice a week, but once at a minimum. This is poking my way around the VDB tools, never used them before - started with a sphere and subtracted other tyflows from it. It's a single mesh, one render from frame 200. I'd love to do an animated version of the tendrils growing out and the bubbles drifting around but it's slow as poo poo to render. I think I should have made the refraction a little more glossy, it's too sharp and the depth is hard to read.  
|
|
#
?
Dec 17, 2020 18:15
|
|
|
I'm making some paper sheet assets in an Unreal 4 environment rn, and I'm wondering what might be the best method for having the back of a sheet of paper blank? would there be a way i can display something else on the back face of a polygon differently using a shader or should i go the simple but dumb route of just duplicating the poly, flipping the normal and making it super close, then shifting the uv?
|
|
#
?
Dec 19, 2020 19:41
|
|
|
Kanine posted:I'm making some paper sheet assets in an Unreal 4 environment rn, and I'm wondering what might be the best method for having the back of a sheet of paper blank? would there be a way i can display something else on the back face of a polygon differently using a shader or should i go the simple but dumb route of just duplicating the poly, flipping the normal and making it super close, then shifting the uv? Duplicate the faces and flip the normals. It's simple and it'll work reliably.
|
|
#
?
Dec 19, 2020 19:50
|
|
|
Kanine posted:--- would there be a way i can display something else on the back face of a polygon differently using a shader or should i go the simple but dumb route of just duplicating the poly, flipping the normal and making it super close, then shifting the uv? Doing a 2-sided thing is probably the best, so you get proper normals for either side, also means you can use 2 different material IDs instead of 1 more complex one. (And as a bonus if you model the edge, there'll also be a tiny edge to catch the light as well. Paper has thickness in reality, after all.) But there is a way to do what you want: There's a material node called TwoSidedTexturing, which lets you assign different sets for the frontface and backface. ( https://docs.unrealengine.com/en-US/RenderingAndGraphics/Materials/Functions/Reference/Texturing/index.html Aaaall the way at the bottom, second-to-last thing.)
|
|
#
?
Dec 19, 2020 19:52
|
|
|
   someone just showed me this on a discord server im on right after i asked the question in here, super simple solution
|
|
#
?
Dec 19, 2020 19:54
|
|
|
I have a fair amount of friends who are illustrators/concept artists and I'm curious y'all. What are some of your pet peeves when you're handed a piece of concept art to work off of and translate to 3d? What are some don'ts that you would tell students/people starting out in concept art to avoid?
|
|
#
?
Dec 20, 2020 16:35
|
|
|
Kanine posted:I have a fair amount of friends who are illustrators/concept artists and I'm curious y'all. As a general rule, the more angles the concept artist can deliver, the better. The more you leave the 3D artist to guess, the less likely they're going to match the concept and the longer it's going to take them. The general rule I recommend here is that the level of importance of the asset should dictate the number and detail of concepts that are delivered. Main characters should have at least three angles: front, side, and back. Random ped #482 can have just a front concept. An important item/weapon should have at least two angles. A wooden box in the corner of an alley only needs one. Material call outs are also very helpful e.g. this buckle is silver, this jacket is aged leather, etc. This is really just very broad advice, though. Some artists prefer more concepts, some prefer fewer. It's always a good idea to discuss with the other artists what they prefer and need and to keep lines of communication open. More communication is always better.
|
|
#
?
Dec 20, 2020 16:59
|
|
|
Sagebrush posted:Post a picture of what you're imagining and if it's not too complex I can probably knock it out in a few minutes. No charge. Awesome! I'll pm you a few drawings. No worries, it's fairly simple. Apologies for the delay, I didn't think to check the thread till now.
|
|
#
?
Dec 21, 2020 08:35
|
|
|
Handiklap posted:https://vimeo.com/458009528 cubicle gangster posted:I think I should have made the refraction a little more glossy, it's too sharp and the depth is hard to read.
|
|
#
?
Dec 21, 2020 08:54
|
|
|
An acquaintance of mine is going to be teaching a generalized intro to 3d art for games next fall at a community college (he's worked in the industry for about 5 years with a background in environment art and wants to pass on his knowledge.) He's going to be dividing the class (about 20 students) between character, environment, and general 3d art for those unsure. I think this makes a lot of sense. He's doing something interesting which I think is really cool, he's actually going 3d-modeling software agnostic, he's telling his students blender/max/maya/modo/zbrush/etc. are all acceptable and the first week will involve showing his students those 4 pieces of software and having them choose one to focus on. Beyond that he's still going to be using the various substance tools as a focus in regards to material creation. He's also stressing using eevee/marmoset/unreal 4 to be used for final deliverables for the final project as the goal by the end of the semester will be to have one good portfolio piece ready. I've been helping him a bit with essential information and tutorials I wish I'd been giving when I started learning 3d as someone who ended up going into 3d environment art. (I can include links to resources and go into more detail for anyone interested.) Beyond the specifics of 3d art, the other focus he'll be having in second half of the semester is showing his students places to get good feedback to continually improve, network and hopefully help land a junior position within the industry. Polycount along with The Club and Dinusty discord servers seem like good places to start? My two main questions are: What do you wish had been shown to you as a novice that would have made becoming a 3d artist a lot smoother? For specifically 3d character artists, what resources do you think are essential for people going into 3d character art?
|
|
#
?
Dec 22, 2020 12:41
|
|
|
Can I ask if anybody would mind telling me where they share their works? I use instagram for finished works twitter I also post WIP Anyone post to reddit? It's a pretty big audience... Just not sure whaat subreddits. Artstation.. Worth it?
|
|
#
?
Dec 27, 2020 13:04
|
|
|
echinopsis posted:Can I ask if anybody would mind telling me where they share their works? Depends on what your goals are. Artstation is pretty much the standard place for portfolios these days. Instagram is better for building and creating a network. The only art subreddit on Reddit that has much reach is r/Blender.
|
|
#
?
Dec 27, 2020 13:34
|
|
|
bigger picture is just more reach, so more people are exposed to my merch
|
|
#
?
Dec 27, 2020 14:17
|
|
|
If you want more reach just do them all, really. It depends on the kind of art you're making, too. Cartoony (or furry,) character art is bigger on twitter, grittier/abstract stuff and environments on artstation, I think reddit seems to love realism judging by the amount of "I drew this person 1:1" sketches on the frontpage, but I personally hate the idea of posting art there. Plus each one has its own audience building requirements, reddit needs none, twitter is a slow burn to gather followers; not sure about artstation, probably in the middle.
|
|
#
?
Dec 27, 2020 15:02
|
|
|
The good thing about Twitter is that a person with very few followers can still have one of their tweets blow up if the conditions are right. People don't seem to respond well if you're clearly just trying to market yourself though (unless you already have a big audience).
|
|
#
?
Dec 27, 2020 21:35
|
|
|
yeah for sure. people can smell a fake a mile away. slowly getting attention of bigger and bigger cg stars, maybe one day someone will retweet or put on their insta story
|
|
#
?
Dec 28, 2020 09:39
|
|
|
echinopsis posted:Can I ask if anybody would mind telling me where they share their works? i would say instagram, artstation, twitter, everywhere tbh. it really doesnt hurt to just have an art account where you post your stuff on all the major platforms
|
|
#
?
Dec 28, 2020 10:32
|
|
|
i should note though, artstation is mostly comprised of portfolios of other artists, its not necessarily where you're going to get randos seeing your work
|
|
#
?
Dec 28, 2020 10:33
|
|
|
https://www.youtube.com/watch?v=TQDYVwSdzCU seriously if you're using blender you should pick up the uv toolkit 2.0 plugin, its only $19 and well worth the money
|
|
#
?
Dec 28, 2020 19:36
|
|
|
thank you so much, dang that's good
|
|
#
?
Dec 28, 2020 20:47
|
|
|
Kanine posted:https://www.youtube.com/watch?v=TQDYVwSdzCU Which marketplace is it better to buy from, Gumroad or Blender Market?
|
|
#
?
Dec 29, 2020 04:01
|
|
|
personally I think blender market is a better way to "hold" your purchases for grabbing updates from later etc but.. gumroad lets people sell stuff for "free" if they want
|
|
#
?
Dec 29, 2020 10:27
|
|
|
Gumroad take less of a cut but blender market is more invested in funding the development of blender.
|
|
#
?
Dec 29, 2020 10:38
|
|
|
https://twitter.com/MemesCG/status/1344056760391364609 i know this is late but happy holidays yall
|
|
#
?
Dec 30, 2020 00:26
|
|
|
Trying my hand at some procedural modelling in Houdini. Very much inspired by this guy's presentation. I spent a good part of today scrubbing through and taking notes any time he does something interesting. Pretty fun to just change the input geometry and see what new thing comes out.  https://i.imgur.com/JiLTPE9.gifv
|
|
#
?
Jan 3, 2021 08:31
|
|
|
i really need to learn houdini since sculpting rocks is fun but gets really monotonous and i could really use a way to automate the process
|
|
#
?
Jan 3, 2021 08:32
|
|
|
it seems like houdini is becoming a tool that a lot of environment artists should learn? or is it mostly the realm of technical artists
|
|
#
?
Jan 3, 2021 08:32
|
|
|
|
| # ? May 31, 2024 09:49 |
|
|
I personally just enjoy learning it because it's interesting to me. I don't know if I'd recommend it for an artist who was purely looking to speed up their workflow. Sinking months into learning Houdini when you could be studying art instead might not pay off in the way that you hope. Having said that, I think it's probably the coolest piece of software I've ever used. If you're looking to get more technical, definitely give it a try! Houdini Apprentice is free, and has all the features of the full version. https://www.sidefx.com/download/ Also it looks like Blender has some pretty decent procedural tools. I came across this today: https://www.youtube.com/watch?v=CBU80mmKImc
|
|
#
?
Jan 3, 2021 11:40
|
|