[a / b / c / d / e / f / g / gif / h / hr / k / m / o / p / r / s / t / u / v / vg / vr / w / wg] [i / ic] [r9k / s4s / vip / qa] [cm / hm / lgbt / y] [3 / aco / adv / an / asp / bant / biz / cgl / ck / co / diy / fa / fit / gd / hc / his / int / jp / lit / mlp / mu / n / news / out / po / pol / qst / sci / soc / sp / tg / toy / trv / tv / vp / wsg / wsr / x] [Settings] [Home]
Board
Settings Home
/3/ - 3DCG



Thread archived.
You cannot reply anymore.



Idea is the following - I, or anybody else, post a short webm of some simple modeling and you have to try beat the speed of the OP. What you MUST do, though, is to follow the operations EXACTLY like they are represented in the video. Some things could possibly be achieved faster, but that is not the point of this thread. Feel free to practice a bit before posting, I kept it really short and simple, but you'll probably have to recreate the object a few times to get to the proper speed.

Since generating content fast is very important in 3DCG, let's see what programs really are the fastest. Not gonna lie, I'll be using Blender since I want to prove to others (and myself) that it really is super fast, but I'll be glad if you beat me. I have experience with 3ds Max and Maya too, but nothing came close to the speed I can achieve in Blender. I really like it, although modeling is the only part I'm using it for.

I'm pretty sure you beat this with buttons, in my experience I can't get to this speed with shortcuts in Max and Maya either, since they still do some things slower, but maybe more experienced users will know the faster way.

We'll see how it'll go, I recorded 2 short clips which use basic operations we are always using, such as extrude, inset, scale etc. No additional add-ons/plugins allowed, let's keep it vanilla for now.

It will be okay even if you can't record but want to talk about your results, but lying is really pointless, so be honest. Maybe we can also watch other people and provide some tips how to achieve some things faster.

FIRST CHALLENGE (21 seconds):

Just a bunch of extrusions and scales, the only important thing you have pay attention to is at the beginning - when you create the cube, you have to rotate it in z axis by 32.7 degrees. The reason I did this is because you can't use snapping with a number like that so you have to enter it manually.
>>
File: speed2.webm (2.91 MB, 1280x720)
2.91 MB
2.91 MB WEBM
And one more. Start in the front view, then scale to a 4 by 4 grid, position it above x and right from the y axis and proceed. You have to move those vertices on top a bit down and up in z axis, so that you can show how you align them after that (with planar, scale or whatever). At the end bevel in the correct amount like I did.

Forgot to mention you can bevel as much as you like in the first challenge.
>>
>>604666
>rotation in object mode
your model is shit.
>>
>>604693
You completely missed the point, good job!
>>
File: 2018-02-05 20-31-57.webm (586 KB, 1280x720)
586 KB
586 KB WEBM
i tried something similiar.
granted, i haven't used blender in weeks
>>
>>604697
Still fast, gj. Only you didn't scale at the beginning but doesn't matter. And ideally you would have to do the operations in the exact order, which was also done on purpose so I can test your switches to different selection modes.

I mean, the models are obviously ugly and pointless, I didn't even think about it, just improvised something on the spot which uses all the operations. If somebody has a better idea how to make these more interesting, feel free to post.

I'd like to see someone with Modo try this. I need like a minute or two to do these in Maya and Max. People say that Max is the fastest modeling program, but when I see pros like Arrimus, Warwick or Chamferzone model in it, they are still slow. Tor Frick, on the other hand, looks pretty fast with Modo. Admittedly, these guys are doing a tutorial so it's probably not a full speed, but I watched some of their works without talking and it still felt slower than Blender.
>>
>>604701
tor frick might take some time because he uses booleans which takes time to get right.

still, i can't really consider booleans to be in the same realm of traditional modeling.
far as speed goes, i assume blender is faster in every aspect if you uses only hotkeys to do the modeling. but for me what matters is the topology and how the artist treats the model.
>>
>>604703
Sure, but it still matters. I'm currently trying to get familiar with Maya, I'll go through some tuts on how to make your workflow faster. From what I see, you can also model completely with shortcuts in it, and it feels a lot faster than Max, not as fast as Blender, but maybe it could be an ideal balance for me.
>>
>>604693
Why is that a bad thing? Either way you can just press ctrl + A and apply rotation.
>>
File: TiredOfDonutsAndAnvils.webm (2.01 MB, 1440x900)
2.01 MB
2.01 MB WEBM
Here's my take. I'm not a power user, I'm more of an autodesk shill but damn I like how fast Blender is when it comes to modeling. Sorry about the shit quality too, I didn't bother playing around with the quality parameters of the online webm converter.
>>
>>604944
Yeah, people can criticize some other aspects like sculpting or unwrapping, but you have to admit that modeling is on point. It is very comfy once you get a hang of it.

I'd like to see somebody replicating these in other software now, it's clear everyone is pretty fast in blender, even when they are not trying.
>>
>>604948

I tried doing it in Maya and it was twice as long (and I have years of experience working in Maya).
>>
>>604948
desu the unwrapping in blender is not bad. i used to really suck with it and now i can unwrap anything without wasting too much texture space
>>
>>604914
>Why is that a bad thing?
just bad habit.
>>
what are you guys using to record your webms? I have nvidia shadowplay or whatever but direct recording to webm would be nice
>>
>>605017
ShareX. Takes screenshots, records .gif, .webm. mp4, anything. Open source.
>>
File: 2018-02-07_09-36-14.webm (1.35 MB, 1054x756)
1.35 MB
1.35 MB WEBM
>>604666
16 seconds
Actually considered making a thread like this myself. So much software war going on, this puts it to the test.
>>
>>604666
>You HAVE to do it the exact same way
>What is bias
It's a good idea I guess, but your requirements to the challenge are cancer.
You admit that it might be faster with buttons, but then you say "you're not allowed to use them, because that might destroy my narrative that blender is fastest"
Better would be to give an object/shape, and then ask people to make that. (without providing instructions) to see who can get the fastest time.

So I suggest doing the same with this object
https://www.turbosquid.com/3d-models/free-obj-mode-chair/909968
There's the .obj if you want to check the wireframe on the original object.
But the point is you don't see how it was made and it requires a bit more than 10 seconds of modeling. It's a lot less bias.
>>
>>605040
Or basically tl;dr, your challenge was DESIGNED for BLENDER, of course blender is going to do it the fastest. Use a real world example. Faggot.
>>
>>605040
>You admit that it might be faster with buttons, but then you say "you're not allowed to use them, because that might destroy my narrative that blender is fastest"
I didn't say that anywhere, read more carefully.

Your idea is also fine, just takes significantly longer to do, and recorded videos would also probably be too long to post here and nobody would care to waste so much time on that. As I said, you can use buttons or shortcuts, your choice, but the idea was only to test the SPEED of repeated execution of various basic operations. That's why the model doesn't matter, but it still makes sense since it all adds up in the actual modeling process. 5 additional seconds on half of the actions you always use will create significant difference in completion time at the end of a long session, won't it? At least it does for me.

>of course blender is going to do it the fastest. Use a real world example.
How is that different? If - just repeating what you said - Blender is by far the fastest, why wouldn't it also be the fastest in the "real world" example then?
>Oh, but other programs have all these specialized tools to make it faste...
I mean, so does Blender. How to judge that? Maybe we could try some simple challenges you can do in any way you want it to, but focused on a particular thing, e.g. creating a pipe in a shape of "S", or using symmetry to do something basic...

The problem with your challenge though is that it also tests the individual's skill, so it's hard to compare the speed of programs in that situation. Here, everyone is on the same level, since the challenges are basic. Feel free to figure out something like that, but better, on your own and post it here.
>>
>>605052
>I didn't say that
Not directly, but you start off by saying "You have to do it EXACTLY like I did"
I'll admit it's not the same as what I paraphrased, but close enough.
>5 additional seconds on half of the actions you always use will create significant difference in completion time at the end of a long session, won't it? At least it does for me.
That's the point of doing a more complex model.
If you really want to test that theory, you have to do so in a real world example.
If - just repeating what you said - Blender is by far the fastest, why wouldn't it also be the fastest in the "real world" example then?
I never said blender was "by far the fastest" I said your example was designed for blender.
>Here everyone is on the same level, since the challenges are basic
Yeah but you have to admit, the basics is where blender shines.

I really think having a few community challenges would really help /3/ come together.
Posting a single piece of concept art, in which users recreate it.
Voting for an object and then allowing complete artistic freedom in creating said object.
Collaboration threads where an obj is shared back and forth to create a model (We've done this before, was super fun.)
Etc.
I think this would help solve a lot of the /3/ nature of only having 2 or 3 good threads up at a time, with the rest being full of shitposts.
>>
>>605057
I agree we need more challenges in general. And like I said - if you have a better idea, go ahead and post it. I think that chair is a bit too much, though. It has to be something which you can do in 5 minutes tops IMO. I said you can make your own challenges and rules. Mine are something which makes sense to me.

Now...
>Not directly, but you start off by saying "You have to do it EXACTLY like I did"
If you bothered reading other replies you would have known why you have to do it "EXACTLY" like I did. I obviously meant "in that particular order". Do those insets and extrudes with whatever tool you usually would do.
>I never said blender was "by far the fastest"
"of course blender is going to do it the fastest"
And the challenges were not designed for Blender. That is the exact reason why I only used the most basic vanilla operations you can find in every program.
>Yeah but you have to admit, the basics is where blender shines.
True, but I think it is good in general as well. I mean, you can say they are "only the basics", but you do those like 30 times in a minute, so it takes up a huge percentage of one's time when modeling. It adds up.
>>
File: 2018-02-07_07-39-48.webm (2.71 MB, 1920x1058)
2.71 MB
2.71 MB WEBM
>>605065
>if you have a better idea, go ahead and post it
Might make something later.
>has to be something which you can do in 5 minutes tops IMO
Some people on here are hobbyists who don't really know what to do next.
Some people still, don't really know how to get better and need a little bit of handholding when making their model. It's much easier if you're working on the same model as a few other people and can ask how other people did something.

>"of course blender is going to do it the fastest"
"DESIGNED for BLENDER" is extremely important in that.
>And the challenges were not designed for Blender
They are. It's a speed challenge, where only the basic box modeling features are touched.
You picked something that blender is clearly better at, then focused your challenge entirely around it.
It makes a good point though.
>True, but I think it is good in general as well
For sure. Back when I was first getting into 3D, I tried Max/Maya/XSI and Blender. blender made the most sense to me so I just went with it.

I think an informative OP with details on how to record a time lapse video and good details on the ideas of the threads would lead to some pretty interesting challenges/threads.
30~60 minutes is probably best to keep the upper limit to.
Also to keep derail to a minimum, here's mine.
My brain doesn't process shit very fast, so it's rather slow compared to everyone else.
>>
File: 2018-02-07_07-56-54.webm (2.41 MB, 1920x1080)
2.41 MB
2.41 MB WEBM
>>604667
>>605074
And because no-one else has yet, here's the second one.
This is what happens when I try to go fast to keep filesize under 3MB
>>
File: 14625836925.png (285 KB, 400x400)
285 KB
285 KB PNG
>not a single non-blender webm has been posted
Wait, I thought Blender was suppose to be shit? Where's all that industry standard Maya and 3DS Max wizardry I've been hearing so much about on this board?
>>
>>605092
Trade secret. Sharing is for commies & blendlets.
>>
>>605092
It should be obvious by now that the loudest loudmouths usually have the least talent
>>
>>605028
fantastic
>>
>>604666
To whoever blaming the challenge for being biased; here's a new one. Mug with handle.

I don't care how you do it. Bonus points for topo / visuals.
>>
>>605116
>Caring about topo on objects that won't deform at all
>>
>>605126
>can't do it
>>
>>605129
It's just a waste of time.
>>
>>605131
Only if you're slow.
>>
>>605092

I'm the autodesk shill who posted this
>>604944


And I'm starting to turn into a blendlet. I really REALLY like blender for modeling. The sculpting part looks pretty good too, but the performance with high poly stuff needs a LOT of work.
>>
>>605132
Besides, how is it even possible to fuck up the topology of something as simple as cup?
>>
>you have to try beat the speed of the OP.
>What you MUST do, though, is to follow the operations EXACTLY like they are represented in the video
What a fucking retard. This is not a modeling challenge, it's a hand reflex challenge. Make it that you do it how you want but say starting from the same primitive, but not fucking repeat everything you've done, like a retard.
>>
>>605135
It's a challenge which tests the speed of a modeling program, not the speed and skill of a modeler, yes. That was the point.
>>
>>605136
Then what's the fucking point, the challenge is set up for Blender from the start. And if that's the case then why are you comparing time in this thread instead of comparing number of actions needed to make the model?

I am genuinely disguisted with you, OP
>>
>>605209
No comment...
>>
File: mugchallenge.webm (2.95 MB, 600x393)
2.95 MB
2.95 MB WEBM
>>605116
Still waiting for anything else than blender to participate.
>>
>>605230
how do you add the subsurface modifier without clicking on the drop down menu?
is there a shortcut?
>>
>>605230
That's pretty neat.
>>
>>605261
Not him, but CTRL + number (on the main keyboard), in object mode. Number represents the subsurf level.
>>
>>605230
What if… everybody here uses Blender, and nothing else?
>>
>>605271
That would explain all the blender hate threads, just meta trolling.

Also this thread is a little bit stupid, being able to model simplistic crap in seconds is worthless.
Being able to model complex things is minutes is worth while, and no one does not imply the other. Modo's boolean support is vastly superior to blender (yes even with booltools or hardops).

t. Blenderfag
>>
>>605302
Okay, you said it a thousand times already - challenges don't make much sense - so think of something better.

The only imporant thing is that you figure out a challenge which does not take individual's skill into account since I wanted to see what modeling program is the fastest regardless of skill. Like, why are so many people saying how 3ds max is the best and fastest modeling program and now by the replies it seems the blender is a winner without a question? How do you even test that, then? Also, I get it that the title could look a bit misleading.
>>
>>605265
that doesnt work for me, I have emulate numpad turned on and it just switches views.
any way I can make it work without having to turn off the emulate numpad option?
>>
>>605074
Use hotkeys, for fuck's sake
>>
>>604944
Why are you using menus and that wheel? Do you WANT to be slow?
>>
>>605261
How did you add those two modifiers in the beginning?
>>
did it in 18 sec in 3dsmax. Have custom shortcuts on extrude and inset. Too lazy to figure out how webms work so you just gotta take my word. Why would blender be faster? Never used it before
>>
>>605334
>Use hotkeys
No.
As I said, I'm slow to process shit.
Hotkeys only slow me down in some areas.
Oh and, good job quoting the one where I only didn't use a hotkey ONCE. Faggot.
>>
What's the point of speed modeling garbage in garbage software?
>>
>>605399
None. That's why everyone on garbage software gave up beforehand and haven't posted a single video.
>>
>>606070
gg
>>
File: DASH_600_K.webm (1.05 MB, 426x240)
1.05 MB
1.05 MB WEBM
>>605399
>>606070
>>606097
stop samefaggting the thread
>>
>>606103
They're all trolls! It's just one guy! My mom says I'm the best. Right?
>>
>>606103
Based webm. Thx.
>>
>>605028
Tried this for a few days. The UI is impressively disappointing. Its like it intentionally avoids typical design philosophies from every other program in existence. :(
>>
>>606210
Surprised to discover you weren’t talking about Blender.

t. Blenderfag.
>>
>>606210
Set it up, use hotkeys. Never touch settings again.
I don't know why it has like 3 different pages all with different tabs for the settings.
>>
File: fuggg.jpg (89 KB, 562x624)
89 KB
89 KB JPG
>>606103
WHY DOESN'T MINE LOOK LIKE THAT
>>
File: Capture.png (99 KB, 586x754)
99 KB
99 KB PNG
>>606318
because you're unable to follow simple instructions?
make cylinder>ctrl+r then scroll for some loop cuts> tap A twice> W - subdivide> ctrl+e - un-subdivide and change iteractions to 1> tap I twice to insert each face then X to delete faces




Delete Post: [File Only] Style:
[Disable Mobile View / Use Desktop Site]

[Enable Mobile View / Use Mobile Site]

All trademarks and copyrights on this page are owned by their respective parties. Images uploaded are the responsibility of the Poster. Comments are owned by the Poster.