OK...File sent mate..had a minor surgery last afternoon so had to runoff quickly. Hope this helps out. When you do boolean operation and when you use the late tool don't forget to weld your points, you had a ton floating in that small section alone. when doing pipes for nurnies and such, 12 to 15 sided disc is well enough. Remember that nurnies once painted don't need a large polycount, specialy things you'll never get that close to. NEVER use double sided polygones like i mentionned, instead make sure your polygones are properly aligned.
A Few exemple to illustrate...no need for description realy.it's self explained realy.
Thanks for the help Al and also thanks for the file . . . .
I will check it out and when I go over the mesh again I will take the optimization methods with me . . .
ciao,
Thanks for the help Al and also thanks for the file . . . .
I will check it out and when I go over the mesh again I will take the optimization methods with me . . .
ciao,
time to park airplanes again . . . .
gp
Glad to help when i can mate...and be carfull with those blind spot when backing those jumbo jets.....;)
Double sided polygons are one of hose horrible artefacts which are tough to remove, like 3 point quads, the gits! (and before you ask, yes you can!) 'Unify polygons' found in the detail tab is most useful for co-located polygons!
Band glue is your freind, when you have rows of quad loops, which are redundant, as is 'disolve edges' when you don't. As a personal recommend, don't have large ngons with holes in them, cut them in such a way they are 2 ngons, without a hole (use the connect tool - l). In my experience when you perform subsequent boolean operations and they break. Every now and again polygons get 'broken' and become inverted or go 'black' one simple solution is to triple them and merge them, which usually fixes them, massivelly pointed ngons are the worst for that, so I just do that as a matter of course these days. (shift t - shift z).
I agree with Al and others, you should clean 'unnecessary' polygons, but things like how many rounding polygons you are using are personal choice. GP - if you want I'm happy to share non-lossy optimisation techniques with you, just PM me or IM me
"Personal choice" is just the polite way of saying a bad choice was made. Most rendered errors with polygins are easily avoided with wise use of edge loops, especially when cutting holes in to any polygon.
It has nothing to do with pro v hobby, it makes life easy regardless of why a model is made.
OK...File sent mate..had a minor surgery last afternoon so had to runoff quickly. Hope this helps out. When you do boolean operation and when you use the late tool don't forget to weld your points, you had a ton floating in that small section alone. when doing pipes for nurnies and such, 12 to 15 sided disc is well enough. Remember that nurnies once painted don't need a large polycount, specialy things you'll never get that close to. NEVER use double sided polygones like i mentionned, instead make sure your polygones are properly aligned.
A Few exemple to illustrate...no need for description realy.it's self explained realy.
Excuse me, but don't put words in my mouth, that is not what I meant at all.
Well how many rounding polygons you use is still an optimisation issue no? And arguably one of the biggest arse-hole jobs to fix later.
Use too many, and the model becomes heavier than it needs to be, like anything, you have to use common sense, not just personally saying 12 segments is fine, when 2 or 3 will look exactly the same when rendered.
If you didn't mean it like that fair enough, but it how it reads.
That is a COMPLETELY subjective opinion that comes directly from your JOB and you and every other 'professional' needs to remember that about a hobbyists/non-pro work. ESPECIALLY when it's NOT asked for to begin with.
Yes everyone can learn something new, but they sure as HELL don't need someone/anybody to stick their noses into anyones projects and indirectly call them out for the most juvenile of reasons on par with the kindergarten level "My Dad's better than Your Dad!" argument/taunt.
Step away from the thread, watch to see if he took your advice, and then shrug if he didn't. Help him if he asked for more. But for the love of God stop hounding the guy! It's up to HIM and HIM alone whether he takes your advice or not, and not your's to repeating attempt to shove your viewpoint down his throat.
That is a COMPLETELY subjective opinion that comes directly from your JOB and you and every other 'professional' needs to remember that about a hobbyists/non-pro work. ESPECIALLY when it's NOT asked for to begin with.
Yes everyone can learn something new, but they sure as HELL don't need someone/anybody to stick their noses into anyones projects and indirectly call them out for the most juvenile of reasons on par with the kindergarten level "My Dad's better than Your Dad!" argument/taunt.
Step away from the thread, watch to see if he took your advice, and then shrug if he didn't. Help him if he asked for more. But for the love of God stop hounding the guy! It's up to HIM and HIM alone whether he takes your advice or not, and not your's to repeating attempt to shove your viewpoint down his throat.
Mate, have'nt you notice that you are the only one posting negative crap right now?. SFM is a place of learning as Ace's put it so often. doing an effective model as NOTHING, and i repeat since you don't read to well, NOTHING to do with been a pro or hobbiest. I don't consider myself a "pro", just a lucky bastard who gets to do this for a few $$$. EVERYONE as something to learn, i learn everyday by working with Pro's and Hobbiest. Tobian is a Hobbiest as you call it, yet he's insanly good with Nodes, so i will not esitate a second to ask him for advice on Nodes if I'm stuck with a problem. It's when someone as an attitude like yours that they get stuck for years never improving realy.
If i was the only one on SFM saying 4.5 million polygones is over the top, then SURE...i'de butt out, but dude, if you knew anything about 3D, you're basicaly stop posting that none sens.
That is a COMPLETELY subjective opinion that comes directly from your JOB and you and every other 'professional' needs to remember that about a hobbyists/non-pro work. ESPECIALLY when it's NOT asked for to begin with.
Yes everyone can learn something new, but they sure as HELL don't need someone/anybody to stick their noses into anyones projects and indirectly call them out for the most juvenile of reasons on par with the kindergarten level "My Dad's better than Your Dad!" argument/taunt.
Step away from the thread, watch to see if he took your advice, and then shrug if he didn't. Help him if he asked for more. But for the love of God stop hounding the guy! It's up to HIM and HIM alone whether he takes your advice or not, and not your's to repeating attempt to shove your viewpoint down his throat.
FYI, I was responding to an attitude blip by someone else, NOT the original poster. Maybe you need to step away and recheck things.... but you know what? fine, what **** would I know about anything anyway.
I can't speak as an expert, but with a mesh the size of this one, 4 million polys at this stage of the build does seem excessive. I would suggest that 4-5 million polys when the build is complete and all elements are in-scene together would not be too outlandish, assuming it was detailed to the level of a mesh like Al's Falcon, and window-boxed. The key thing to remember, is that a render with close up elements of the station mesh in it (say a docking sequence, for example) would take advantage of a modular build technique, and only show the mesh elements needed to make the scene.
Either way, as a "detail mesh" of this size, you can either have it all on screen, and lower the apparent detail level (due to the fact the naked eye wouldn't be able to see small vents, switches, rivets etc') or a section of it on screen at a higher detail level. It's a principle similar to detail scaling in modern game engines and in motion picture modelling.
Al's got a good point. I know from experience that if you don't clean this stuff up early it will come back to bite you hard later on. And at that point it's such an enormous pain in the ass to clean up that it pretty much doesn't get done at all - and you're left with a model that takes 5 minutes to render in a frame of animation, rather than one that takes 30 seconds. Remember that Earth-mars transit thing I did a while back? Ended up around 7.5M polys - and that was as absolutely optimized as it could possibly be - and still hold up close. And that monster was 1KM of exposed girders, fuel lines, and complicated nasa rip-off tech. Nightmare. Ended up having to render it in sections! It's absolutely a good idea to get as much of this stuff cleaned up now - long before you do any texturing - rather than have to deal with the pain of reducing Geo, UVs, and no longer instanced geometry all at the same time.
Thanks very much for your compliments al However I do separate non-destructive reductions and destructive one's as a separate thing. If you are going for a production thing then yes, it's excessive, but this is a hobby project. If he wants too many polies for his pipes THEN HE CAN, it's not an optimisation competition! If he wants the hints, he wants them, if not, he doesn't and I can't blame him from staying away, there are some heavyweights weighting in with their opinions, but it's just that: Lets remember this is about HELP not persecution, hence my post was about suggestion and ideas, not just opinion. Calm down people! this is about help, hints and suggestions, not absolutes of opinions, remeber: You can take a horse to water, but you can't make it drink!
I kind of wish we spent half as much time as we spent on all of this, discussing this in a new thread where we could show models and wire, talk about different optimization techniques, tips and 'rules', suggestion, etc. We could show models before and after. Show areas that needed to be adjusted. Show 'optimized' models, and talk about why in specific areas they were built a certain way. I know there is no absolute but there could be some general guidelines for such things that people in 'the industry' use on a daily basis. Now that would be really helpful for people that wanted to learn, and best of all it wouldn't really be connected to any one's specific model.
^ Why not? I think it would be a useful and informative thing to discuss- you could put it in the 3D Q&A Section- and people could post their before and afters, and discuss the different techniques and approaches for modelling high-detail pieces. After all, Al will have approached his Falcon mesh in a different way to Craig's approach of his Veyron, but it is important for both of them to make effective use of the polygons available.
The key is to be intelligent with your poly allowance and distribution- more poly's is not necessarily "bad" or "wrong", when used appropriately- it's more important to ensure that the polygons that you do use are effective, and that I think is an area that many modellers need advice with.
Thanks very much for your compliments al However I do separate non-destructive reductions and destructive one's as a separate thing. If you are going for a production thing then yes, it's excessive, but this is a hobby project. If he wants too many polies for his pipes THEN HE CAN, it's not an optimisation competition! If he wants the hints, he wants them, if not, he doesn't and I can't blame him from staying away, there are some heavyweights weighting in with their opinions, but it's just that: Lets remember this is about HELP not persecution, hence my post was about suggestion and ideas, not just opinion. Calm down people! this is about help, hints and suggestions, not absolutes of opinions, remeber: You can take a horse to water, but you can't make it drink!
You're looking at this the wrong way mate. now one is talking about destructive reduction. As to the pro and hobbiest thing, that's BS and you know it..hehe. a good model is a good model period. Optimization as a much an part of CG as Painting, modeling, surfacing, etc etc. how many people on SFM can model as they want, yet can't paint for a New York Minute!..some for optimization realy. it's not a Choice, it's a skill every modeler needs to learn.
WOW . . . lots of points of view going on in here . . . . cool.
Let me start by saying sorry for pumping up the aggression in this thread . . . I get protective of my work. @ Chrono, Tobian, Thanks for the assistance, and everyone else, thanks for the help and interest with my model. I am assuming everyone here has the best of intentions, otherwise you wouldn't have posted at all . . . .
The funny thing is everyone is right in a way . . . a high poly mesh servers no one any good. How on Earth are you going to use it in a scene Animated or not if the Mesh is too big? good point. . . . . Trust me when DS9 is done I will go back in and cut out unneeded polys.
The station is in it's basic form, when I do full renders I have to merge all the layers, mirror, radial array to get a full section, so going back in to fix things will be very easy.
The other side of the coin is my modeling methodology, as I said I am still learning but I have been doing this long enough to make better choices about SOME things rather than letting them slide.
Personal Choice is not a Bad Choice, you may prefer using 3 poly's in your rounded edges, I prefer using 5 . . .
You know this mesh isn't as out of control as you may think. I gave AL the part of the mesh which had the most concentration of detail in one spot and a lot of it was un needed polys, If you look at the images he posted you can see some of them.
My intention was to get help with this portion because as I said it was almost 1 million polys when you added in all 12 docks. What set me off a little was the way that section was made to reflect the entire mesh as being a mess. . . and that was incorrect.
My bevels uses one poly, this is a normal method . . . my rounded edges use no more than 5 poly's. This is my preference, if you think 3 is fine . . . Bravo, . . .
5 works well for me . . . I use 5 maximum everywhere on the station. The section I gave Al only had 2 rounded edges, one was the main face of the Docking port, and the second was the rubber sealing ring/docking collar, which I believe has 6 polys.
The only other place were there is concentrated detail is the sensor pallet but I constructed that with Optimization in mind so I know there is no waisted poly's there, hell I rebuilt that 3 times . . .
I could go on all morning but I am getting distracted now . . .
How about this . . . . when I do the Optimization on the station I will post the wires and work and get help were needed, plus you will get a chance to see all the so-called mess . . . .
gp
Sorry for the long delay . . . Life got in the way . . .
I started back on this model today and started hashing in the pylons . . . These are the last sections I have to build before Optimizing the mesh and doing some texturing . . anyway in true form of the model here are the pylons basic shape.
gp
I was away for a long while, I think I went to 2 places of employment since the last post, but between work and doing websites I drop 3D altogether.
I will try to get the juices going again and finish this puppy off asap.
more update to come . . .
gp
Posts
every once in a while I get to do something cool like that . . . .:D
gp
A Few exemple to illustrate...no need for description realy.it's self explained realy.
I will check it out and when I go over the mesh again I will take the optimization methods with me . . .
ciao,
time to park airplanes again . . . .
gp
Glad to help when i can mate...and be carfull with those blind spot when backing those jumbo jets.....;)
Band glue is your freind, when you have rows of quad loops, which are redundant, as is 'disolve edges' when you don't. As a personal recommend, don't have large ngons with holes in them, cut them in such a way they are 2 ngons, without a hole (use the connect tool - l). In my experience when you perform subsequent boolean operations and they break. Every now and again polygons get 'broken' and become inverted or go 'black' one simple solution is to triple them and merge them, which usually fixes them, massivelly pointed ngons are the worst for that, so I just do that as a matter of course these days. (shift t - shift z).
I agree with Al and others, you should clean 'unnecessary' polygons, but things like how many rounding polygons you are using are personal choice. GP - if you want I'm happy to share non-lossy optimisation techniques with you, just PM me or IM me
It has nothing to do with pro v hobby, it makes life easy regardless of why a model is made.
nice work, al, hope they surgery went fine.
Well how many rounding polygons you use is still an optimisation issue no? And arguably one of the biggest arse-hole jobs to fix later.
Use too many, and the model becomes heavier than it needs to be, like anything, you have to use common sense, not just personally saying 12 segments is fine, when 2 or 3 will look exactly the same when rendered.
If you didn't mean it like that fair enough, but it how it reads.
Cheers!
That is a COMPLETELY subjective opinion that comes directly from your JOB and you and every other 'professional' needs to remember that about a hobbyists/non-pro work. ESPECIALLY when it's NOT asked for to begin with.
Yes everyone can learn something new, but they sure as HELL don't need someone/anybody to stick their noses into anyones projects and indirectly call them out for the most juvenile of reasons on par with the kindergarten level "My Dad's better than Your Dad!" argument/taunt.
Step away from the thread, watch to see if he took your advice, and then shrug if he didn't. Help him if he asked for more. But for the love of God stop hounding the guy! It's up to HIM and HIM alone whether he takes your advice or not, and not your's to repeating attempt to shove your viewpoint down his throat.
Mate, have'nt you notice that you are the only one posting negative crap right now?. SFM is a place of learning as Ace's put it so often. doing an effective model as NOTHING, and i repeat since you don't read to well, NOTHING to do with been a pro or hobbiest. I don't consider myself a "pro", just a lucky bastard who gets to do this for a few $$$. EVERYONE as something to learn, i learn everyday by working with Pro's and Hobbiest. Tobian is a Hobbiest as you call it, yet he's insanly good with Nodes, so i will not esitate a second to ask him for advice on Nodes if I'm stuck with a problem. It's when someone as an attitude like yours that they get stuck for years never improving realy.
If i was the only one on SFM saying 4.5 million polygones is over the top, then SURE...i'de butt out, but dude, if you knew anything about 3D, you're basicaly stop posting that none sens.
1.85m. And the falcon is built for close-up..it's like 1% the size of DS9
FYI, I was responding to an attitude blip by someone else, NOT the original poster. Maybe you need to step away and recheck things.... but you know what? fine, what **** would I know about anything anyway.
Knock yourselves out.
But he is building DS9 for close up, and like you said, its much, much bigger.
Even for close-up 4.5m is WAY to much. i cleaned up a section and it as 30% of the original polycount and still looks exactly the same.
Either way, as a "detail mesh" of this size, you can either have it all on screen, and lower the apparent detail level (due to the fact the naked eye wouldn't be able to see small vents, switches, rivets etc') or a section of it on screen at a higher detail level. It's a principle similar to detail scaling in modern game engines and in motion picture modelling.
The key is to be intelligent with your poly allowance and distribution- more poly's is not necessarily "bad" or "wrong", when used appropriately- it's more important to ensure that the polygons that you do use are effective, and that I think is an area that many modellers need advice with.
You're looking at this the wrong way mate. now one is talking about destructive reduction. As to the pro and hobbiest thing, that's BS and you know it..hehe. a good model is a good model period. Optimization as a much an part of CG as Painting, modeling, surfacing, etc etc. how many people on SFM can model as they want, yet can't paint for a New York Minute!..some for optimization realy. it's not a Choice, it's a skill every modeler needs to learn.
Let me start by saying sorry for pumping up the aggression in this thread . . . I get protective of my work. @ Chrono, Tobian, Thanks for the assistance, and everyone else, thanks for the help and interest with my model. I am assuming everyone here has the best of intentions, otherwise you wouldn't have posted at all . . . .
The funny thing is everyone is right in a way . . . a high poly mesh servers no one any good. How on Earth are you going to use it in a scene Animated or not if the Mesh is too big? good point. . . . . Trust me when DS9 is done I will go back in and cut out unneeded polys.
The station is in it's basic form, when I do full renders I have to merge all the layers, mirror, radial array to get a full section, so going back in to fix things will be very easy.
The other side of the coin is my modeling methodology, as I said I am still learning but I have been doing this long enough to make better choices about SOME things rather than letting them slide.
Personal Choice is not a Bad Choice, you may prefer using 3 poly's in your rounded edges, I prefer using 5 . . .
You know this mesh isn't as out of control as you may think. I gave AL the part of the mesh which had the most concentration of detail in one spot and a lot of it was un needed polys, If you look at the images he posted you can see some of them.
My intention was to get help with this portion because as I said it was almost 1 million polys when you added in all 12 docks. What set me off a little was the way that section was made to reflect the entire mesh as being a mess. . . and that was incorrect.
My bevels uses one poly, this is a normal method . . . my rounded edges use no more than 5 poly's. This is my preference, if you think 3 is fine . . . Bravo, . . .
5 works well for me . . . I use 5 maximum everywhere on the station. The section I gave Al only had 2 rounded edges, one was the main face of the Docking port, and the second was the rubber sealing ring/docking collar, which I believe has 6 polys.
The only other place were there is concentrated detail is the sensor pallet but I constructed that with Optimization in mind so I know there is no waisted poly's there, hell I rebuilt that 3 times . . .
I could go on all morning but I am getting distracted now . . .
How about this . . . . when I do the Optimization on the station I will post the wires and work and get help were needed, plus you will get a chance to see all the so-called mess . . . .
gp
I started back on this model today and started hashing in the pylons . . . These are the last sections I have to build before Optimizing the mesh and doing some texturing . . anyway in true form of the model here are the pylons basic shape.
gp
I will try to get the juices going again and finish this puppy off asap.
more update to come . . .
gp