my DC-8 Project

[TheOptimist]

Can I give some feedback/constructive criticism?

Whilst your ambition to construct a well-performing product is admirable, you've certainly got the capacity to use a few more polys!

Performance in FSX isn't particularly related to poly numbers (unless you get to several hundred thousand). Texturing/materials/drawcalls/systems (in terms of MFD's etc) is where performance issues start to arise.

By the looks of the wiremesh you posted below, I'd anticipate you could at least triple your coly count before any noticeable effect on performance.

Whack a few mesh smooths all over the place. It won't have any effect on performance and will look 10x better.

 

[Mr.FaosFX]

Can I give some feedback/constructive criticism?

Whilst your ambition to construct a well-performing product is admirable, you've certainly got the capacity to use a few more polys!

Performance in FSX isn't particularly related to poly numbers (unless you get to several hundred thousand). Texturing/materials/drawcalls/systems (in terms of MFD's etc) is where performance issues start to arise.

By the looks of the wiremesh you posted below, I'd anticipate you could at least triple your coly count before any noticeable effect on performance.

Whack a few mesh smooths all over the place. It won't have any effect on performance and will look 10x better.

Thanks but that isn't my style; MeshSmooth is the easy way out in some cases but not entirely but it is not good practice to throw a meshsmooth on a surface that already has lots of polys only to attempt to achieve a smoother looking surface, when you can't figure out how to make the mesh look smooth. FSX is already horrid in performance, so why even stress the end users system further by not caring?, I only use the amount of polys I need and heavily research/tweak so much in a way that throwing a mesh-smooth even on the fuselage will have absolutely no change visual detail other than just blowing up the polys, throwing a mesh-smooth at certain places will only have minimal results, as I've already tried this for experiments ... working with models that are VERY DENSE in polygons is very difficult and time consuming.

I think in terms of the end user who wants a great looking plane & good performance. While I understand your view on detail and this is just my constructive critic, I see no benefit and I would ALWAYS rather have full control of my mesh rather than letting a modifier do all the work for me, hopefully I have not intimidated you or offended you in anyway. I'm just pointing out my doings from my 9 years of 3d modeling.

I did a few tests and the only benefit gained is higher poly count and I think everyone will agree.

>>> DC8 fuselage with MeshSmooth thrown on top of the stack.

>>> DC8 fuselage with MeshSmooth removed.

Do you see anything that is different? Hardly not because I've already used polygons needed to achieve a nice looking mesh. You can see almost no benefit of throwing a meshsmooth on top of the stack other than just blowing up your polycount.

MeshSmooth is subdivision modeling used to form meshes from VERY LOW POLY models, it should never be used to smooth a surface for flight simulator because it creates more damage than it strives to fix and you only end up with excess polygons, this is really bad for business. In FlightSimulator we can use a smooth modifier to achieve the same results using lower poly counts, if your going to use MeshSmooth you might as well forget about smoothing groups. Here is some more info >>> http://en.wikipedia.org/wiki/Subdivision_surface

To finalize, thanks for your advice but I would never use MeshSmooth, on any of my projects! :-]

Here is a comment from a very talented 3d artist...

If you're using meshsmooth (that is, subdivisions) you will get unnecessary polys in your model. If you're doing organic modeling work for anything but games, this is acceptable, and even necessary. Just using smoothing groups probably won't give you the level of detail that you need for film work.
They're right though, I can't imagine any cases where you'd want to use subdivisions for game models. It's getting to the point where you can work with a lot of polys, but you're probably not going to want to have as many polys as a mesh smooth iteration creates in a game model.

Don't completely rule out subdivisions. There are a lot of cases where smoothing groups simply won't cut it. But if you're doing models for games, you still have to learn how to use smoothing groups.
Keep in mind though that mesh smooth isn't always the 'easy way out.' There's more to it than just throwing the modifier on top of your object and calling it good. You have to learn how to predict the way that the smoothing works with your model, how to control creasing, you have to have a good understanding of edge-loop modeling, and how to create a complex mesh with as few non-quad polys as you can get. Personally, I'd say that there's a lot more to subdivision modeling than there is to using smoothing groups to get the look you need. It's hardly the "easy way out" that these students are suggesting it is. Well, unless you don't care about creasing and shading errors... :shrug:

 

[TheOptimist]

Whilst I understand you don't like to use MeshSmooths, they're a perfectly valid tool as long as you know how to use and apply them. They certainly don't take away any control over your mesh, assuming you use them properly.

There's little point in me posting a rebuttal to every point, as you sound very set in your ways, which is no bad thing.

However, when I suggested subdivision it was not a suggestion for the fuselage, but the landing gear assembly in particular. Those tires are several of the struts are very low poly - unnecessarily low, you might argue, given the nil effect some subdivision would have on the performance. No amount of clever work with smoothing groups is going to help with those! The smooth modifier certainly doesn't do the same job.

Of course you can't just apply a subdivision to a whole part, the relevant planes/vertices/edges have to be carefully selected, but given the time and understanding it will produce results much cleaner and much quicker than can be done manually.

 

[Mr.FaosFX]

Whilst I understand you don't like to use MeshSmooths, they're a perfectly valid tool as long as you know how to use and apply them. They certainly don't take away any control over your mesh, assuming you use them properly.

There's little point in me posting a rebuttal to every point, as you sound very set in your ways, which is no bad thing.

However, when I suggested subdivision it was not a suggestion for the fuselage, but the landing gear assembly in particular. Those tires are several of the struts are very low poly - unnecessarily low, you might argue, given the nil effect some subdivision would have on the performance. No amount of clever work with smoothing groups is going to help with those! The smooth modifier certainly doesn't do the same job.

Of course you can't just apply a subdivision to a whole part, the relevant planes/vertices/edges have to be carefully selected, but given the time and understanding it will produce results much cleaner and much quicker than can be done manually.

That is because I intended for them too be low because the part is so small, if I modeled the DC8 landing gears to complete perfection and I could, I would create an assembly that easily encompasses at least 38,000 polygons on just that assembly alone, when you take into account both of the assemblies, that's easily 78,000 polys on just landing gears

Anyway, I am done with the re-modeling work and just need to animate...

I appreciate your feedback but while I am trying to reduce the polycount, you are swaying me to do the complete opposite =/ If this was an aircraft of much smaller scale, I would invest in details but certainly not with subdivision.

I'm going to be late for work and must be on my way.

 

[Mr.FaosFX]

Update!

Everyone, I need your input on this, it will decide the fate of the current hydraulic wire assembly and its implementation into the DC8 model. I experimented with this over the last few days and it works well in FlightSim but, it looks some what ambitious and unrealistic but, that is just my view.

To animate the compression on the main landing gears, the hydraulic wire assembly is composed of sliding tubes that each slide in place to give the illusion that the wire is stretching as well, here is what it looks like...

Full compression of the gears, slider tubes are all neatly tucked inside. I know the torque strut is not animated, just pointing that out. [the mess on the left]

No compression, shock strut is fully out of cylinder, all slider tubes now visible, they slide out in progression as the compression loosens.

So what do you guys think? The hydraulic wire assembly looks pretty neat and I would love to keep it but the idea of how it works, seems un-realistic, why? Well, because if you look at DC8 photos, you probably won't see something like this.

The landing gear looks equally well with out the wire assembly so, if I get enough votes to dismiss it, I'll ditch it and stick with no wires. HJG you guys pitch in as well and leave your comments on our private saloon or here where ever you wish, everyone keeping track of the DC8, all pitch in.

If FlightSim had the same kind of physics as Half-Life 2 implemented, we could have such nice wire assemblies and better animations. Wires would dangle in mid air like in Half-Life 2 and animations would be more realistic, as you could add 'weight' to parts but this isn't Half-Life 2, lol. If you'eve ever played Garry's Mod, you'll know what I'm talking about.

Later guys, gotta go eat now.

 

[n4gix]

Given the limitations of FS9, that seems a reasonable compromise.

In FSX it would be a trivial exercise to have the hydraulic hoses fully animated using skinned-bones.

 

[hairyspin]

Looks great, and is probably never going to be noticed on such a big bird! I'm looking forward to seeing the gear fully animated.

 

[Mr.FaosFX]

I'll keep it then, I didn't do all that work for nothing. The front of that area has a cage that keeps the wires back, i'll throw that in and finish up work on the landing gears.

EDIT:

LOL I threw the cage thingy in and had to re-organize the wire assembly but now, I need modify the animation =P I still have time to get this done, hopefully with a preview of it.

 

[Michael2]

Found some landing gear pictures today. You may not need them now, but here they are anyway. I've been looking for a long time -- they aren't easy to find.

http://aircraftwalkaround.hobbyvista.com/airliner_dc-8/airliner_dc-8.htm

 

[Mr.FaosFX]

Found some landing gear pictures today. You may not need them now, but here they are anyway. I've been looking for a long time -- they aren't easy to find.

http://aircraftwalkaround.hobbyvista.com/airliner_dc-8/airliner_dc-8.htm

Thanks but I actually had these already :-] Thanks for taking the time to look tho, your efforts are appreciated.

 

[Michael2]

I'm afraid I can't claim a total lack of self-interest -- I am also working on a DC-8.

 

[Mr.FaosFX]

I'm afraid I can't claim a total lack of self-interest -- I am also working on a DC-8.

That is awesome, I can't wait to see your model! They say the more the merrier.

 

[Mr.FaosFX]

It is amazing how time has passed since I started this project and it only seems like yesterday!

- September
- October
- November

And now, it is December ... so that is 3 months of development and I'm still not done with the external model Hopefully this thread will show others how long it really takes to build a project from scratch, in a way. I've been meaning to take all of the pictures I posted over the days and compile them into a video but I just don't have the time at all for this.

I'm almost done with the gear guys, just fixing some stuff up and then a very detailed review will follow on the website I posted, updates to that website will be weekly, here at FSDeveloper I post daily updates.

 

[hairyspin]

Did you include this photo in your research??

 

[Mr.FaosFX]

LOL I have that picture as well and it demonstrates how the DC8 can deploy reverse thrust in mid-air something which I plan to do Will require some XML coding and other tid-bits but I believe it can be done...as for the fly oh boy...that man must have been crazy!

 

[Mr.FaosFX]

Just what exactly did I change on the main landing gear assembly?

Left :: Old | Right :: New/Improved

How many modifications/additions can you spot?

 

[Mr.FaosFX]

Update!

Alright, going to animate the gears now. I am going to redo the entire animation of parts as I lost some keys during the remodel proccess. Because I modified parts in the gear assembly I may have to modify the mesh ever so slightly or signicantly to get the new assembly to enter the wheelwells with out anything stick out the top of the wing, etc.

Be back with some updates later.

 

[Mr.FaosFX]

Update!

Moving the main shaft into maximum stow position, the apparent issues are visible in the picture bellow. The issue is the door panel sticking out some distance from the wing, the
one I added the struts to...

This is because of the door now feature struts that connect to the main shaft, hold the door at an angle, this is how it is on the DC8, so how will go about fixing this?

Hmm well, I'll have to modify the wing root area and wing to some extend while observing DC8 photographs to keep as accurate as possible. The wing root is actually bigger at the bottom but I was too lazy to stretch it down when I did the older main gears so now, it came back to haunt me lol.

This is where it gets fun for me ... as I adjust the wing root and root of the wings, I need to continue making slight adjustments to the gear assembly specially the door placement in the stow position.

When I've got everything in working order, i'll post another picture of the modifications.

I think I also might have made the tires wider by mistake, observing the previous picture, that seems to be the case, was not aware I did that! oops.

Later guys.

 

[Xpand]

Hey!
One question: How do you make the animation for the nose steering hydraulics? Like, how do you maintain the hydraulic linked to the rotating piece without the pivots jerking around. I tried to use a look at constrain in mine but it didn't work on FSX. And when I try to key frame it the rotating points never align between frames.
I hope you understand my question.

 

[Mr.FaosFX]

Hey!
One question: How do you make the animation for the nose steering hydraulics? Like, how do you maintain the hydraulic linked to the rotating piece without the pivots jerking around. I tried to use a look at constrain in mine but it didn't work on FSX. And when I try to key frame it the rotating points never align between frames.
I hope you understand my question.

The hydraulic shafts are linked to the rotating parts and the cylinders are linked to l_gear, however, they pivot independent of themselves, when I did the animation, I keyed it so that the shafts always align with the cylinders tweaking both for precise alignment...I'm not sure if I'm making any sense.

Also, for the animation to work the parts copy the animation of the RUDDER PEDAL POSITION; from 0 to 50 is deflected left from 50 to 100 is deflected right...if this is not done, the animation doesn't work and you have to copy that animation for every part in the assembly that is supposed to move.

The concept is easy but some what hard to explain ... I'll upload a video on my channel of how I did the whole thing step by step......its a very delicate animation that has to be done in stages otherwise it becomes very difficult to work with with it takes into account gear extension/retraction as well as compression and pivoting, it took me days to rig the whole thing up to work properly.

If your problem is that they wont align between frames, re-align them ... that is what I had to do until it looked satisfactory to me...