AF99 vs. SCASM : The Ultimate Confrontation... well... not really...

[hubbabubba]

The muzzle bore hole is 0.04 in diameter.

[Ivan]

Hello Hubbabubba,

I suspect 0.04 foot diameter probably won't work very well, but I will give it a try anyway when I get the chance. On something this small, doesn't it make better sense to just do it as a texture?

- Ivan.

[hubbabubba]

Hello Ivan,

If you persist working within AF99 limitations, I guess the answer is yes.

Some of the holes and "rivets" of the MG cradle are made the same way. The gun sight blade is only 0.01 foot large... but it split the right-left axis. With AF99, that blade would be as thin, but on one side or the other.

0.01 of a foot is very good, but 0.005 is even better!

[Ivan]

Hi Hubbabubba,

Attached are screenshots from AF99 with the point editor showing the coordinates for the 0.24 Octagon and my solution to the .50 caliber gun muzzle. Yes, the build process corrupted the dimensions a bit, but it still sorta works. Keep in mind that we are discussing a polygon that is only 0.02 in diameter. I had trouble finding it. As you can see the 0.24 Octagon turned out well enough. keep in mind that although rivets can be represented by polygons, that is computationally expensive. If you are going to texture the object anyway, why not put the rivets there? The only reason I can think of is that the smooting process corrupts the image a bit.

Yes, you are correct, SCASM does better, but I still don't think we are giving up much in precision using SCASM. As for quality of model, that is another story entirely. I agree that for a good model, AF99 output NEEDS some SCASM tweaking for things it cannot do and for removing some general stupids from the assembly process.

From a philosophical standpoint, we need to stop somewhere. If 0.01 foot is good and 0.005 foot is better then 0.001 foot must be even better??? AF5 had issues with 0.1 foot resolution which in my opinion made for very poor models, but 0.01 isn't that bad.

- Ivan.

[hubbabubba]

Hello Ivan,

Yeah, 0.01 isn't that bad, but the MG jeep would not exist in its present form if I hadn't used SCASM. I too would like to go to 0.001, but SCASM stops at 0.005 foot. On a philosophical standpoint, as you say, we are only limited by the tools we use; this is where we have to stop. Don't tell me that, if Abacus was to announce a "patch" bringing the precision of AF99 to 0.001 foot, you wouldn't download and install it?

The 0.24 octagon is too "fat" at the 45° angle vertices and the 0.02 octagon in radius is not round at all. Here is a picture showing the barrel from the sleeve to the muzzle. Compare "my" bore hole to yours. As you say, texturing would probably have "blurred" it. And you don't have to search for long to find it.

Attachment 13884

On the left I'm showing the cradle and pintle. I could have textured the hole thing, but it was more expedient to use polygons and, besides that, the jeep already had enough textures. Incidentally, it is much more "computationally expensive" to apply texture than to create polygons. In its notes, Kukushkin , creator of AF5 BAO and AF99, had different SCASM commands run in a loop.

The command "Poly()", which I used for these polygons, was capable of running 460.29 instructions per second (it was probably an old 386...) while the command "ShadedTexPoly()" would only run 25.72 instructions in the same time. All things being equal, I could make more than 17 polygons for the "computational price" of one textured polygon.

But higher precision was only gravy...

[Ivan]

Hello Hubbabubba,

You are right: If there were ANY kind of a patch for AF99, I would probably install it.

You are the builder of the Jeep. If you believe SCASM was necessary for proper construction, who am I to argue? I will point out however that my original argument still holds: if you found that there needed to be minor tweaks for alignment to get a body panel to curve (Jeep? Curve?) correctly, you would have an extremely difficult time fixing things. I also believe that without a visual design package, I could not have done the fine tuning of contours for the recent P-40C.

With few exceptions, I prefer to texture pretty much everything in a model because the reflections and shadows look strange without textures. Besides, How much of the speed differences are due to texturing and how much to shading?

I actually had no luck with an octagon for the muzzle, so I used a small hexagon. I agree with you that even that hexagon doesn't look quite right. Regarding fuzzing due to textuers, I noticed that you have that issue with the rear wheel wells and the effect is much more visible than for the muzzle on a machine gun.

Your comment about the 45 degree vertices of the 0.24 polygon being too fat is interesting. You said *I* have good eyes. I am thinking yours must be much better: The dimension from the screenshot in AF99 is 0.17 feet for the 45 degree vertices. The actual number SHOULD be 0.169705627484 feet. It is too big by just under 0.0003 feet (0.0035 inch) or a bit less than the thickness of two sheets of paper.....

- Ivan.

[hubbabubba]

Your comment about the 45 degree vertices of the 0.24 polygon being too fat is interesting. You said *I* have good eyes. I am thinking yours must be much better: The dimension from the screenshot in AF99 is 0.17 feet for the 45 degree vertices. The actual number SHOULD be 0.169705627484 feet. It is too big by just under 0.0003 feet (0.0035 inch) or a bit less than the thickness of two sheets of paper.....

I was making an aesthetic appreciation here, not a Pythagorean calculation. To settle the matter, I took your "GunMuzzle.jpg", removed the 0.25 octagon to only keep the 0.24 and the little hexagon, made a double of the image and made the octagon yellow. After turning it clockwise by 90°, it was over imposed on the original one;

Attachment 13929

If AF99 was making a perfect octagon, the yellow lines would cover the white lines, it is simply not the case here. The yellow octagon appears to be higher but narrower than the original white one. The 45° angles appear to be more or less at the same distance from the center but, given the wider original figure, they appear also wider. I repeated the experiment making my own 0.24 foot octagon along your method and got exactly the same distortion.

So, it is very possible - certain in fact - that AF99 blueprints are distorted OR that it lies about coordinates.

Which makes me rethink my comments about AF99 being a good CAD...

[Ivan]

Hello Hubbabubba,

I suspect there is more to the visual interface of AF99 than that. I have noticed that sometimes when adjusting a point using the point editor, the apparent movement of that point isn't consistent. I thought this was perhaps it uses metric measurements internally and it was a matter of snapping to points. (In other words, perhaps the SCASM scale conversion happens within AF99.

I have also noticed that when I run CFS, it screws up my display in AF99 and I need to restart AF99 to get a proper parts display. The change is in the aspect ratio; The left/right scale on the screen is doubled so that aircraft start looking long and skinny.

I always thought this was my computer because I also have a minor issue when using AF5Paint.

- Ivan.

[hubbabubba]

This is also why my octagon, that was short by two units right-left, looked OK to me. I had noticed that when the mouse cursor was hovering over the coordinates in the upper-right corner, the picture was stretched horizontally.

I used more or less the same visual trick on my octagon in game (yours is ¾ frontal) and it is not distorted, give or take a few pixels.

I guess we'll have to live with it. But it good to know.

[Ivan]

Hi Hubbabubba,

It wasn't my intent to get a 3/4 frontal screenshot. It was as close as I could get to a full frontal shot. It isn't all that far off considering that this almost cylinder is much longer than it is wide. Perhaps I should have just altered the AIR file to put the cockpit viewpoint right in front of the cylinder. I didn't do it because I have been putting these these test cases together in a few minutes and taking screen shots just before I leave on my commute to work.

Do you remember that way back when you were building your Taifun, I was commenting that textures for edges were not as good as polygons? Seems like our positions in this case are somewhat reversed when discussing the Machine Gun Muzzle. I suppose that with the .50 M2HB on a pintle mount, a proper muzzle is needed for "The Look", but for an airplane model, I generally don't even try to put a muzzle on any of the guns. Most of the time, I have even left off the gun barrels themselves.

- Ivan.

[Ivan]

Hello All,

Considering the discussion about representing the muzzle of a .50 caliber machine gun, consider how well such an object would show up in the simulator. The Spitfire's cannon are 20 mm or 0.787 inch and the Corsair uses .50 caliber guns. Muzzles of neither type of gun are typically visible in these views.

- Ivan.

[hubbabubba]

Hello All,

Considering the discussion about representing the muzzle of a .50 caliber machine gun, consider how well such an object would show up in the simulator. The Spitfire's cannon are 20 mm or 0.787 inch and the Corsair uses .50 caliber guns. Muzzles of neither type of gun are typically visible in these views.

- Ivan.

Hi Ivan,

As long as the muzzle is not seen directly in the external view, like the "Ma Deuce" jeep, it can be omitted. But what about a Bf109 with a motorkanone? For many Allied pilots, it was the last thing they could remember before being deported to the POW camps. The less fortunate would probably have said the same if they had survived.

[Ivan]

Is it the opening in the spinner which is only the blast tube or is it the muzzle of the cannon which can't be seen about 5 feet behind the spinner or is it the "Flush" (DPED term) that they are seeing? Or is it just a black dot inside a spiralling spinner which was typical of Luftwaffe aircraft of the period? Did they say the same about the FW 190A which didn't have Motorkanonen?

The FW 190A would have been a more common interceptor and the more dangerous one to bombers.

By my understanding, the 20 mm MG151/20 was a more common motor cannon than the MK108 even though the MK108 was more destructive. Also, from the characteristics of the guns, the MG151/20 should have put out a larger muzzle flush than the MK108. The MK108 was a very low velocity gun with very little propellant and a big A$$ shell which blew rather large holes in aircraft structure.

Any details to your assertion regarding bomber crew? If there were any statistics to be collected, the meticulous Luftwaffe would probably have done it.

BTW, a Me 109K is in my plans and yes, it MUST have an opening in the spinner for the cannon. Whether this is visible in combat is debatable though.... As designers, we all choose what we represent. In the case of the 109, I don't think I will bother with the aileron balance weights.

- Ivan.

[hubbabubba]

LOL!

I have been using DPEd for ages and I never remarked that it was spelled "Flush"! I think it should read "Flash", but who am I to judge? LOL again!

I was talking of the opening in the spinner, of course. What I was trying to explain is that they are circumstances where "detailing" is more important, like the motorkanone that the simmer will inevitably see from quite close quarter if he he goes front-back-front in external views. I remember reading an account by, I think, Gentile in which he was trying to shake-off a "Gustaf" by going vertical, the 109 on his tail. He could see the spinner cannon, probably in his rear mirror, coming closer and closer as his follower was climbing. I'm not convinced that Gentile, or whoever it was, debated the merits of the Mk 108 vs. MG 151/20 while its servant was trying to aim for a no-deflection shot.

Any details to your assertion regarding bomber crew?

Which assertion? I make so many...

In the case of the 109, I don't think I will bother with the aileron balance weights.

With the Taifun, I did bother. And I think I would still bother now that I'm SCASMing my projects.

[Ivan]

You actually wrote "Pilots". I interpreted this to be "Bomber Crew" because for the most part, fighter pilots who get shot down never see their attacker. A dogfight isn't quite as common.

The typical Luftwaffe fighter of the time had a black spiral on the spinner or al all black spinner or a black and white spinner. I am guessing that in combat, most of the time the muzzle could not be seen. Gentile and the typical fighter pilot had excellent eyesight, but I am guessing that sometimes it was just imagination.

Imagine what would have happened had Gentile fought a G-10 Gustav or a K-4 Konig? The result may not have been the same against a superior fighter.

- Ivan.