Junkers Ju-52/3m

[aleatorylamp]

Transparent Cabin

Hello Smilo,
I see. OK. Then I´ll give it a go!
It should really work out OK with the single crew member.
The AT-9 Jeep cockpit had a similar layout, with TWO crew members plus their seat-backs.

Update:
I´ve just got the head and torso in, with the cabin aft-wall, floor and dashboard, and I´ve split glazing and framework components into left and right halves each, and it all fits! I´m now on 148.9%, and there´s still 10 free parts for glue templates to do the glue sequencing. Getting there slowly...

Update 2:
OK, it worked. Now I´m on a cramped 149.6%.
Ivan´s "Conga" has done it again! It was a bit difficult because the cabin is embedded into the fuselage behind the nose. Here´s a screenshot.

Cheers,
Aleatorylamp.

 

[aleatorylamp]

... and with 2 pentagon-shaped heads?

Hello Smilo,
Both crew members CAN be put in,
but with simplified structure cross-sections.

Heads get pentagons instead of octagons, and
torsos have flat instead of slanted fronts and backs.
Perhaps the shapes don´t look so good, but I could
live with it. How about you?

... and maybe the crew shouldn´t have goggles?

We also get Ivan´s "Cockpit Conga" in its full splendour:
In the sequencing list, cockpit elements and glue templates
do their full, fascinating, charismatic dance!

Hello Ivan: I am still in awe about this glue sequencing
technique that you demonstrated with the At-9 Jeep!

Parts count is now at 149.6%...
Cheers,
Aleatorylamp

 

[smilo]

goggles?
we don't need no stinking goggles.
...do we?

of course a person wants more detail.
anyone who spends too much time
looking at the crew will be disappointed.
but, considering the options
of non transparent windows,
or even, no crew at all,
i certainly can and will live with it.
well done, Stephan

although, i am a little upset
there are no paratroopers
looking out the side windows.
...just kidding.

 

[aleatorylamp]

Hello Smilo,
I suppose only the gunners would be needing goggles out in freezing open,
so I´ll exchange the bitmap and use the one Ivan supplied some time ago.
While I´m at it, I´ll fix up the dashboard bitmap a bit too - some of the
instruments are cut off by the slanted cabin sides.

I just finished a preliminary SCASM correction of the Virtual Cockpit, to check
how it works. Thanks to the simplified structures that had to be used, perhaps
the co-pilot looks a bit scrawny, so I may add a proper torso and head to the
end of the SCASM listing to correct that shape. We´ll see how it goes.

Then I´ll see if I can get him to turn his head with the rudder like I had on the
Dornier, although that head wasn´t an added one at the end of the listing - it
was just called again for the VC view.

For the moment, at any rate, it´s turning out better than before.

Update: OK, I got a better shaped co-pilot into the Virtual Cabin view, and
the head turns with the rudder!

It helps a lot - or rather, it was the only way I could do it - to identify and
get the correct SCASM code for elements to be added, by running individual
elements through SCASM, elements which have previously been compiled with
AF99 and animated with Aircraft Animator.

SCASM (even at this light level) requires an aging brain to be properly plugged
into concentration mode with logic supercharging to track down all the label
numbers for all those calls and jumps! Where´s my meths?!

I´m just fixing the textures a bit and then I´ll post another screenshot. ... or
better, methinks I´ll upload the machine then, because it´ll be finished!

Paratroopers in the side windows!! ...I know you were joking, but I actually
thought it might not be a bad idea - not exactly to have them clearly looking out
the windows, grinning and waving, but perhaps showing them as shadows wouldn´t
be too distracting. It would have to be done rather artistically, otherwise it would
look quite silly, I suppose.

Cheers,
Aleatorylamp

 

[aleatorylamp]

SCASMing the Crew External view too.

Hello Ivan,
This would be a technical SCASM-related question,
about the best way to go about having the cockpit
crew in both Virtual Cockpit and External View Modes.

There are insufficient resources for AF99 to make a
decently shaped crew for the transparent cockpit in
external view.

For the moment I´ve resorted to simplified pentagon
cross-section head structures, which look insect-like,
and non-slanting, vertical dome shaped torso structures,
which have too thin a profile. (They would be OK, I
suppose, but perhaps there is SCASMing option here...).

With SCASM, I know how to include a decently shaped
co-pilot´s head and torso structures into the virtual
cockpit (octagon cross-section head and dome cross-section
torso slanting fore and aft), including a rudder-triggered
turning head.

I know the routine is added to the end of the SCASM listing
and called by a CALL32 from a Cockpit Section which is
placed before the marker *** Start of Main Aircraft Code *** .

So, now I was thinking of using SCASM to substitute the
simplified crew shapes in the external view with
decently-shaped crew-member shapes, so here I´d have some
questions for you if possible, for the event that you could
have the time and energy to comment on.

There could be a second routine at the end of the SCASM
listing for the pilot (head and torso), and for the external
view, it would be called together with the already existing
co-pilot routine.

Now the question is where to call it from and how to do it.
Identifying where in the listing the simplified crew shapes
are done should be no problem.

Then, I suppose and that just after the labels, a CALL or
CALL32 would be inserted, to call the new pilot and co-pilot
head and torso routines. So four calls would do the trick,
and they would be done in the glue-sequence that is pre-defined
by AF99 in the same order of the existing SCASM listing.

But: My question is, what to do after the calls to prevent the
simplified heads and torsos from being drawn. Do I simply delete
the lines, or is that too risky and it´s better have a kind of
"go to line X" instruction?

Update: Or maybe it´s better to overwrite and substitute the code of the
simplified heads and torsos with that of the new ones, retaining the original
labels.

Thanks in advance for your much appreciated advice, if you were
to have the time and energy for it!

Cheers,
Aleatorylamp

 

[smilo]

whew...and you thought ad2k was too complicated.
(sorry, i couldn't help myself)
it will be very cool, if you can figure out a procedure.
if so, it's something you can use on all your models.

you might want to hold off on that release for a bit.

thinking about paratroopers waving and grinning
made me smile....too funny.
subtle shadows of their backs is intriguing, though.

 

[aleatorylamp]

Hello Smilo,
Hello Ivan, too!

Ha ha! Well, for the way my brain seems to work, (or not to work - that could be debatable!), AD2K seems more difficult to me, but I suppose it depends on the person!

Anyway, I couldn´t wait for a reply and decided to jump in at the deep end and start experimenting with this (for me) more complicated idea which was burning in my finger tips!

It turns out it works! I substituted the SCASM code of the simplified heads and torsos with that of the properly shaped ones. I expect that the technique will work for the other models too!

The longer code off course caused instances in several places of the listing, of Jump instructions which were out of range and thwarted SCASM compilation, but I remembered what Ivan had said about long Jumps, and just substituted them one after the other for Jump32´s, hoping for the best. I was totally surprised when I saw that it worked!

The technique is really quite straight forward, but requires concentration and attention with the code, which is all numbers and letters of course... The elements to be substituted were quite easy to identify because of the texture numbers they had, and the new code was easily obtained by just building separate crew members alone and SCASMing those by themselves. So it was all possible!

All very pleasing.
Now instead of changing the textures (improved dashboard and no goggles) with AF99 and having to do all the SCASM work again for Ju-52 Transport, I´ll just make new textures bitmaps and put them into the texture folder!

The idea of subtle shadows of paratroopers´ backs behind the windows deserves a try too. I realize they are sitting on benches with their backs to the windows, and maybe one window could have the silouhette of one person, and in another it could be offset, and yet another could have partial silhouettes of two, for example. We´ll see.

With the other models that have less margin for parts, I can make 2D crew figures to get the AF99 Glue sequence right, so that they show up in the SCASM listing, and then put in the code for the real crew members.

Anyway, all that Ivan showed me to do is certainly bearing more fruit! Thanks Ivan, again!
Cheers,
Aleatorylamp

 

[smilo]

yes, they do look much better.
somehow, i knew you could do it.
well done, Stephan.

 

[aleatorylamp]

Auntie Ju Paratrooper Transport almost finished.

Hello Smilo,
Thanks! They do look better now, don´t they?
The scrawny insectoids looked more like aliens than a crew.
I suppose preserverance helps there... thanks for your faith!

Then I discovered a minor glitch in the canopy windows.
They all bled momentarily through the cabin body seen from
below - difficult to spot because the wing covers most of it.
Not really serious, but annoying!

I hadn´t declared the window component as smooth, so the
transparency was visible from underneath, but I was able to
correct it just by substituting the cabin window component
SCASM code with the corrected code of smooth one.

Easy, and quite amazing! This way I didn´t have to re-do
the SCASMing on the whole model.

Never a dull moment - more preserverance.
Soon I´ll upload the Unarmed Transport Auntie Ju!
Cheers,
Aleatorylamp

 

[aleatorylamp]

Final Unarmed Transport Version

Hello Folks, hello Smilo,
So far, it´s done! I managed to improve the dashboard the textures,
put a texture on the cabin wall (with SCASM), and also incorporate
the paratroopers´silhouettes behind the windows, and my secretary
is just writing the paperwork for the upload.

Just in case I may have overlooked any problems, here is the final
transport model attached to this post - it´s easier to undertake
corrections before an upload than after. (I´ll have to delete model
attachments to previous posts for attachment cuota reasons).

So, any comments will be very much appreciated!
Here are also a few more pictures!
Cheers,
Aleatorylamp

 

[aleatorylamp]

Now for the CASA 352-L Transparent cockpit

Hello Smilo, hello all!

After the successful additions of crew members into a transparent, external-view
cockpit, as well as a virtual cockpit, to the German Paratrooper Transport, now
comes the turn of the Spanish paratrooper, using the same principle with SCASM.

It´s not so easy this time, because parts count is higher, leaving less room to
manouever.

Then, the armed version is for the moment totally impossible, because there´s no
room for manouevering, but I´ll probably be able to find a way.

The technique for SCASM-adding extra elements for the Spanish version, first
requires simplifying several existing elements in the AF99 build, to make space to
put in the additional ones, also in simplified form, so that the resulting SCASM
listing is complete and includes all glue sequences.

This way I can then identify and pinpoint all those simplified elements in the SCASM
listing, that have to be correctly substituted with the new, fully shaped ones, in
the same places with the same labels.

The new elements are made separately, coded in SCASM separately, and then their
code list is copied into the model listing, overwriting that of the simplified elements.

It seems to be working! The Spanish version with its simplified elements is now slowly
getting the new fully shaped substitutes, one after the other. An interesting process
that requires a lot of attention!

I never thought I´d get this far with SCASM! I bet Ivan will be glad when he reads this!

More tomorrow!
Cheers,
Aleatorylamp

 

[aleatorylamp]

Hello Smilo, Hello Folks,

The Spanish one is done too. The question now would be, do we want paratroopers´silouhettes in the windows on this one as well?

Probably yes, but I´m asking just in case it´s carrying cargo - maybe some silhouettes like the top of boxes or something?

Cheers,
Aleatorylamp

 

[Ivan]

Been a little Busy lately

Hello Aleatorylamp, Smilo,

I had not been checking in on this thread for about a week, so missed the SCASM discussion.
There has been a lot of parent type stuff and trying to keep peace in the house.
We drove my Daughter back to college yesterday. So between a couple appointments for her and getting her settled on grounds, we did not get back until pretty late.
The keeping peace in the house is because until the interim report this morning, my Son was still making a D in history and C's in a couple other classes. The C's are still there, but at least the D is gone. This kind of thing gets Anna Honey VERY upset.

Also been hanging out in another forum and doing a lot of research on the FW 190 fighter for the discussions.

Glad it worked out with SCASM. Your description was pretty good to begin with though I would have done things SLIGHTLY differently.
I would have just found the Pentagon Pilot and put a Call32 at the front of its code followed by an immediate return.
The rest would remain unreachable code. It sounds sloppy, but I prefer leaving things as close to possible as the machine generated version so that it can be recognized if the next model needs to be have some edits from AF99.
It is just a preference. It is not the cleaner method.

If you look at Smilo's Ju 88, I added a pair of FW 190 propellers but did not take out the originals. They remain as unreachable code and can be seen in DPED.

- Ivan.

 

[smilo]

i finally got time to check out the Transport version.
at first, she did not show in game.
i had a look at the Sound.cfg and found the alias call to be off.
i changed alias=FW190Snd to the standard alias=FW190\Sound
and there she was, listed in the Change/Modify Aircraft menu.
i haven't had enough time to wring her out,
but, at first blush, she looks very good.

 

[aleatorylamp]

Hello Smilo, Hello Ivan,

Smilo, Sorry about the incomplete sound alias. It´s because
I have the sounds in separate folders with Minuteman10´s
Sound.cfg. But I´m glad it wasn´t too much of a pain!
Thanks in advance for wanting to try the plane out. It´s just in case
you see something I´ve missed, or have some recommendation!
I much appreciate it, and I know you don´t have too much time,
but there´s no hurry.

Ivan, It´s good when you see the kids improve step by step.
I´m sure they realize it´s worth the effort when they see
how parents are concerned, even at the risk of them coming
along a little heavy. Good job!
So Dped shows unreachable code? Interesting...
Thanks for the indication that the most practical way would have
been to bypass the unused code, but I was getting confused with this.

Thus, I overwrote all the simplified elements with the fully shaped ones
once I saw that it worked. I found it easier, and better this way to oversee
the whole thing as it´s all cleaner. It´s also curious to see, as the listing gets
longer, how some Call´s need changing to CALL32´s for successful compilation.

I had one thing pending on the German transport version - the external
view co-pilot didn´t move his head, because I hadn´t animated it, but
the VC view co-pilot I´d added to the end, was animated.

The annoying thing was that in VC view you could see a bit of the unanimated
head! As I didn´t want to re-do everything all over again, I copied the animation
code from the end to just before the external view co-pilot´s head code, correcting
all the lables of all the conditional calls involved, and changed the VC cockpit call
to the co-pilot´s head so that I could delete the one at the end of the listing.

It took a while and a few tries until it worked correctly, and it ended up clean
and satisfying, and the external view co´pilot turns his head.
I could feel the steam coming out of my ears towards the end!

Anyway, it´s definitely fascinating to see how SCASM and AF99 seem to work in
tandem, and how it puts everything together.

I put comment markers for the new elements into the SCASM code, and make a
back-up of the listing, so that I can use it as reference for the next AF99 model
version to be SCASMed, and I can also copy and paste things from there.

Now, the armed version is also taking shape!
I even reduced the wheels to textured discs to make enough space for the cockpit
and simplified crew element glue sequence, and it´s just under AF99 max.

I also had to further reduce the head pentagon to rectangles, but it worked, and
the AF99 model is ready for SCASMing once the dark camo textures are ready. It´s
painted corrugated aluminium, and this is a very lengthy job.

On the Ju-52/3m, the co-pilot doubled up as front gunner, but I won´t be having him
standing up next to the pilot with his head above the cabin roof, outside behind the gun.
I´d have to build his arms, hips and legs, and I hate to think what kind of a glue sequence
that would need. So, I´ll have him sitting like on the other versions.

I don´t know whether to put the paratroopers´silhouettes on the window textures of this
one either. Well, we´ll see.
Cheers,
Aleatorylamp

 

[Ivan]

Hello Aleatorylamp,

Do you remember seeing screenshots of my Dornier Do 17Z project before I put it on hold?
Each Assembly only had one Group that was a 3D model. The other Groups were all just flat templates.

I think you are now understanding what I was suggesting back then:
Build a representative shape for each AF99 Group.
Build each Group to as much complexity as you wish.
Combine them into one model via SCASM.
I know others have already done this.

This should prove interesting in the long run with the recent programming I have been doing to calculate the VectorJump parameters based on a Glue Part from AF99. If I can make that work, then we don't have to deal with the silly Group Glue sequence that AF99 generates with all viewing planes going through the Origin.
....but to do this, I need a more reliable Development machine.
I am thinking about trying a 5:4 Flat Screen on another old machine and run at 1280 x 1024.
What kind of monitors do you all use?
I just need to clear some space for it and that is not easy.

- Ivan.

 

[aleatorylamp]

Hello Ivan,
I have a 19-inch (6:4) flatscreen that I make to work at 1024 x 768, and for some reason, it was in 16bit mode. I don´t know why, because the only reason was FS98, but that won´t run now with the graphics card I have. I´ve just put it into 32bit mode - AF99 and AA still work...

1280 x 1024 makes things too small. Perhaps a bigger flatscreen would allow the higher resolution. It would be more comfortable because you wouldn´t have to zoom in and out so much with AF99. A bigger monitor on my desk, which is tucked under and next to the stairs, would however be knocked onto the floor by people´s elbows coming down the stairs!

When you were doing the separate Do 17Z builds, I´m afraid that I wasn´t around in the Forum yet, but I understand the method you mean. If you use templates for the reference elements that are to be substituted with SCASM later, the AF99 manouevering space limit problem disappears. It would be much better than the full-build using simplified elements system I´m using.

Doing the glue templates in SCASM could be extremely taxing on the way my brain functions - I have a good built-in 3D graphic booster, but my abstract number/letter interpreter is only generic, with which I can just about manage mediocre Haiku and simple trigonometry.

Thinking about the different ways in which AF99 and SCASM work, it´s apparent that you don´t get to see any AF99 glue elements in SCASM, but you get their result as an ordered sequence of the elements involved.

Looking at the SCASM code for the animated co-pilot´s head, and then correcting labels after copying it to the unanimated one is one thing, but trying to understand it or write a new one myself would dehydrate my brain.

It´s a pity that graphic visualization is absent in SCASM. I daydream about a routine into which you could paste copied SCASM code elements, to get to see them in some order, but it still wouldn´t show glue!

Nevertheless, your suggestion sounds very appetizing, and opens up new ground for building!
Cheers,
Aleatorylamp

 

[aleatorylamp]

Seemingly absurd Jump

Hello Ivan,
I wonder if you would know the origin of a strange SCASM error message.
I also don´t know how to substitute Jump instruction with insufficient
range for a longer range Jump instruction.

The error message says:
-> destination ":L009CF0" out of range (33030), source line 1662
Scasm compilation status: error(s) 1

Line 1662 says:
Jump( :L009CF0 )

Investigating further, I tracked down label :L009CF0.
It appears in line 4930, and is the middle of the listing for all the
ShadedTexPoly parts of the left wing component, and is also called
by a jump in Line 1706.

It looks very absurd, and I felt like just deleting the jump lines and the
seemingly ad-lib placed label. So I just did it and the model compiles and
works perfectly.

What would you rekon could be the cause of this this strange behaviour?
A glitch in my AF99 construction of the left wing, probably...

Hello Smilo,
I know you aren´t so interested in the armed version, or in bombers, but as they existed,
it is probably a good idea to supply one...

Here´s a couple of screenshots of the g4e armed transport in darker camo livery. I still
have to darken and camo the wing textures, and put on the CA+JY registration of the
fuselage of the Ju-52 that came out of the Norwegian lake in 1986.

These could also carry 1100 lbs bombs - i.e. 10 x 100 lb, but they didn´t have the ventral
"stew pot" where the bomb-aimer/ventral machine-gunner sat.
Perhaps they eye-balled the bomb-release moment place to get a general sort accuracy.
I think I´ll leave the bombs in the Dp files for this armed version, so this option can also
be available to simmers.

Cheers,
Aleatorylamp

 

[Ivan]

Hello Aleatorylamp,

I would be very hesitant to do what you just did.
Trying to sort out machine generated code (or what programmers call "Spaghetti Code") is always a chore and I hate it.
I also presume that the compiler (AF99) knows what it is doing unless I KNOW that it clearly is wrong.
Did you possibly add a large block of code between the JUMP instruction and its destination?
If you did add something large enough and the target was barely within range of the JUMP, then it would cause this kind of error.
This is one of the reasons I do minimal modification of the existing code and try to all my new procedures at the end.
That way, when something goes wrong, you can do a quick comparison between the original MdlDisAs generated code and the version you are working with now and see the differences more easily.

These days, even my changes for a Interior Virtual Cockpit are pretty minimal for code size inline.
The Inline stuff is all just a bunch of Call32's to the actual additions at the very end of the original source file.

If you want to retain this JUMP instruction, then you can add a new Label about half way in between the JUMP instruction and its target with a JUMP instruction there to take it the rest of the way.
I believe that right after a Return statement is probably a safe place to add the new Label and JUMP instruction.

Then again, you could be so close to the AF99 limits that there are internal errors now.
Check out my FW 190 Revisited thread to see some silly things that can happen in that case.

On a slightly different note:
The biggest thing holding me up right now is the fact that I cannot get a mouse function to work with my compiled MS VC programs.
If I could, that would open up a lot of possibilities.

- Ivan.

 

[aleatorylamp]

Hello Ivan,
Many thanks for your views! Yes, I had inserted the more fully shaped crew´s heads and torsos, when the "normal Jump" turned into a "too long Jump". Previously I had inserted the Wheel structures that were 2D discs before, but probably before the Jump´s origin. I´ll see.

This Jump right into the middle of the left wing component parts (there´s none in the right wing) must of course have already been present in the original AF99 generated code, with 148.9 % parts count, so your AF99 generated error idea will most probably be the case.

I´ll try and simplify some more of the plane, to get the Jump to go away. It wasn´t there on the other two models, whose original AF99 generated code in fact had a higher parts count of 149.6% in each case.
But, there are some differences in the build, so I´ll back-track. I prefer to eliminate the Jump at its origin than try to fix it with an in-between Jump (but thanks for that remedy too!).

I seem to remember that the MS VC Program you are having the mouse problem with, is an Art or a Photo re-touching programme you intend to use for your texture work. What is it exactly so good at, that prevents you from using another more mousable programme?

I use MGI Photosuite 8.5 generally (lines, cut and paste, simple brightening and darkening, and some special effects), Paint Shop Pro 4.15 for screenshots, and Photoscape for colour curves and more complicated contrast things. Then, I have never been able to cope with Photoshop and its layers.

Also, funnily enough, modern freeware Art and Photo Re-touching programmes seemingly don´t give you a
cut-and-paste-into-the-same-window option anymore. You need Payware versions to get that feature!
Crazy...

Cheers,
Aleatorylamp