Wednesday, January 27, 2010

F-35 Closing the alpha test

As the current build of the F-35 is now significantly different from the Alpha build, I am closing the Alpha test stage - I would like to thank all the testers for their detailed and competent feedback.
For now, most of the changes and improvements I've implemented were on the external and the flight model, while the VC still needs more functionality.

I will keep the Alpha WIP updated for a while, as I'll be going though all the list, and will post some updated screenshots soon.

Also, I'm experiencing some problems with my home email - my apologies for any delay in replying to your messages - rest assured that I will read all the messages and reply asap.

Tuesday, January 19, 2010

F-35 Alpha test Work-in-progress post

Below is a list of issues either known or found in the Alpha build as released to testers yesterday night. This list includes only the "accepted" issues - i.e. things that I acknowledge as actual defects (if compared towards the scope of the project). I will keep this list updated, as reference for me, for the testers and also for anyone interested in the project progress.

Note to testers: no need to rush - I won't have much time until the next weekend. While evaluating the flight model, please keep in mind the fact that the aircraft is loaded with two JDAMs.

UPDATED 23-01-2010

External model

1) Nozzle animation incorrect.
Action: redo animation sequence. Evaluate link to nozzle animation variable of FSX:Acceleration (SP2 compatibility?) or link to rpm for smoother animation - DONE in model build 43
Status:closed - solution approved for Beta - animation sequence redone. Still linked to throttle but closer to the actual behaviour.


2) Jaggies nearby the engine intake aera. Visible on close inspection only, partially due to geometry/smoothing, partially due to self-shadowing algorithm.
Action: evaluate mitigation by local application of Meshsmooth command. It may be impossible to fix completely - but will not be noticeable by most users.
NOTE: Upper to lower side discontinuity verified correct. Minor issue spotted with intake bulge curvature. Inspection highighted UVW misplacement in the engine intake area
Status:open - high priority

3) Excessive gap between engine nozzle and fuselage.
Action: check against photos and eventually modify the mesh.
Status:CLOSED - fixed on model buidl 46


4) Afterburner effect/1: effect poking through A/C fuselage. NOTE: FSX effects positioning an visibility is somewhat dependent on graphic card model and setting. Having the card set in "performance" mode makes things worse. E.g.not perceiveable on my system.
Action: modify .fx file (shrink them a little - check nominal position)
Status: open

5) Afterburner effect/2: flame should be much longer in stage 3. NOTE: a/b flame is confined into the engine nozzle on purpose for F-35B preparation. Also, my experience with "big" flames is not really positive (different users had different results).
Action: evaluate modification in .fx file. Stage 1 effect seems correct if compared to F-35 a/b videos.
Status: open

6) Wheel/tires animation incorrect. Status: CLOSED (Fixed dynamic modle for tires, diameters were wrong)

7) Navigation lights missing. Note: they were left out as all the F-35AA-1 pics I have are showing only anticollision lights. Status: CLOSED (Nav lights placed into the lights subassembly)

8) Covers for auxiliary air intakes missing when using SHIFT+E+4 command. Note: never seen such covers in any picture. Status: CLOSED (Auxiliary inlet cover modeled - but no ribbons as I'm saving polys for the weapons ).

9) Irregular bumps in the external model. Action: modify the extenal model. Sratus DONE: (model build 44) Auxiliary air intake untouched, other bumps geometry modified and new textures.

10) Poor detail in the refuel receptacle. Action: While the comment is correct, I do not have pictures clear enough to make a porper model - Edit: will model a stardard receptacle based on similar devices. Status: CLOSED. Refuel receptacle modeled in build 46

11) Poor detail in the nose landing gear. Action: Increase detail in the 3d model. Status: CLOSED - Mitigated issues with additional gear detail, previously missing.

12) Modeling mistakes in the main gear. Action: evaluate complete redo of the kinematism. Status: Closed on model build 45 - Redone most parts of the main gear and applied new textures.

13) AIM-120 not attached to hardpoints. Action: will model LAU-147 (although not actually mounted in AA-1). Status: CLOSED modeled hardpoints for AA-1 will model LAU-147 for operative model

14) Nose gear door detail/accuracy not good enough. Action: impossible to modify the shape, but will redo the hinges and add the actuator. NOTE: Applies only to AA-1 model - other models have a different door (2 parts). Status: Closed (model build 45 - Redesigned hinge and added actuator and new texture)

15) Request to model the lights in more detail. Action: modify the model. Note: initial evaluation shows that boolean operation is requested: such operation will inhibit model roll-back. Status: CLOSED (No additional modeling, but better texture)

Miscellaneous other actions (not requested by testers):

- Resize main gear wheels / tires - DONE

- Add detail to the main fuselage texture diffuse/spec/bump - DONE

- Evaluate changes in texture brightness /contrast - DONE, also miscellaneous enhancements to AA-1 texture coloring. Note: experiments with primer-only AF-1 texture looking good.

- Fix mistake in texture UVW in the inner side

- Add separate moving part for weapon door hinge - visible on real plane on closer inspection, not modeled DONE

- A/c is sitting rather low on the main gear. Changes to the animation sequence and contact points needed. FIXED


Virtual cockpit

1) Virtual cockpit frame rate low / excessive number of polygons.
Action: optimize virtual cockpit model mainly by removing some of the parts of the external model attached for track IR users (eg. lower surfaces fuselage, flaperons, tails). Evaluate reduction of invisble polys/faces. NOTE: Only one tester has highlighted this issue. Performance seems on par with similar products for the others.
Status: open

2) Temporary HUD useability (keeping this open in case the new HUD/HMD is not available on time)
- Proposal to reduce the size of the 3D part
- Enlarge the G meter text window
- Enlarge various text windows in 2D panel

CLOSED / NEW HUD IMPLEMENTED AND AUTHORIZED

3) Minor texturing UWV misplacement on cockpit floor.
Action: fix mapping in .mdl file. DONE in VC build 19
Status: CLOSED


4) VC night lighting too bright.
Action: dim the brightness in .fx files.
Status: CLOSED - modified .fx file


5) Incorrect tooltip showing "parking brake" in the battery area.
Action: check .XML embedded in the .mdl. DONE in VC model build 19
Status: CLOSED


6) Miscellaneous autopilot interface:
- Autopilot master hotspot not working. NOTE: A/P master will be part of the FCS panel
- Autothrottle miscellaneous issues. NOTE: At first check it .XML appears correct. I had similar issues in the tomcat.
- Autopilot max bank activation. NOTE: Presence of this command is a mistake and will be removed.
Action: check associated .XML
Status: open

7) Engine needles behaviour not correct
Action: check associated .XML
Status: CLOSED. COMPLETELY NEW GAUGE

8) Visual hole aboce the pedals (polygons missing); Action: modify the Vc model (add a cap) Status: CLOSED (VC model build 20)

9) "No Hold" texture stretched above ejection seat; Action: modify the Vc model / texture; Status: CLOSED (VC build 23)

10) "patches" on the side of the Vc glass are unconnected to the canopy; Action: modify the Vc model; Status: Unclear how the thing should look like. CLOSED for now.

11) Explosivc cord not connected to glass; Action: pending evaluation - this was intentional to prevent clipping that happened during initial tests; Status: Closed with no action - Current geometry prevents clipping. Leave as is.

12) Stick movement excessive if compared to real F-35 stick. Action: modify animation sequence. Status: closed (VC model build 20 movement reduced, still not as little as the real one but provides a "visual feedback" that enhances the experience IHMO).

NOTE: Track IR testing generally good. No major issues, apart from the "usual" seat clipping and minor flaws in extreme conditions.


Flight model

1) Excessive top speed. Top speed in excess of 2.05 Mach. Real plane rated 1.6 to 1.8 (varying according to sources).
Action: change the wave drag table in the .air file - DONE
Status: CLOSED - approved for Beta. New wave drag / Mach table makes it extremely difficult to exceed Mach 1.69


2) Weak brakes. Brakes cannot hold the plane still even at relatively low RPMs.
Action:modify brake effectiveness either in the .air or .cfg file - DONE
Status: closed - promoted to beta. The brakes are now strong enough to keep the plane still up to almost full military power. However, brakes seem excessively strong when stopping the plane after landing.


3) A/c prone to yaw instability during takeoff run. NOTE:Verified only with crosswind.
Action: root cause unknown. Check for asymmetric loads. Possibly increase yaw stability/MOI or adjust scalars.
Status: Modified values in the air file parameter. Symmetric load checked and correct. Still happening with significant crosswind (15 kts or above). Should be manageable by most pilots. Awaiting flight test to close. CLOSED. LEFT AS IS FOR BETA TESTING

4) Micro-oscillations while on ground. Probably due to contact points instability.
Action:modify contact points parameters in the .cfg file. NOTE: was not present in earlier build (with incorrect auxiliary contact points) DONE in latest aircraft.cfg
Status: CLOSED


5) Pitch changes depending on throttle setting. FCC/FCS should compensate such effects in the real plane. Effect seems related to center-of-lft vs. center-of-weight misalignment, as proper .air files parameter seem correct.
Action: Check .air and modify a/c balance in order to minimize such effect. NOTE: for FCC/FCS emulation, it may desireable to set to 0 all the secondary effects of aerodynamic surfaces.
UPDATE: All fore/aft loads artificially balanced towards the aerodyamic center. Load is now balanced versus pitch. This is not theoretically correct, but makes up for the FCC/FCS absence. Issue seem mitigated to the point of being acceptable for the vast majority of the users. UPDATE: I forgot to to balance the engine pitch momentum - my bad, thanks Robby.

Status closed: Behaviour balanced in most situations.


6) Pitch changes when activating virtual speedbrake. FCC/FCS should compensate such effects. See 5) - it is probably the same issue.
Action: See 5
Status: Closed. Not completely solved but definitely acceptable.


7) Low roll rate. User report low roll rate for a fighter. NOTE: it is unclear what the actual roll rate should be. Current roll rate seems realistic for a plane with 2 JDAMs and 2 slammers, but may be less agile than real F-16,
Action: Aileron action enhanced by 10% - but awaiting further evaluation and feedback.
Status: CLOSED on latest fligh model - enhanced ailerons effects and travel.
As a good side effect, a/c is more maneuverable at low speeds.

8) Excessive loss of energy during high-G turns. Action: unclear what the target should be for this plane in the depicted configuration. Assuming 5 - 6 g sustainable turn. Will see what I can do. Status: CLOSED with thanks to Roberto Leonardo

9) High AoA dynamics causes excessively low stall speed (75kts). Action: TBD some users are happy with the high AoA dynamics, as in line with some claims by LM. Will evaluate minor changes. Status: open - unclear if it is better to leave it as is. CLOSED. LEFT AS IS FOR BETA TESTING

10) Check landing attitude AoA (approach at 140kts?). Action: TBD - The only information I have is about X-35 whose appoaches were conducted at rather high speeds (155/165 kts). NOTE: Most users report the landing behaviour is good. Status: open - unclear if any action is needed. CLOSED. LEFT AS IS FOR BETA TESTING

11) Full flap deployment happens at low speed. Action: TBD - flap is controlled by FSX via the maneuvering flap option. Check if it possible to modify the speeds, otherwise no action. Status: open. CLOSED. LEFT AS IS FOR BETA TESTING

12) Virtual speedbrakes not much effective. Action: easy fix to the .air file. Status: CLOSED - Boosted speedbrakes effect


APPENDIX A: LIST OF MISSING (BUT PLANNED) FEATURES IN THE VC

- ICAWS (currently not working at all)
- FCS window to include touch screen functions for flaps/brakes/trim/autopilot master switch (note: this may not be realistic as those are typical HOTAS functions...still may have some use for FSX)
- FCS window will show more information
- Context sensitive weapon window pop-up (with no actual functionality)
- Touch screen radio window within the MFD
- Context sensitive HUD backup window within the MFD
- Fuel receptacle/weapon bay opening activation via touchscreen (currently implemented but not working due to an unidentfied bug...)

APPENDIX B: LIST OF OTHER MISSING (BUT PLANNED) FEATURES

1) AF-1 vsual model:
- new two piece nose landing gear doors
- new APU exhaust location
- new probes
- jagged fuselage/nozzle interface
- weapon bay: undecided between real and mock JDAM
Livery: still undecided between official artwork or actual primer-only appearance

2) Production F-35 model - same as above but:

- Removal of the nose probe
- EOTS
Livery: Most probably USAF 58th FS "Mighty Gorillas", which I understand is the first confirmed unit to receive F-35s.

3) Generic F-35 model for repaint

APPENDIX C: LIST OF NICE-TO-HAVE FEATURES (NOT PLANNED)

1) F-35A Thunderbirds (according to promotional LM artwork).NOTE: would require a special flight/visual model (with no weapons)

2) F-35A with external loads

3) F-35B STOVL version. Would require a lot of work to modify the visual model. Also, it would have the "superflaps" flight model.I know it is pretty lame but STOVL dynamics is not natively implemented in FSX... FSUIPC.dll methods exist but I do not want the model to be dependant on external modules

4) F-35C Would require a lot of work to modify the visual model and extensive flight test for carrier trials.

5) Custom sounds - I know it would be nice...

Color key:
Green - issue solved, or mitigated to the point is acceptable or not perceived by most users
Yellow -issue open, and could be perceived as product defect by many users. Awaiting action or investigation. Or issue open, mitigated to the best of my knowledge but will still be perceived as defect by users. Or issue open but appears easy to fix.
Red - root cause not identified as of today. Or the problem is beyond my knowledge or impossible or difficult to solve.

Monday, January 18, 2010

F-35 Alpha Test starting

First I'd like to thank you once again for the outstanding quantity and quality of applications to Beta. I'm sorry I had to left out the vast majority of them, but I had to keep the test group to a manageable size.
Yesterday I've finished replying to all of the emails. I'm sorry it took so long, but they were so many that it took several hours to read them, evaluate and reply to each one. By the way, a couple of the replies were bounced back as spam - my apologies if you did not get one.

Also, I've compiled an Alpha build which will be distributed to the testers tonight. The visual model is complete and so is the virtual cockpit model.
Flight model is "almost right" in terms of numbers (meaning think I entered the right ones and had a resonable guess to the rest) - with possible exception of some aerodynamic data. After loading the model with 2 JDAMs and 2 AMRAAMs, I'm under the impression that the plane seems underpowered. Maybe I should relax the wave drag vs.Mach table.

This will be an Alpha just because it is still missing some of the planned features, mainly in terms of cockpit functionality. But I think it is good enough to get some useful feedback from it.

Wednesday, January 13, 2010

F-35 Beta...will actually be an Alpha.





I'll be reviewing the applications for Beta testing tonight - my apologies for the delay, but these are pretty busy days for me...

Speaking of the Beta...well technically it will be an Alpha, i.e. will not be feature complete. More specifically, the MFD will work as initially intended...but I am having some ideas on how to improve it. Also, Jivko Rusev is working to an accurate HMD based on all the information available on the real one so far - but may not be included in the Beta build.

Long story short, this Beta will be focused on the visual model (external and virtual cockpit) and the flight model, while the cockpit will lack some of the planned functionality.

The Beta will feature the AA-1 prototype only, while the final packge will also have AF-1 (I'm still undecided if it will be in the actual "upainted" colors, or in the AF-1 scheme showed a while ago by LM) and an operational version.
F-35B and F-35C, at least for the moment, are being left out. I know that many of you are waiting for them but I want to be sure that the F-35A is as good as it can (reasonably) be before making anything else.

On the good side, I've tried it under SP2 and seems to work (with the exception of the features borrowed from the stock Hornet).

Sunday, January 10, 2010

F-35 project - Beta testing application / UPDATE

Finished model of the Beta (and yes, I've added the wheel chocks!).

Thanks for the outstanding response to my request for Beta test pilots. The quantity (and quality) of the applications is largely above my expectations, and definitely much more than what I can handle.
I will pick 10-12 testers maximum - my apologies in advance to those whose will be left out. I really appreciate the support of everyone, but, being this a small (almost)one-man-band project, I do not have the time to manage the feedback and information-flow of a much larger group.
I will try to reply to each e-mail, but it could take a couple of days.
By the way, as per emails and previous posts, Beta version will be sent, upon request, to anyone who has sent a Paypal donation for my freeware projects before January 7th. These Beta copies will not count towards the 10/12 limit.

Friday, January 8, 2010

Goodbye to the real F-35 AA-1...

The virtual F-35 AA-1 model parked at Edwards AFB. As visual treat, SHIFT+E+4 will remove the pilot and add engine covers and "remove before flight" ribbons as shown in this picture. I should probably add wheel chocks. By the way SHIFT+E+2 will open the in-flight refuel receptacle and SHIFT+E+3 will open the weapon bay doors.

I've just read that the real life F-35 AA-1 has been officially retired by Lockheed Martin after 91 test flights and 3 years of service. But instead of being placed in a museum, it will be the subject of live fire testing for battle damage assessment purposes.

AA-1 is(was) the first F-35 test aircraft built, and it is(was) roughly 5000 lbs overweight compared to the current CTOL test aircraft, named AF-1 which took flight for the first time in November 2009. And it is the only model that will be depicted in the Beta.

Goodbye AA-1.

EDIT: Here is a link to a montage of the AA-1 test flights straight from the JSF website

http://www.jsf.mil/video/f35test/AA-1%20Montage_high.wmv

Thursday, January 7, 2010

F-35 cockpit geometry complete and a word of clarification about money and Paypal

This is the final cockpit geometry for the Beta - but some of the planned "functional" features are still missing....

I have to confess I spent most of this Wednesday (which was bank holiday here in Italy) playing FSX and flying above NAS Forth Worth in the F-35. Quite fun. Some more "touch" functionality has been added to the MFD (but I repeat: do not expect anything super - no windows popping/resizing or other cool stuff the real F-35 does). So far I'm quite happy. This build in still somewhat Acceleration dependant - but should work in SP2 with some restrictions (I still have to check).

But this is not "really" a post about the F-35. It is a post about money. The F-35 will probably be my first payware project ever. "Probably" means that I have to iron out a thing or two before taking a final decision. This does NOT mean that I will abandon the freeware. In particular, if I will ever find the time, I would like to update both the T-45 and the F-14.

Back to the Lighting II, it has been tought from the beginning as a payware and I thought I was clear on this in this blog. It has been designed to be a low price/low budget/reasonable quality project, meaning some of its features may not be on par with some other AAA aircrafts out there. But I am doing my best and should be on par with my previous FSX projects.

As for the price, it will be probably in the 10-15€ range - depending on the distribution method. I know some of you may be disappointed but keep in mind that a fair FSX plane could easily take 500 hours to make or more. As for the distribution method I've not decided yet. SimMarket seems the best option at the moment.

One more thing about the Paypal donations. First I'd like to say THANK YOU to anyone who made a donation so far. Much appreciated. Then, I'd like to clarify once again that this button was placed on this blog on request of several T-45 and F-14 users who asked for it. As the F-35 will be payware, this may be misleading - and I hope that none of those who have donated felt cheated. If so, I'm sorry it was not my intention. As compensation for any possible mistake, whoever has done a donation so far will receive the test versions of the F-35 (once the Beta is ready and upon request).

I'm thinking to remove the button completely, but, for the moment, I'm just changing the text. Let me know what you think.

Tuesday, January 5, 2010

F-35 project update

Finally the cockpit model is almost complete!

Experimenting fixed camers angles

Quick update - I've made huge steps towards the completion of the cockpit model. It is almost done: actually the only thing that needs some work is the canopy - which I hope it will be completed by tomorrow. From the functionality standpoint - the basic things are working. It should not take too time to have everything you need for VFR flying working. MFD will have a fixed layout (i.e. you cannot change windows type/size/position like in the real thing) - it is possible to make a folly working MFD in FSX - but I'm not a "gauge-guy" and it will take a while to make. So I decided for a fixed layout.
Also I'm experimenting with the HMD - some things work fine some other don't. Basic flight data will be superimposed on the VC via a transparent gauge - while I still need to find a suitable soution for the horizon projection.
Fligh model seems to work fine - albeit it is hard to tell if it is realistic...
As for the external model it does not need much work, I think - but it will not change it before the beta. Speaking of the beta, I will first complete the canopy, add some functionality to the VC and implement basic HMD. But it should not take more than few days, I hope.