Print

Please login or register.
1 Hour 1 Day 1 Week 1 Month Forever
Login with username, password and session length

[Herra Tohtori]

I use a GeForce GTX560. Asus GTX560 OC DirectCu II, to be specific.

EDIT:

in fact yes, it is a matter of convention and it works consistently. So Convergence ~= Inward crossed view ~= Positive Values.

I found the problem.

PAL3DConvLeft changes the right eye point's convergence angle.
PAL3DConvRight changes left eye point's convergence angle.

I tested this by setting

PAL3DConvLeft=0
PAL3DConvRight=-45

The end result was right eye looking straight forward through the gun sight, but left eye was looking 45 degrees to the right (inward).

Naturally this inverts everything else regarding convergence as left becomes right and inward becomes outward.

[benitomuso]

I use a GeForce GTX560. Asus GTX560 OC DirectCu II, to be specific.

It seems to be the same. So please, confirm if after some seconds that "first tilt" is not normalized by itself letting later select whatever you want.

Regards.

[benitomuso]

I use a GeForce GTX560. Asus GTX560 OC DirectCu II, to be specific.

EDIT:

in fact yes, it is a matter of convention and it works consistently. So Convergence ~= Inward crossed view ~= Positive Values.

I found the problem.

PAL3DConvLeft changes the right eye point's convergence angle.
PAL3DConvRight changes left eye point's convergence angle.

I tested this by setting

PAL3DConvLeft=0
PAL3DConvRight=-45

The end result was right eye looking straight forward through the gun sight, but left eye was looking 45 degrees to the right (inward).

Naturally this inverts everything else regarding convergence as left becomes right and inward becomes outward.

Are you positive about that? You mean you see the Cyan image in normal front oriented position and the Red one twisted?

  Here I don't have a way of testing the game, but I'm checking the code and I don't see any inversion on what angle is applied to each eye.

  Very strange!!! In a couple of hours I'll be able to test it live.

  Regards,
                         Pablo

[Herra Tohtori]

It seems to be the same. So please, confirm if after some seconds that "first tilt" is not normalized by itself letting later select whatever you want.

Sadly, I'll have to take a rain check on that at this time.

I'm going to my parents' to spend christmas and new year's with family, and my night train leaves in about two hours. I just booted into my Linux partition so that I can use SSH/SFTP to connect to my stuff when I'm away, but I can't test the mod right at the moment... and probably not really until some time in January, really.

I'm happy I got to participate in the feature testing so far, though. This mod is quite incredible already and will only get better. I probably can still make some graphics if you require some, but I will not be quite as accessible once I get home, for obvious reasons.

Are you positive about that? You mean you see the Cyan image in normal front oriented position and the Red one twisted?

  Here I don't have a way of testing the game, but I'm checking the code and I don't see any inversion on what angle is applied to each eye.

  Very strange!!! In a couple of hours I'll be able to test it live.

Well, it's not so strange, variable switches can happen in programming just like sign flips can happen in mathematics, it happens to the best of us. I suppose there is one equation somewhere, where left and right have switched places, and all the rest of the code follows from that simple error. Good thing is, it doesn't cause any functional problems - just a minor interface glitch.

But yes, the cyan image (right eye) was looking dead center forward, while red one was twisted 45 degrees clockwise, port, right, inward or however you want to express the direction. I was also surprised to see that, but it did explain why I kept seeing negative convergence values push the objects away, despite the opposite intent.

[benitomuso]

OK Herra.

   So Merry Christmas and Happy New Year for you.

  Thank you for your help and for being a tester who knows about the subject and knows what to test without too much explanation.

  Your graphics are very clear and detailed and as I told you I will add the last one to the ReadMe (Manual of the MOD) obviously with the proper credits for your contribution.

  Later I will take a look at what you recently mentioned about the inversion of the eyes twist. One last question: do you use Mode A? Because now I'm thinking that the problem can come from the different between Mode B (the one I'm using) and Mode A, generating that in Mode A only. I'll see.

  Regards,
                      Pablo

[Herra Tohtori]

I don't remember seeing any drastic shift between mode A and mode B, it looked quite similar to me - but, I wasn't really watching for changes happening in the transition.

However, shouldn't the view vectors be quite independent from the rendering method used? It's like, if you put a camera on a tripod mount, then put in black and white film, take a picture, switch to colour film, and suddenly the camera would be pointing into different direction.

I don't know how the mod is coded, but it would sound quite a strange reaction for viewpoints to change due to switch in rendering method.


Merry Christmas to you as well, and a happy new year.

[santobr]

I noticed that the values shown in the screen sometimes are correct, but sometimes they are wrong.
It looks like the value is reseted to zero, but the 3D view stays the same, losing the reference.

Is there any chance of customizable colors for each eye in conf.ini.
If so, we could test another color scheme like Trioscopic or optimize the red and the cyan to our glasses.
Maybe even reach the color of INFICOLOR scheme.


Thank you so much, Pablo.

Merry Christmas.



santobr.

[shardana]

Hi pablo, i'll be out of sardinia for a week. I'll try to get glasses when i come back home, and i'll check if this works for me and i'll let you know. Thanks again

[benitomuso]

I noticed that the values shown in the screen sometimes are correct, but sometimes they are wrong.
It looks like the value is reseted to zero, but the 3D view stays the same, losing the reference.

Is there any chance of customizable colors for each eye in conf.ini.
If so, we could test another color scheme like Trioscopic or optimize the red and the cyan to our glasses.
Maybe even reach the color of INFICOLOR scheme.


Thank you so much, Pablo.

Merry Christmas.



santobr.

Santo,
                   I've reviewed different things for what is going to be the V6 of the MOD, and I hope the final version. The changes in V6 respect of V5 are:

  -Implementation of Convergence in Stereo view. It allows deeper stereo perception with less eye strain. Now it formally follows the convention of considering positive values those ones that cross your eyes "clockwise, port, right, inward or however you want to express the direction" and negative ones the oposite.
  -Correction of a mistaken way of alternating the eye shift that at the beginning caused some troubles with Convergence, but it was a problem in Separation not noticeable before.
  -Fixation and different way of determining if the plane is seen in external or internal view. This makes fail safe the adjust of Separation.
  -Reviewed some complaints made by Santobr regarding incorrect adjustment in flight.
  -Extra precision to adjust Separation and Convergence.
  -Established mechanism to avoid freezing of the game when switching to Mode C and your video card doesn't manage Auxiliary Buffers. Now in the worst case you will not see stereo at all but the game will keep working.

  Using this version with Convergence Adjustment, I realized that after having set the Separation is more importante to be able to adjust at any time convergence, so I established this new setting that allows you to have Convergence as default:

[Mods]
PAL3DAdjustConv=1

  The default value is 0, it means the old fashion operation that when the game starts you can change Separation until in non-Stereo Mode you switch to Convergence adjustment.  But with PAL3DAdjustConv=1 the game starts adjusting Convergence instead and you have to switch to Separation if needed (less needed now, that's why I offer this chance).

  As a wrap up of all the current settings until now, this is the list with the default values:

[Mods]
PAL3DStart=1
PAL3DSeparation=1.0
PAL3DSeparationExt=25.0
PAL3DConvLeft=0.0
PAL3DConvRight=0.0
PAL3DAdjustConv=0
PAL3DScrShtDir=MyScreenShots/
PAL3DScrShtExt=jpg
PAL3D6DOF=1
PAL3D6DOFAutoReset=0
PAL3DEcranWide=0

  Regards and Merry Christmas,
                                                          Pablo

[Whiskey_Sierra_972]

This new version make my 409 installation don't even start....the exe window run , disappear and reappear after few seconds!

[santobr]

Thank you Pablo, when I grow up I want to be like you.



santobr.

[benitomuso]

This new version make my 409 installation don't even start....the exe window run , disappear and reappear after few seconds!

Walter,
             sorry but I don't have the chance of checking compatibility with 4.09. I develop in 4.101 basically DBW oriented that is "our standard".

  I can figure out what's the only difference ni this MOD that could make 4.09 to crash, but it will be a problem having to consider every time to create a 4.10 and 4.09 version.

  Regards,
                     Pablo