Varying transparency of star & galaxy labels

[ElChristou]

I agree totally with neo albireo. The new fading effect is very nice. However, the last orbits are still disappearing abruptly.

I don't see this effect of the last orbits disappearing abruptly . . .

I also confirm this behavior...

[chris]

I agree totally with neo albireo. The new fading effect is very nice. However, the last orbits are still disappearing abruptly.

I don't see this effect of the last orbits disappearing abruptly . . .

I also confirm this behavior...

It seems very much like a Mac-only issue

--Chris

[chris]

I agree totally with neo albireo. The new fading effect is very nice. However, the last orbits are still disappearing abruptly.

I don't see this effect of the last orbits disappearing abruptly . . .

I also confirm this behavior...

Strange question: do you see this no matter what angle you zoom out from the solar system? That is, does the same thing happen whether you're looking at the solar system either edge-on or from above?

--Chris

[Cham]

It seems very much like a Mac-only issue

--Chris

Well, the rest is working. No problems with the labels, and the other orbits are fading correctly. The moons orbits are all working correctly too. If it's a Mac problem, I wonder why it affects only the last orbits. I suspect it's related to the screen size/ratio.

[Cham]

Strange question: do you see this no matter what angle you zoom out from the solar system? That is, does the same thing happen whether you're looking at the solar system either edge-on or from above?

--Chris

I just made a test : the last orbits are disappearing abruptly in all cases, no matter what is the viewing angle.

[ElChristou]

Strange question: do you see this no matter what angle you zoom out from the solar system? That is, does the same thing happen whether you're looking at the solar system either edge-on or from above?

--Chris

I just made a test : the last orbits are disappearing abruptly in all cases, no matter what is the viewing angle.

Confirmed!

[neo albireo]

Strange question: do you see this no matter what angle you zoom out from the solar system? That is, does the same thing happen whether you're looking at the solar system either edge-on or from above?

--Chris

I just made a test : the last orbits are disappearing abruptly in all cases, no matter what is the viewing angle.

I do not fully confirm this. I think that the behaviour IS different at various angles. I'm at work now and can't check this now, but the distance of the disappearing was different if edge-on or from above.

[edit]Mac users, just go to a distance where the orbits just disappeared and move around the sun - on my computer, the orbits will appear again in some positions that are related to the viewing angle. Just try with different distances and you will probably see the same effect.[/edit]

[chris]

Thanks for the test. I think that I may have a grasp on what the problem is. I should have a fix coming shortly.

--Chris

[neo albireo]

It is really hard to reproduce the behaviour of this bug. Sometimes you can make a full turn around the sun with the orbits visible and the next turn will be without orbits. Sometimes you are very far and can make a full turn around the sun (complete with edge-on and view from above) with orbits and sometimes you are nearer and the orbits disappear for a whole turn.

One thing happens quite often, though: The orbits pop out again when you reach view from above, or zenith. They seem more likely to disappear near edge-on view.

[dirkpitt]

Mac users, just go to a distance where the orbits just disappeared and move around the sun - on my computer, the orbits will appear again in some positions that are related to the viewing angle. Just try with different distances and you will probably see the same effect.

I am a Mac user, and indeed I can confirm this angle-dependence.

[ElChristou]

I cannot determinate this angle dependence on my config, the behavior is quite chaotic... Perhaps it's my card...

[t00fri]

...
Do I understand correctly that the fading is based on apparent magnitude and not distance?

--Chris

It may be useful if I explain some of my label-fading code here. It is basically simple and very similar for galaxies and stars. The label transparency 'distr' (and thus the label visibility) ranges between 0 and 1.

Code: Select all

float distr = 3.5f * (labelThresholdMag - appMag)/labelThresholdMag;


It only depends on the ratio appMag/labelThresholdMag

The more tricky variable is labelThresholdMag. It is basically constructed following my automag philosophy. For a given sky brightness, it essentially depends on the field of view(FoV) or equivalently, the "effective screen distance" of the user. In a way this emulates the magnification of a telescope and keeps the density of visible labels on the screen about constant.

When the FoV decreases ( <=> increasing magnification) the labels of the brighter objects move out of the field while the labels of dimmer ones appear.

This crucial trick avoids overcrowded/overlapping labels!

labelThresholdMag and thus FoV is typically altered via SHIFT + mouse-LEFT movements.

The rest is easy. The scaling factor in my current implementation is indeed apparent magnitude (appMag) rather than a function of distance alone. The latter would easily be possible as well, by defining and using above --instead of appMag-- a modified apparent magnitude, referring to some suitable constant absolute magnitude for all stars, say.

Using appMag, however, has the advantage that all prominent stars in the constellation patterns are mostly labelled initially, since here apparent magnitude and NOT distance is the relevant quantity.

Bye Fridger

[t00fri]

I have committed my fading star label code to CVS (render.cpp)

Bye Fridger

[chris]

Mac users, just go to a distance where the orbits just disappeared and move around the sun - on my computer, the orbits will appear again in some positions that are related to the viewing angle. Just try with different distances and you will probably see the same effect.

I am a Mac user, and indeed I can confirm this angle-dependence.

I believe that I may have fixed the problem. I committed the fix to CVS, but I haven't had a chance to test it on my MacBook yet. I'd be interested in hearing reports from Mac users who've updated from the CVS (after 10am PDT).

--Chris

[ElChristou]

...I'd be interested in hearing reports from Mac users who've updated from the CVS (after 10am PDT).

Build 3 minutes ago, the orbits are just fine for me. Tx Chris!

The stars labels are also very well, one can appreciate a full during a star tour... nice work Fridger!

[chris]

...I'd be interested in hearing reports from Mac users who've updated from the CVS (after 10am PDT).

Build 3 minutes ago, the orbits are just fine for me. Tx Chris!

Thanks for testing the fix!

--Chris

[tech2000]

There seems to have been alot of fixes and ofcourse Fridger's new fading star labels gone into the CVS.

Sorry for breaking into this thread like this but are we close to pre3 release?

Cheers, Anders

[ElChristou]

Sorry for the noob question; while playing with the fading orbits (and labels btw nice!) I was wondering if Venus with it's 0.77 albedo is so bright to see it's dot disappearing after Mars orbit... is that correct?

Edit: Indeed seems odd to me. At this distance the Sun is rendered with one pixel and Venus 4... ok Venus is not pure white like the Sun, but still very present (too much to my taste)...

[chris]

There seems to have been alot of fixes and ofcourse Fridger's new fading star labels gone into the CVS.

Sorry for breaking into this thread like this but are we close to pre3 release?

Cheers, Anders

Yes, very close.

--Chris

[ElChristou]

Still playing with the orbits, but now it's about the planet's dots; at aprox 7 ua of distance, they disappear abruptly... someone confirm?