VisualPSX Suite released

[ahaka]

Great to hear that you are still continuing to develop this software, Garry!

To those asking about the smoothness: I have come to conclusion that no networked solutions can give you 100% smoothness like PSX does, without using some sort of artificial frame creation.

At the moment, the only thing that works for me is the X-Plane 10 with XView, but even though I get constant 70hz from PSX boost server, and my visual computers are running at a stable 60Hz (60fps), there is never a complete smoothness in the visuals unless I enable the "extrapolation" function in XView, which, to my understanding, creates some artificial "frames" to smooth out the visual. Why it is so, is beyond my understanding, as it would seem logical that 70fps of data from the boost server is enough for perfect smoothness on a 60Hz display, but lo and behold, it isn't.

[Hardy Heinlin]

Hi Antti,

how do you know that your FPS indications are constant? How often are they updated? Once per second?

For example, when the indication reads 50, you get 50 frames per second indeed. And you may get a frame every 20 ms. But you may also get 50 frames in the first half of the second, and no frame within the other half of the second. For example! If the FPS indicator refreshes every 1 second, it indicates just the sum of the past second. It says nothing about equal frame distribution within this 1 second.

PSX has two indicators:
- a frame rate indicator (refreshes at 1 Hz)
- a frame time indicator (refreshes after each frame)

The frame time indicator is the green-magenta bar under the FPS digits. The more green, the shorter the frame time. The more steady the bar, the smoother the frame distribution. A momentary flicker towards magenta indicates a momentary frame rate reduction. Such a short reduction will not affect the sum of frames within a second, hence it won't be visible on the frame rate indicator. It will be visible only on the frame time indicator and on moving objects.

Re extrapolation in Mike's XView: I think in Garry's VisualPSX there's an optional extrapolation as well, isn't it?


Cheers,

|-|ardy

[ahaka]

Excellent point Hardy. I don't know the update rates of FPS indicators in the other simulators, but they seem to be more than 1Hz in any case.

However, I know that all of the simulators are constant/smooth enough alone (in perception, I'm not claiming mathematically), because my eyes can't detect any judder in their motion. Only when a network setup comes into play, the micro stuttering starts, until the extrapolation is enabled. The only problem with having to use extrapolation is, that if you use multiple computers to drive parts of the visual system, it occasionally causes some disturbances/sync problems, as the extrapolation is done by each computer individually.

So I think you might be right that the frames don't perhaps spread out evenly over 1 second when they are sent over the network. But it really seems to be network related, not that the simulators themselves would have any problems doing their job. Whether it's the network card/switch or the tasking the network traffic puts on the CPU, I have no idea, but something's at play here. For the record, I have tested it with different network cards/switches/computers, and the results have so far been pretty much the same.

[Hardy Heinlin]

OK, if your eyes can't detect any judder, the FPS is obviously constant.

Of course, another way of indicating the FPS is this: Refresh the indicator after each frame, and devide the last frame time (in ms) by 1000 ms. The other simulators probably use this method. The disadvantage is that the indicator digits update too fast to make a short value drop visible. An analog bar indicator shows the smoothness better, I think. Anyway, if you don't see any judder in the graphics, that's proof enough.


|-|ardy

[Avi]

Hi,

After a very (very very) long time I decided to connect PSX to FSX.
Since I have FSX Acceleration version I didn't installed SP1 and SP2 (I don't know if this has anything to do with my problem).

I followed the installation instructions (including creating the Runways_PSX_v53.vpx file) and started VisualPSX.

FSX did jump to LLBG (where PSX is) but it wasn't aligned with the runway (26).
I followed the alignment instructions. I set PSX become lights to OFF (got FSX Position Locked in VisualPSX), selected LLBG in the PSX aircraft position page, selected Take-off for runway 26 but in VisualPSX I got "No matching FSX runway found" in yellow.
How come? What should I do (or do wrong)?

Thanks,

[Garry Richards]

A question with Visual PSX. Does PSX send the date to FSX?

Hi Blake,

Sorry to have overlooked this question. VisualPSX does send the PSX date to FSX but the year is fixed (at 2011 if memory serves me correctly). The month, day and time are all sent and match the settings in the PSX instructor.

Cheers,

[Garry Richards]

... selected LLBG in the PSX aircraft position page, selected Take-off for runway 26 but in VisualPSX I got "No matching FSX runway found" in yellow.

Hi Avi,

You have found a bug in VisualPSX.   I don't yet know why it happens with that particular runway at that airport but will investigate.

Cheers,

[Garry Richards]

Hi Avi and Hardy,

I have found and fixed an anomaly in VisualPSX that prevents it from assigning runway offsets in some circumstances. I refer to it as an anomaly rather than a bug because I am not sure if it is due to some change I made to VisualPSX or that Hardy made to PSX. In any event it was a one line fix that will appear in the next update to VisualPSX.

Hardy, the problem was that when changing runways through the instructor, PSX sends data for the runway it is leaving before sending data for the new runway. VisualPSX will now ignore repeated runway data so no change to PSX is required.

LLBG runway 26 is a different matter. There appears to be an error in the PSX runway database. When you select runway 26 from the list in the instructor, PSX moves to runway 26 but sends data for runway 21 instead. VisualPSX then cannot reconcile the two.

I have now cleared all known bugs from the VisualPSX suite and have finished adding new features. I will be able to release it shortly after I have updated the manual and other documentation.

Cheers,

[Hardy Heinlin]

Hi Garry,

I think the database is correct; I guess this is an algorithmic problem of the runway determination in the EGPWS model (you get EGPWS TCF [terrain clearance floor] system data from PSX). LLBG 26 has a displaced landing threshold. If the aircraft is in the takeoff position of 26, the aircraft is closer to the landing threshold of 21 than to that of 26. And the runway headings differ by less than 60°. So the EGPWS thinks 26 is the desired runway.


Cheers,

|-|ardy


I just reduced the tolerance for the distance criterion from 0.15 to 0.03 nm. This should fix the problem. Available soon in 10.0.4-beta1.

[brian747]

Hi Garry!

As I posted in the "lock" thread (http://aerowinx.com/forum/topic.php?post=27713#post27713) but forgot to cross-reference here:

"...with my PSX aircraft at a stand at Heathrow EGLL" [stand 534, if I remember correctly] ", selecting runways 09L or 09R results in the VisualPSX message "No matching FSX runway found". (Although runways 27L & 27R perform as expected).

(I'm using the Aerosoft version of Heathrow, incidentally...)."

I mention this just in case you may wish to test the new version at EGLL prior to release.    

Cheers,

Brian

[Hardy Heinlin]

Re LLBG problem:

http://aerowinx.com/forum/topic.php?id=2782

[Garry Richards]

I mention this just in case you may wish to test the new version at EGLL prior to release.    

I have and it works.  

[Garry Richards]

I just reduced the tolerance for the distance criterion from 0.15 to 0.03 nm. This should fix the problem. Available soon in 10.0.4-beta1.

Thanks Hardy.

Cheers,

[Avi]

Thanks Garry and Hardy.

If there is a problem with the runways at LLBG, and there is, it is in FSX and not PSX since both runways 21 and 26 are today (much) longer than they are in FSX.

Cheers,

[brian747]

>> "I mention this just in case you may wish to test the new version at EGLL prior to release. "

> "I have and it works. "

Fantastic!!  Thank you so much, Garry!    :mrgreen:  :mrgreen:  :mrgreen:

Huge thanks to you, too, Hardy (I have also thanked you in your 10.0.4. Update thread).    

Sincere congratulations, gentlemen! And enormous thanks for making PSX and its add-ons the very best of the best.

Cheers,

Brian

[simbro]

When using RunwaysPSX.exe to create the matching table I have set the tickbox to include Addon airports.

However, it seems to me that only addons inside the Addon-Scenery folder are being processed and the two addon folder structures UK2000 and Aerosoft directly under the main FSX folder are not being included in the results.

Because I can find no way of checking this I thought the simple way of finding out was to ask here.  

I haven't noticed any problems with misaligned runways (sofar) but like having things straight in my head!

Kind regards
Simon

[Garry Richards]

Hi Simon,

That problem is fixed in the next update, not yet released.

Cheers,

[GodAtum]

Is there a good 747 model (not the default) I can use that has all the lights working realistically?

[GodAtum]

Hi gary, your website appears to be down.

[Avi]

Hi Garry,

Something interesting I noticed.
When I select Disable all approach offsets there are no alignment problems for runway 26 (maybe because it doesn't confuse with runway 21, I don't know).

One more thing I noticed is that any tiny touch on the brake pedals makes the outside view to pitch aggressively. Can that be solved?

Cheers,