LOD polygons: Alpha India 757 vs. Aardvark 757

[edemeijer 19]

Hi everyone!Today I see the Alpha India Group released their 757, short after AI Aardvark released their 757w version.Since the Alpha India Group 757 is newer, it has more features and will surely look a bit nicer.But how about performance and LOD's? They both have 5 LOD's available, but how about the polygons per model. Via the Alpha India site I could find the polygons per LOD, but I do not know them for the AI Aardvark model.So, in short, if I put Schiphol Airport full of 757's which one will perform better, because it has less polygons?Thnx!

[geoffco 0]

It's on the respective websites for the non-winglet aircraft - http://www.ai-aardvark.com/aiaircraft/aiaircraft.htmlhttp://www.alpha-india.net/content/view/17/33/AIA B757-200 PW has 2266 polygons at LOD 1 and has 5 LOD levels.AIG 757-200 PW has 3696 polygons at LOD 1 but has 10 LOD levels.I don't know enough about it but I'd guess that up close, the AIG is more detailed and therefore more resource hungry. However, perhaps the extra LOD levels will make it more efficient as an AI aircraft when you are not so close to it...GeoffEDIT: For what it's worth this is about equivalent to The Fruit Stand 777 models of which (if your AI is set up anything like mine) you probably have a fair number in Schiphol

[edemeijer 19]

Hey Geoff!Thanks for your reply. Could not find the polygons for the AIA model... Indeed it looks very close to each other, so I think about switching to the AIG modelThnx!

[ailchim 206]

Its to be hoped that the extra levels of detail offset the higher polygon count on the Alpha India 752.The visual models don't look a lot different to me, although the Alpha India does have a few extra bits like the do-berry wot-sit on the nose gear. It also has rotating fans - I'm not sure if the Aardvark does.Of course, what you also get are take-off flap deployment on taxy and flexing undercarriage and spolier and reverse thrust deployment on landing, so the model itself is more sophisticated than the Aardvark 757, which was one of their earlier offerings.

[geoffco 0]

I've (obviously) only had a quick look at each but I think that the windows etc of the AIG just look better... Anyway, I was slightly annoyed that after waiting so long for an AI B757 with winglets, two appear within days of each other, but after looking at both, I don't mind the time I "wasted" installing the AIA model since the AIG looks so good...Geoff

[ailchim 206]

Yes, did the same thing.Strange things happen out there amongst the AI experts I think, seems to be pretty competitive at times, which isn't always to the users advantage.

[Guest ladannen]

I did FPS testing on both aircraft, using FS9 at a test airport in the middle of nowhere.AIA first, then AIG.

[geoffco 0]

Ouch! That's no insignificant difference! I think I'll go with the AIG for airlines that I notice more / care about and use AIA for the rest (and, as always, PAI to fill out gaps).Ladannen, did you tat test without aircraft shadows being displayed?Geoff

[ailchim 206]

OUCH!Mind you, the point I had in mind was that frame rates on approach are the biggest issue.I hope 40 Continental 757's are not the biggest issue 2 miles from the thheshold

[ailchim 206]

40+ actually isn't it?Do Continental have that many aircraft lying idle?

[Guest ladannen]

Same test with "aircraft cast shadows" disabled: 34 FPS / 18 FPS.The test has 76 aircraft, to be exact.No, there is no scenario that would have 76 Continental 757s at a single airport at one time, and I can't imagine having 76 757s from a mix of airlines anywhere, at any one time.So the test does greatly exaggerate the FPS effect, but there still is an effect.I do not want to give the impression I am bashing the model. I think the new AIG model does look better. But with a large FPS drop in my system, my computer does not have the power to handle it, otherwise I would change models in a second.Edit: I did see on another board that Jannik Dahl is aware that the FPS impact could be better, and I look forward to seeing it.

[ailchim 206]

HMMMMObviously the poly count is an issue and I'm no expert, but in the real world we taxy out with a few aircraft in view at a distance and then take off with a few more in distant view.Then we land with other aeroplanes at a distance.What's the trade off between polygons & LOD's? - I don't know.The test is on the take-off roll and on final approach I'd suggest, offset by what we see in taxy mode.Sure as ####, it isn't 76 x 757 aircraft in a row, in detail.Unless one has entirely inefficient AI flightplans. Or an entirely inefficiant airline in mind,NB I'm from the UK, so an entirely inefficiant flagcarrier isn't impossible........

[Guest ladannen]

If anyone would like to test these or any other AI aircraft under your conditions, I've uploaded a package here:http://library.avsim.net/esearch.php?DLID=...hor=&CatID=root

[geoffco 0]

Have you tried the test with the newly released updated models? Would be interesting to see if they make a difference... If not, I'll do the tests myself tonight.Geoff

[Guest ladannen]

Yes, I have tried the new models.Lots of people have sent me IMs and emails asking for the results of them, as well as the FPS of all kinds of different models.That's why I uploaded my AI test, so everyone can test it themselves.Me testing them and posting the results do not have the same effect as others seeing the test results on their system.