My video review of Aerofly 2. It looks way better in person sadly shadow plays record quality sucks

no atc

no weather

no water

no ground crews

no car/plane traffic

no lights at night

etc.

may i ask how is this better than P3D and FSX as you say in your video review :rolleyes: I understand it's Early Access but it's not better yet.  Let's see how sales and development go first before concluding that this is better.

no atc

no weather

no water

no ground crews

no car/plane traffic

no lights at night

etc.

may i ask how is this better than P3D and FSX as you say in your video review :rolleyes: I understand it's Early Access but it's not better yet. Let's see how sales and development go first before concluding that this is better.

I say it's better because it looks way better than FSX or P3D and looks are about 90% of immersion. Also its performance is top notch no OOMs ect. To talk about what you said in detail

1.ATC is going to be added according to the devs but even if it isn't who cares FSX's default ATC sucks,and I never use it because of its horrible vectors for approaches and its extramly out of date.

2.I'll give you that currently you can't make it storm but you can change cloud cover,visibility,and wind.

3.Water is going to be added,it's early access so it wasn't a priority a stable program was

4.FSX by default didn't have ground crews,you'd need GSX or ASE for that ie addons,I don't use either as its a performance killer and I don't spend enough time on the ground to use it.

5.Many FSX addon photoreal also lacks night textures,I'm sure it will be added if not an addon will be made for it with the upcoming SDK.

The program has potential as its not based on a 10 year old code that DTG and LM continue to flog to get life out of it. Is it perfect no but nither is FSX and it certainly wasnt when it was released 10 years ago.

I like many parts of it better already. Certainly it seems a breath of fresh air compared to what I've seen of Flight School so far.

 

I'm tired of re-warmed servings of FSX, and perfectly willing to start fresh with a 21st century engine.

No lights at night!??

 

For real?

 

Wow.

LOL! And I have to wait till the end of the month to get it. After downloading 5 different mmo's at about 30gigs each, plus Flight School, Plus re-downloading P3D all in the course of about a week, My ISP called to ask me if I was out of my mind. (and that I was more than 3/4ths of the way through a limit I never knew I had)    

 

Never downloaded like that before, and guess now I never will again.

That provider is a chicken!  Come to Europe Devon, come to Europe...

 

And regarding the pluses of Aerofly, no one is mentioning the flight dynamics, which are better than the best in FSX or X-Plane, at least from what I recall from Aerofly v1...

That provider is a chicken!  Come to Europe Devon, come to Europe...

 

Well, once Flight School finally unlocks, I'll be at least as far as Virtual England............  :smile:

No lights at night!??

For real?

Wow.

Just the ground textures lack lights, the cockpits are beautiful at night and feature full adjustable lighting control,the runway lights work as do the lighting in the 3d buildings

I say it's better because it looks way better than FSX or P3D and looks are about 90% of immersion. Also its performance is top notch no OOMs ect.

 

I don't agree that it looks better than P3D. If you compare stock scenery of both products, it compares quite favorably. However, there are severe restrictions due to the use of satellite imagery.

 

-No autogen. Airports and major cities have quite a few objects manually placed, but outside of that, it's very flat.

 

-There's a tree coverage database, but it's not very accurate. You can see trees sticking out of roads or 2D building imprints etc and the tree coverage isn't very dense.

 

-The texture resolution rapidly tapers off away from major airports and cities. For example, take off from KHAF and fly south along the coast at < 6000 feet. It's worse than Flight Unlimited II from 1997.

 

The superior performance is mostly thanks to the sim being very basic at the moment. Compare the number of 3D objects being rendered at a typical OrbX airport in P3D, or the crazy number of autogen buildings in X-Plane's "plausible world", to the few scattered 3D objects that shown up in AFS 2. It's basically just an elevation mesh draped with satellite imagery, which is not very demanding for the computer to render. With only basic systems being simulated, no ATC etc., the sim is simply doing less, thus performing better.

If they would add the ability to input OSM data like world2XP it'd be perfect because OSM works great with photo textures and would add autogen buildings to replace the flat world. What I don't get is why people expect the default planes to have PMDG level detail. FSX's default planes weren't anything special,and Aerofly's have about the same level of systems modeling. 

Great setup!

I for one will buy this because I think the experience will be very immersive!

 

So what if systems depth is not so great - the airflow modelling by comparison seems to be much better than FSX, at least in AeroFly 1.

 

Trouble with most posts is that everyone assumes their own preference should be exactly what any new sim should provide.

 

Well to corner enough customers it would need to do everything perfectly which wold be economically unvianble - or else compromise.

 

Me I'll settle for superb graphics and will await eagerly whatever future developments will come our way ...

 

 

The superior performance is mostly thanks to the sim being very basic at the moment. Compare the number of 3D objects being rendered at a typical OrbX airport in P3D, or the crazy number of autogen buildings in X-Plane's "plausible world", to the few scattered 3D objects that shown up in AFS 2. It's basically just an elevation mesh draped with satellite imagery, which is not very demanding for the computer to render. With only basic systems being simulated, no ATC etc., the sim is simply doing less, thus performing better.

 

Until we actually see it running with tons of buildings, pretty much every bit of this is unsupported supposition. We'll have to wait and see.

Until we actually see it running with tons of buildings, pretty much every bit of this is unsupported supposition. We'll have to wait and see.

 

It's not unsupported. It's a fact that rendering more 3D objects reduces performance, and rendering fewer results in better performance. It's simple maths. The less work you do per-frame, the more frames you can output every second (FPS).

Even FSX performs great if you're using something like Megascenery with no autogen.

 

They're clearly going for the continuous satellite imagery approach, which dictates what they can and cannot do. It's the same "generic" vs "photoreal" argument that has been going on since the 1990's. Both methods have advantages and disadvantages, but they are distinctly separate techniques.

 

We can wait all we want, but we will only see refinements to their current approach, unless they completely change their approach to scenery in the next 6 months (it's supposed to come out of Early Access this year) and go for something like X-Plane's "Plausible World" or FSX/P3D's landclass+vector data. I don't think they would compile 30GB of photoreal data and write an engine to support it, only to throw it out the window at the last minute and switch to something completely different...

It's not unsupported. It's a fact that rendering more 3D objects reduces performance, and rendering fewer results in better performance. It's simple maths. The less work you do per-frame, the more frames you can output every second (FPS).

 

In general terms, absolutely. But you can't prove it will actually start stuttering until you test it. Obviously the performance decreases as you raise the number of 3D objects, but this does not mean the experience will be necessarily bad. Only a test can provide conclusive statements. Until then, it is only speculation.

It's not unsupported. It's a fact that rendering more 3D objects reduces performance, and rendering fewer results in better performance. It's simple maths. The less work you do per-frame, the more frames you can output every second (FPS).

 

Its unsupported. You're taking a generally known fact and then reaching out to apply it broadly to a completely different and more modern 64bit architecture, likely designed to make much better use of today's computer hardware.

 

There's simply no reference point available to infer the performance of this new engine yet, and any attempt to do so is unsupported supposition based on analogy with a much older engine.

 

Its like saying "My 1929 ford only goes so fast, and that proves this 2016 vehicle can only go so fast as well."

 

Yes there are limits, but I doubt FSX or anything derived from it defines them.

 

I don't think they would compile 30GB of photoreal data and write an engine to support it, only to throw it out the window at the last minute and switch to something completely different...

 

I'm not sure of the relevance..... Whats to stop them from taking advantage of osm technology just the same as some X-Plane photoscenery? Especially when Ipacs has explicitly stated that's this is exactly what they intend to do eventually?

 

Its already proven to be a powerful combination in the right hands. And the tech is ripe for enhancement and further refinement

 

It's not unsupported. It's a fact that rendering more 3D objects reduces performance, and rendering fewer results in better performance. It's simple maths. The less work you do per-frame, the more frames you can output every second (FPS).

Even FSX performs great if you're using something like Megascenery with no autogen.

 

They're clearly going for the continuous satellite imagery approach, which dictates what they can and cannot do. It's the same "generic" vs "photoreal" argument that has been going on since the 1990's. Both methods have advantages and disadvantages, but they are distinctly separate techniques.

 

We can wait all we want, but we will only see refinements to their current approach, unless they completely change their approach to scenery in the next 6 months (it's supposed to come out of Early Access this year) and go for something like X-Plane's "Plausible World" or FSX/P3D's landclass+vector data. I don't think they would compile 30GB of photoreal data and write an engine to support it, only to throw it out the window at the last minute and switch to something completely different...

They have said they looked into OSM but the data for the US is poor. As for buildings killing performance they do hurt it but they hurt it way more in FSX then they do in Xplane. Personally I think we need to step away from the FSX engine it's 10 years old and still doesn't perform well even on top systems with planes that arent that complex systems wise (Carenado for example) there is somthing wrong at its core that makes it run that way. I'm tired of having to deal blurry ground textures just a couple of miles from the plane because FSX or P3D cant handle loading more of the full res textures due to OOMs.

 I'm tired of having to deal blurry ground textures just a couple of miles from the plane because FSX or P3D cant handle loading more of the full res textures due to OOMs.

 

This kind of gross over exaggeration isn't helpful to discussion. 

 

I have no problems loading full res textures in P3D without OOMs with lots of add ons with no blurry textures. I currently have zero tweaks in the .cfg as well. And to answer the obligatory "but do you do it flying the 777!!!" response, no I don't fly that thing and it's not a litmus test for me.