P3D 3.0 Speculation & General Nonsense

[n4gix 4,947]

DTG may well achieve this but...as it based on MS FLight we shall see...

I guess you missed Aimee's post where she categorically stated that the new sim DTG is developing is not based on "MS Flight."

[FLAPS15 8]

I guess you missed Aimee's post where she categorically stated that the new sim DTG is developing is not based on "MS Flight."

 

Indeed I did! thanks

[Bobsk8 12,586]

I guess you missed Aimee's post where she categorically stated that the new sim DTG is developing is not based on "MS Flight."

 

What is it best on "TrainSimulator"?  

[Pugilist2 878]

That is actually disappointing news to me.  I think Flight had a lot of potential in the right developers hands.  Lets hope it's not based on a reworked Flight Simulator X code.

[N400QW 205]

Lets hope it's not based on a reworked Flight Simulator X code.

What should it be? Every sim anyone flies these days has that as it basis. The exception of course is X-plane.

[Bobsk8 12,586]

That is actually disappointing news to me.  I think Flight had a lot of potential in the right developers hands.  Lets hope it's not based on a reworked Flight Simulator X code.

 

I totally agree with you and flew it all the time, but when Microsoft bailed out on it too, I decided that was the end of that and got it off my hard drive. The flight dynamics in flight were very realistic, and it looked good, but all the FS9 and FSX fans on the different forums wanted it to be just like what they were used to, and trashed it every chance they got. So what happened, MS said "the heck with this" and pulled the plug. That is why I hate to see these LM bashing threads, because they could do the same thing in a heartbeat. 

[Guest]

Primarily because the more powerful computing of programmable GPU's have been harnessed and are leveraged in those games to take advantage of their remarkable abilities of their Memory Speeds, Rasters, and the texture and pixel fill rates. It's not the hardware, it's the software, and with FSX and P3D in their current form, they just weren't built for it.

 

Not this again ... take a close look at those other games you listed ... notice the distances everything's blurred out and static?  This is how developers manage FPS by governing the draw distance in a restricted scene.  That's the primary difference is draw distances.  P3D is DX11 shader based, GTA V is DX11 shader based ... they both use compressed textures and tessellation, and they both work on quad-tree design.  

 

As you increase draw distance your GPU/CPU loads increase exponentially (as does VAS ... that's why all global simulators have a LOD radius).  I don't see any "volumetric fog" in GTA V?  Weather is always the same canned weather pending the scene I'm "following"  GTA V is highly optimized to work within it's limited world, it's NOT a global simulator.  Also, where is the SDK in GTA V?  If it has one how flexible is it -- does it support individual aircraft models.

 

Most of these debates are really about "not accepting the limitations of one's hardware" because some completely different type of game works well on the same hardware ... this is the equivalent of racing a 125cc shifter Kart on a tiny < 0.5mi track and beating a F1 Ferrari ... now move that to say a 3 mi track and watch the F1 Ferrari beat the 125cc shifter kart ... see what happens to GTA V if you bump the view distance to 110 mi or more ... you'll be measuring performance in terms of seconds per frame and not frames per second.

 

Event GTA V, Twitcher 3, etc. etc. ... they do stutter, turn up the AA (just like in P3D) and graphics settings and they will stutter away also.  What GTA V and similar games do is build a world that is highly optimized for a view distance ... everything from the object geometry (increase geometry edge segments to reduce AA issues) to textures are highly optimized.  One could do the same with P3D ... some 3rd party devs are doing exactly that but it's an incredibly time consuming process and the revenue return on work investment has a finite limit.

 

Yes it's about how pixels are set in a physical 2D coordinate system ... to obtain "true" fluidity -- defined as any scene motion will never exceed more than 1 adjacent pixel change ... I did a long response about this in another thread a year or two ago ... basically at 500 Kts 200 ft AGL looking 90 degrees (side) you would need about 240 fps and a monitor that can operate at 240 Hz ... it's a long complicated formula but that's about what it came out to be IF you wanted "True fluidity".

 

Best way to validate what I'm talking about is to set up a situation (500 kts, 200 ft AGL 90 view out of VC) where you are locked to 60 Hz (60 fps) use ShadowPlay to record at 60 FPS ... now use a good video editor (i.e. Adobe PP CC 2015 or equivalent) ... playback the recording moving one frame at a time (pending your video software zoom in to pixel level and note the position of an object, now move one frame and note the position of the same object ... it would have shifted about 4-5 pixels ... this is NOT true fluidity.  For it to be true fluidity moving one frame would result in the object moving one pixel.

 

And yes, obviously pixels don't move in the 2D physical coordinate, they change color and how that color is changed is by MANY render passes to produce a final frame.

 

 

 

Xplane actually does a fairly good job of this today.

 

Not on my PC isn't doesn't ... if I turn on HDR, performance and AA are not acceptable (worse than P3D) ... the only way I can fly XPlane to reach just 30 FPS is with HDR OFF and clouds draw distance set to VERY short (like 10 mi max).  But as with most global world simulators, I've adjusted graphics to meet my goals of a steady 30 FPS.

 

The argument "my favorite 3D Shooter runs well, so P3D should run well" is just not valid ... like I said, when that 3D shooter starts drawing out to 110mi to 150mi or more then lets talk comparisons.

 

Cheers, Rob.

[WarpD 824]

Thanks Rob... I remember you had posted about this before. But of course they really refuse to believe me.

[Bobsk8 12,586]

It's nice to read a post from someone who really knows what they are talking about instead of some of the garbage that the pseudo experts come up with. 

[vgbaron 3,509]

Thanks Rob... I remember you had posted about this before. But of course they really refuse to believe me.

That's because you're a developer Ed. NOBODY believes a developer!! 

 

Seriously, it really amazes me how many times I read the SAME old argument making comparisons between FSX/P3D and GTA, Warcraft, etc.  The reality of a limited world and a global world gets completely ignored every time.

 

Vic

[KenG 492]

 

 

Not on my PC isn't doesn't ... if I turn on HDR, performance and AA are not acceptable (worse than P3D) ... the only way I can fly XPlane to reach just 30 FPS is with HDR OFF and clouds draw distance set to VERY short (like 10 mi max).  But as with most global world simulators, I've adjusted graphics to meet my goals of a steady 30 FPS.

 

The argument "my favorite 3D Shooter runs well, so P3D should run well" is just not valid ... like I said, when that 3D shooter starts drawing out to 110mi to 150mi or more then lets talk comparisons.

 

Cheers, Rob.

 

Too bad, worked fine on my system, just did not want to pay again for all the add-ons I already own. Then again I am not trying to sell people anything or have emotional attachment to a paticular sim.

 

I never discussed 3D shooters. I was discussing what I use at work where I train real pilots. No mystical 240hz projectors and Cray Supercomputers there. Seems the FAA is satisified that real pilot training can occur on something much less sophicated. I don't see then why entertainment software would require such outragious requirements. Considering that I have had the FFS off motion and when I commented, time to go on motion, I have had a number of pilots exclaim they did not realize they weren't on motion. I don't think we even need that perfect pixel you mathmatically computed. The human brain is so much easier to fool.

 

In the end I don't fly an aircraft based on the movement of a single pixel. I use the sum of all feedback to execute control movements to adjust my flightpath. I don't care if the programmer used techniques to trick my brain, good for him.

 

Maybe someday we will have the perfect artifical flight environment, but I am betting I will see it on our $17 million training devices long before it hits your $3,000 PC.

[Bobsk8 12,586]

Maybe someday we will have the perfect artifical flight environment, but I am betting I will see it on our $17 million training devices long before it hits your $3,000 PC.

 

 

I have flown $30+ million dollars sims made by CAE, and next to them, X plane is a nintendo game, and not a very good one at that. 

Edited August 24, 2015 by n4gix
Please stop with the FULL quoting!

[N101ER 0]

Maybe someday we will have the perfect artifical flight environment, but I am betting I will see it on our $17 million training devices long before it hits your $3,000 PC.

 

The Lear 45 sim, I used to teach in, was a modern electric sim first certified in late 2009.  It used 8 MS Windows computers networked through a switch to operate.  Because of FAA certification issues the video cards were GTX 480s (3).  Sim did many things well but daytime visuals were not very high resolution.  It was easier to do circling approaches at night because of approach and runway lights were easier to identify as runway environment.

 

Mike

Edited August 24, 2015 by n4gix
Please stop with the FULL quoting!

[Guest]

Too bad, worked fine on my system

 

I've got XPlane working on my system too, I've just had to turn OFF more options than I had hoped to get my goal of 30 fps and good time frame consistency ... just like all simulators I've used since the 1980's, it's been a process of compromises regardless of hardware used.

 

 

 

I don't think we even need that perfect pixel you mathmatically computed. The human brain is so much easier to fool.

 

I agree we don't ... that's why I'm puzzled by why so many people insist on it?  My flight experience is perfectly acceptable at 30 Hz / 30 fps so long a my time between frames are consistent with no wild variance and it doesn't hamper input control.

 

I believe the FAA require 60Hz / 60 fps (but don't quote me ... Ed would probably know this info).  I've flown RedBird motion simulators and they are most definitely "missing" world details, but they get the basics done in terms of objects to avoid (powerlines, towers, etc.) and weather conditions.

 

 

 

Maybe someday we will have the perfect artifical flight environment, but I am betting I will see it on our $17 million training devices long before it hits your $3,000 PC.

 

I have no idea ... estimated development costs to bring even FSX to the point where it was in 2006 was around $100 Million over the years.  End user global environment has considerably more buying potential the strictly single purpose commercial.  I recall the days when Apple purchased Shake (from Nothing Real) that was selling it for about $10,000 per license and Apple ultimately dropped that to $500 (when I bought it) and is now EOL ... Shake's revenue was higher with the product at $500 than at $10000 a license.  Point being, I'd never ignore the global end user buying potential vs. single purpose commercial. 

 

Sorry, didn't mean to suggest you had commented about 3D shooters, that was more a "generic" statement than one aimed specifically at you.

 

Cheers, Rob.

[FaxCap 13]

I have flown $30+ million dollars sims made by CAE, and next to them, X plane is a nintendo game, and not a very good one at that. 

Funny you should mention that....I was in the 767-200ER SIM at CYVR.....the graphics were about FS200 level.

I remember thinking my FSX graphics are much better than this.  :smile:

 

FaxCap