The New Faster, Leaner Next Generation Flight Simulator SIM-Posium is Now Available for Your Input

[srborick 0]

Good stuff.... and yes copied and pasted along with several other items of worth from this thread... 

This is indeed the kind of information that is helpful.

 

n4gix... we may have met... if you recall the meeting where one guy raised his hand and asked if MS intended to include force feedback, that was me. My wife just the other day cut up my T-Shirt for rags, but saved me the logo. 

 

OH... and I did send Bruce an email that very day...

 

Stephen B.

 

___________________________________________________________________________________________________

dtmicro... are you sure you don't want to take a look at the UNIGINE Sim eval kit? We have access until Sept 30th

 

Here's a piece of an email I received from the UNIGINE Sim rep just yesterday...

 

When I get back To my computer I'll send you a roadmap that will be of interest to SIM-posium as it pertains to terrain tools and GIS data support.

[HiFlyer 9,360]

dtmicro... are you sure you don't want to take a look at the UNIGINE Sim eval kit? We have access until Sept 30th

 

It would indeed be interesting to see what you think. I thought a lot of clever things were being done to keep things fast and save memory (I've seen the same basic techniques in outerra) but the outdoor illusion seemed to be right on the bleeding edge of breaking down to me. A lot of just-in-time loading going on, causing hitches.

 

I was unable to determine if it was the engine, the incomplete state of the demo, or my computer that was causing some of the issues.

 

This demo might be at its best on an SSD.

[dtmicro 1]

I am trying to review all the engines, but it's going to take me a long time. It is important to see if Vue D Esprit textures still retain a nice look at lower-resolutions, and which terrain generator will most easily be the most "morph-friendly" into aerial images with blending tricks. I basically envision the first environment as a real world GIS mesh, with Vue or Terragen on top of it for lower-res, use those textures and blend them into the aerial images.

 

I say morphable because I think the first attempt would be to render the ground with shaders based on inheriting the pixel colors of the high-res aerial imagery, and then use the default shading generated from the aerial on the ground level and modify further.

 

That way you still retain a realistic look of an area and appropriate colors. Now I know what immediately raises a red flag here, aren't we trying to get away from the terrible coloring of Aerial imagery, yes very much so.

 

The one advantage to this method however is that the color tables in the aerial imagery usually are fairly consistent (if anyone knows Excel, think in terms of absolute vs. relative referencing). Even though as a whole the colors are terrible or far off, as they are remastered there is a "believable" consistency to acheiving real life color. Also you'd only remaster a small amount of aerials (not nearly the whole world), then you'd generate the ground textures from the land class attributes like Outerra does (sort of but I envision a simpler way to do it).

 

An application like Vue D Esprit could save incredible amounts of time, especially since they give you so much vegetation stuff, it's kind of expensive now and last time I used it was (umm 10 year ago?), maybe longer?
 

See the thing is that our hardware is just now getting to the point where I wouldn't have to downsample the images so much as to ruin them. We had really good terrain generators for years and years (Vue and Terragen), but it was unusable in a gaming environment. Now it's getting to the point where you can use some of it.

 

[HiFlyer 9,360]

Uhmmmmm, Yah........

 

*Blinks at video*

 

My hard drive is crying even thinking about that many textures. I think its going into shock......

[dtmicro 1]

Uhmmmmm, Yah........

 

*Blinks at video*

 

My hard drive is crying even thinking about that many textures. I think its going into shock......

 

Yah, that's why I see it as a morphing mode, as you get to around 1000-2000 feet, other than the buildings, the ground textures will smear into the aerial textures with a luminance filter and blending mask. I do think I can make higher-res low-altitude ground textures using one of these tricks. As you are flying low, there is going to have to be some type of blur mask or lower-res textures in the distance beyond just standard MIP MAPs, and the highest res limited to directly above the ground as GTA V does.

 

I however don't want to make higher up altitudes look anything like GTA V does, GTA V actually looks pretty bad from high elevations (try it), I think JC2 looked better in that respect.

 

I'm not sure the best trick yet, but I know several tricks on how to do it. I don't know because I've never tried blending close-up ground textures into a higher-up aerial image.

 

Using aerial images I could also create the roads without the Open Street Vector data initially just for a prototype in a smaller area, so save me a lot of time.

I will post a screenshot of what I want to acheive in a bit.

Uhmmmmm, Yah........

 

*Blinks at video*

 

My hard drive is crying even thinking about that many textures. I think its going into shock......

 

I picture the game being about 1 Terrabyte before downloading the expansion areas

[srborick 0]

dtmicro

 

You have at least got to go onto the SIM-Posium and download the tech demo... I think you'd be rather surprised at the results as I saw similar scenes as well as lighting effects as I was moving around the 50 km^2. 

 

at the very least look HERE

 

Just in case you missed this... cuz I added it as an edit previously...

 

 

When I get back To my computer I'll send you a roadmap that will be of interest to SIM-posium as it pertains to terrain tools and GIS data support.- Unigine Sim Support

[dtmicro 1]

Yah, those screenshots look very very good.

I guess a big decision to make is whether or not to completely abandon the high-res aerial imagery, but if you do that, you're going to lose a lot of the reality of an area, especially as you get higher up it's not going to look like the real area.

 

It may actually be better just to abandon aerial imagery all together.

 

EDIT:

I'm looking into the Unigine engine now, PM me the GIS road map. It looks promising from what I've seen so far.

 

Here is how I would want desert rock textures to look for a game :(

http://www.mediafire.com/view/l7wpznnt7layp9y/Utah_ReferenceImage.png#

 

--CANNOT POST A PNG FILE, so you'll have to click the above link to download.

 

Here is a better remastered 1.2m NAIP image than the previous images I posted, this is literally at the END of the ROPES when it comes to 1.2m NAIP imagery, and not everywhere can look this good.

 

The only reason I post that, is it gives us a visual reference as to what the possibilities are with the poorer quality aerial imagery... It looks good enough for CURRENT-GEN, but not for NEXT-GEN. This is 1.2m, but even 60cm would have a similar limitation for a NEXT-GEN SIM, it is just that you wouldn't see it until you go below about 2500 AGL.

For land-class though, you'd always ONLY want to use 30cm or better and you MIGHT get away with downsampling a 30cm to as low as 60cm (then re-sharpening), never 1.2m, remastered or not... Only if you were going to go with a pure Photoreal area (no repeating textures), could we live with 60cm or worse source imagery...

So the above image would do us no good really.

 

So if we didn't use any aerial imagery, at least we wouldn't face the issue of re-tiling land class from one aerial image that was from a different part of a state. Otherwise, we would have to use the limited 30cm data coverage and then re-tile it.

[Wobbie 938]

Well done Stephen for the site, for opening dialogue & for piloting this discussion.

A personal view of what interests me..

Well, in P3D & FS9, I can sail ships (thanks Milton & team), drive on the roads, drive a submarine & even fly. There is a whole bunch of local (for me in South Africa) freeware scenery that adds to my enjoyment. So, what more do I need?

I quite enjoy reading all the wish lists & what people 'demand' from a new sim/game. We don't all need all the bells & whistles, some of us also want to have fun.

 

A big distinction should be made for going the realistic route OR the immersion route..

What actually do we want? In my instance, we built up 2x simulators, an Avro Shackleton sim & an Aeromacchi MB326 twin seater sim. (this aircraft is known as an 'Impala' here in South Africa & was used as a trainer by our Airforce)

 

Why have I mentioned this? Well, we are using FS9 in the sims, because of freeware scenery for our country, as well as ease of use.

 

When we had real pilots flying the sim, & we saw that they were 'bumping' or bracing themselves in the seats (real ones), we knew that we had hit the 'soft spot'as far as immersion goes, EVEN thought, as far as realism goes, we have our instrumentation on a screen, & use a single projector for scenery.

 

So, maybe what is needed, or would be nice to have, would be a basic sim, for newbies. Then an 'Intermediate model' (optional) for those that want something better, then an 'Expert' module for the 'PMDG' type guys that want all the bells & whistles.

 

From what I'v read here, everybody is looking for the holy grail.. It's nowhere out there, as far as a complete sim goes. We all have diferent needs & expectations.

The gamers, the guys that just want to have fun (even sailing around) & of course the guys that wan every button working.

 

Also, do not forget the developers, both freeware & payware, that have so many choices of what to develop for. From FS9, the various FSx's, P3D's, the reportedly future 64bit P3D, X-Plane & the rest. We have never had so many choices. 

Obviously there is much talk of our hobby becoming diluted & fragmented. That is a debate in itself.

 

We need to get more people into our hobby, obsession, or whateve we call it. Somewhere out there are the future developers.

 

So, my personal thoughts are as I mention above.. an entry level nursery module to get people interested without scaring them off, an intermediate module & an 'expert' one. There is no 'One size fits all'.

 

Just a few thoughts......     :p0503:  

[dtmicro 1]

I think the basic model vs complex realistic model (as far as plane models go) is more up to the add-on developers than it would be to the developers building the sim.
 

1) As far as ground scenery textures type goes, it comes down to availability. Anything better than 1.2m NAIP is only covering mostly the east coast, and even sometimes the 30cm stuff that is available has pretty OFF COLOR as well (or was taken during a bad time of year for vegetation)... Yes you can get 30cm for the west coast, but it's mostly around the city areas (Denver, LA, whatever)... And, some of that stuff isn't perfect either, still has to be gamma, exposure, contrast, and color-table corrected to some degree, especially if making ALTERNATE seasons from one set of PR scenery.

 

It is still (in some cases) more pleasing to remaster aerial images, but aerial images are inconsistent and to maintain consistency is incredible amounts of labor even with a full set of automation tools like I have (and I have everything needed, including almost fully auto-water mask injection). To try to use tools to build similar textures such as land-class type textures straight from generators (which will become real repetitive if you are not VERY careful). That is why we are still in the alternate phase of experimental thinking, almost as if our thoughts are transmutating dimensionally about the graphic formats and how to intermix to achieve the best results overall.

 

If using say Terragen or something for non-PR like scenery, meaning you are not going to store 100GB+ 30cm for an area, the only option is to extract the textures and try to resurface using repeating land-class textures that are only based on the Terragen texture, while keeping the terragen and FSX results open side-by-side.

 

When doing this method:

The checkerboard effect is a problem I've had before doing that method, it limits the type of textures you can do. Yes I know like in Instant Object Studio you can even make textures and they have blur-ties at the edges to try to get rid of that effect, but it only works so well. It's not that you cannot get rid of it, it's just the effect limits your ability to texture things. Of course I can do that in Photoshop or even Terragen itself (but in old mostly plugins, I haven't used Terragen since a way older version, same with Vue). I cannot speak for the current version.

 

GTA V definitely has the checkerboard issue too (so I cannot be critiqued for that issue fully), but they hid it somewhat ok. GTA V has very nice rock textures and beach textures, it's the grass textures at a distance I don't care for, but their grass is fine close up (great).

[TooWings 17]

What I've noticed (StephenB,  and perhaps others have, is that you're attributing much more value to your efforts (and to yourself) than would otherwise be warranted.  Merely setting up an Internet forum isn't going to let you or any other lay person change the direction of flight simulation, any more than it would for brain surgery.

 

I totally agree - 100%. 

..Like I've said before, this thread and Stephen's Sim-Posium site are nothing more than the discussion/chat threads we've enjoyed contributing to over the years..the difference is that the Sim-Posium site has unfortunately failed to attract anything more than a handful of regular contributors -while this AVSIM thread has many hundreds. 

...But at the end of the day it doesn't matter as it's evident to many that nothing concrete will come of it, and that we sadly wont see the nex-gen flightsim / world-sim on the market  through Stephen's and co's efforts no matter how many video clips and tech demos get posted here..  But it's been an enjoyable debate nonetheless. So thanks everyone..and bye!.

[dtmicro 1]

I totally agree - 100%. 

..Like I've said before, this thread and Stephen's Sim-Posium site are nothing more than the discussion-chat threads we've enjoyed contributing to over the years ..the difference is that the Sim-Posium site has unfortunately failed to attract anything more than a handful of regular contributors -while this AVSIM thread has many hundreds. 

...But at the end of the day it doesn't matter as it's evident to many that we sadly wont see the nex-gen flightsim / world-sim on the market  through Stephen's and co's efforts no matter how many video clips and tech demos get posted here..  But it's been an enjoyable debate nonetheless. So thanks everyone and bye!.

 

OK fine - I'll tell you what it was, I wrote an app for the funny farm, and looking at the database 20% of the people in here are in my APP's database!

 

If we really can find enough passionate people, we can in fact get something going. It is all about motivation, forming the right team, proper management techniques, funding, and what not.

 

I currently pay for server racks with about 20u space still OPEN and available for use and I have plenty of hardware already, though would need to pay if the bandwidth use creeped up too high. Tell you what I'll do to shutup the nay-sayers, I'll donate 24 to 48 cores of 24/7 off-site rendering capability running at a Tier3 Data Center to anyone that wants to help me, that has the right skillset. It's a start... There now I'm doing something...

 

The problem is being one person in this stuff is too hard, need more people to help.

[codechris 9]

Well that is very kind!

[dtmicro 1]

Exactly my point! ..and thats why the nex-gen flightsim / world-sim will never come to fruition. Sure you might 'do something', tinker round the edges...but it wont amount to a successful retail flightsim getting onto the market.  I'm not being a naysayer here, I'm just echoing what many others have said...injecting some 'real world' into the debate as someone wrote recently.

 

We aren't trying to build a successful retail flight sim right away, just evaluate the market potential and get some funding.

 

I personally believe srBorick has done a nice job from a Macro perspective already, he has identified the Unigine engine to me, as I did not know about it. He has also noted some key points and pitfalls about other engines in his page, even though he has been very humble in identifying and admitting his own shortcomings in not being a developer. The problem is in today's world people expect everyone to act like a politician and just LOBBY for more war. Because there are plenty of forum war lobbyists in this thread right now, and it will not deter any efforts one bit if we do decide to move forward.

 

I have no problem with his web-site, looks fairly nicely done to me, and well organized. He is pitching himself as an organizer after all, and he seems to have that skill (I can see it), so no issues here with him.

[TooWings 17]

If we really can find enough passionate people, we can in fact get something going. It is all about motivation, forming the right team, proper management techniques, funding, and what not.

 

Exactly my point! ..and that's why the nex-gen flightsim / world-sim will never come to fruition. Heck its only been a couple of weeks into the project and its a bit of a shambles already for various reasons already mentioned by myself and others.

.  

Sure you might 'do something' - you'll tinker round the edges, but it wont amount to a successful retail flightsim getting onto the market.  I'm not being a naysayer here, I'm just echoing what many others have said...injecting some 'real world' into the debate as one experienced commentator wrote recently..

 

 

 

We aren't trying to build a successful retail flight sim right away, just evaluate the market potential and get some funding.

 

Yes I know ...but you're still just 'tinkering around the edges' of the project, as I said.

 

Anyway, have fun.  It all learning.

 

Meanwhile we'll watch this space.

[codechris 9]

Well you could argue all we need to do is lay some ground work, then maybe someone will notice and pick it up, in a more professional capacity.