Print Page | Close Window

New Product ‘X’

Printed From: SimForums.com
Category: Ultimate Terrain and XpressSim Products
Forum Name: Ultimate Terrain And XpressSim Support
Forum Description: Support for the Ultimate Terrain and XpressSim Product Line
URL: https://www.simforums.com/forums/forum_posts.asp?TID=30107
Printed Date: May-12-2021 at 4:50am
Software Version: Web Wiz Forums 12.01 - http://www.webwizforums.com


Topic: New Product ‘X’
Posted By: akriesman
Subject: New Product ‘X’
Date Posted: April-11-2009 at 11:47pm

For the last few years we have been listening to some of the things our users have been requesting.   Mainly, these are things that will greatly improve VFR flight in FSX and allow users to more readily recognize local landmarks (i.e. shopping malls,  strip centers, schools, etc).  And, we have long heard requests to actually have houses that follow accurate roads, instead of having housing autogen in FSX that follows the “fake” roads in the landclass textures.

A lot of people have been wondering what we could possibly be up to for the last 6 months.  Well, we have been very, very busy working on a new product development system (internal software).   This system will allow us to create large, semi-detailed scenery areas with a reasonable amount of manual work.

Each scenery area that we will release will provide:

1) Semi-Custom objects that are accurately sized, shaped and positioned according to there real world positions.   We call them semi-custom because the textures are generic, even though the building shapes and locations are accurate (generic textures will help performance).

2) Generic buildings from TeleAtlas POI (Point-Of-Interest) data.   These are buildings that are placed using POI data and oriented to face the accurately positioned roads.  So, you will see a generic "Auto Parts" building where one exists in real-life (or close to it).  Same for churches, restaurants, etc.

3) Finally, you will get housing terrain imprints and actual housing objects that follow accurate roads.    Urban landclass is completely removed.  So, you will no longer see the "fake" roads that are embedded in the landclass textures.  The only roads you will see are accurately positioned ones.

For years we have been discussing this potential product line.  Internally, we called it Ultimate Objects.  Since the project does not have an official name right now, I am just calling it project ‘X’ in this thread.

The goal of scenery packages we create with our new system is not to provide incredibly detailed buildings at close range.   Instead, we are going for "perception of reality" from VFR flights.   So, users familiar with an area will see something that looks very close to reality, without completely hammering frame rates.   Things like large shopping malls, universities, high schools will all be custom drawn in a very efficient manner.   Other areas based on the POI data used will have appropriate generic buildings placed.   In residential areas, you will finally be able to locate your street and see houses following it, although the houses are generic and not spaced like they would be in the real-world.

At this time, there is no information available as to when products will be released with this new system, how they will be packaged or how they will be priced.    One very probable situation is that we include the technology in our upcoming Ultimate Terrain X Alaska package as a kind of a demo.   Alaska might work well because there will be very few urban areas.   So, there would be no need for a separate Alaska product using this new technology.

The new product series will work terrific with Ultimate Terrain X.  Although, Ultimate Terrain is not necessary.   This new product is kind of like an Ultimate Terrain for buildings.   Ultimate Terrain provides the ground terrain features.   This new product provides the 3D objects to work with the accurate terrain in Ultimate Terrain.

Our internal test scenery area is a 2,500sq kilometer area around Tyler, Texas.  I will post a couple comparison shots showing the new technology.  But, please remember that while the system is in place, a lot of the textures and building repetitions need to be improved upon.    I just wanted to show you guys what has been going on to build some excitement.

Enjoy !   Screen shots to follow (hopefully they will post today).

 



-------------
Allen




Replies:
Posted By: akriesman
Date Posted: April-11-2009 at 11:50pm

OK.  Here is the default Tyler, Texas.   Tyler is a 100,000+ population city.  As with most cities in FSX, Tyler has been largely forgotten.    Here is the FSX version of Tyler.



-------------
Allen



Posted By: akriesman
Date Posted: April-11-2009 at 11:54pm

Here is our version of Tyler, Texas.   You will notice a noticably present downtown area, stadiums, storage tanks, businesses, schools and accurately placed residential roads with houses.



-------------
Allen



Posted By: akriesman
Date Posted: April-12-2009 at 12:06am

Here is a shot that demonstrates what we call "Semi-Custom" buildings.   These are buildings that can be quickly drawn using our new development system, but still provide objects in the correct size, shape and positioning using generic textures.   So, from the air, they should look very accurate.

With our new software, we can create and position about 150-180 of these objects an hour, which makes this kind of product feasible.



-------------
Allen



Posted By: akriesman
Date Posted: April-12-2009 at 12:11am

Our new development system has a feature called “dynamic object positioning”.    This allows us to create things like apartment complexes, warehouse districts and even downtown areas without having to draw out each object by hand.   Instead, we mark off an area and the software populates it with appropriate buildings aligning the roads.   This will allow us to included city centers for even the smallest cities and include them in a product covering a very large area.

Here we have a couple of apartment complexes using the dynamic object technology.



-------------
Allen



Posted By: egntpilot
Date Posted: April-12-2009 at 3:29am
Hi,

that looks very impressive, will look out with interest for more details.


Steve


Posted By: Axelb9
Date Posted: April-12-2009 at 11:03am

Hi,

This is nothing short of revolutionary! Been thinking about this idea for a long time - obviously I have zero skills to actually pull it through so it is great to see that you are working on it. The screenshots demonstrate excellently the difference - finally cities might actually take shape replacing the current blob of buildings and the archaic landclass system at least for towns.

Finally also a tool that could solve the issue of autogenless photoscenery. BTW how are the green areas addressed inside a city such as parks, greens and trees alongside the roads?

I am really looking forward to see the demonstration of the technology in UTX Alaska.

Thank you,

Alex



Posted By: akriesman
Date Posted: April-12-2009 at 12:50pm
Originally posted by Axelb9 Axelb9 wrote:

Finally also a tool that could solve the issue of autogenless photoscenery. BTW how are the green areas addressed inside a city such as parks, greens and trees alongside the roads?

Thanks Alex.   I am glad that you understand what it is we are trying to accomplish.

I actually started looking into this type of thing several years ago.   But, it has taken that long to work out a system that looks good, can be accomplished with reasonable amount of manual labor, and won't just kill frame rates.

City parks will mostly be provided by UTX, as they are now.   As we work on individual areas, we will add more detailed ground polygons as necessary (like we do with the parking lots here and other underlying terrain areas).    More or less trees could be added by changing a vector autogen entry associated with the houses.

Yes, this should work very well with photoscenery also.  Dean Mountford (FS Dreamscapes) and I are in touch often.   So, I can see a lot of our stuff working well with photoscenery.   Exept for the semi-custom drawn buildings, most other buildings won't overlay the photoscenery perfectly in most cases, because the precision of what we are providing won't be exact (but it is a big improvement over urban landclass).

This system is highly configurable from the users end.   So, you can do a lot of tuning for frame rates if necessary.



-------------
Allen



Posted By: David Wooddell
Date Posted: April-12-2009 at 4:32pm

Way to go team.

My areas to fly in FS are Alaska from Southcentral north and west, Upstate NY, Montana, and East Texas. After all, I live within 10 minutes of Tyler, TX. I have a reworked version of Tyler Pounds Field with the new terminal and AFCAD for UTX if this doesn't address that.

Mine also includes the Air Museum static display by the old terminal, control tower.

Keep up the good work, and I'm looking forward to this and UTX Alaska so I can update several sceneries to work with it.

David



Posted By: Brian W Keske
Date Posted: April-12-2009 at 10:06pm
Very interesting Allen. It looks great.

And as above....Alaska can not come soon enough for me Wink

You guys have spoiled me to the point that I cannot fly outside UT areas....and I have been missing my trips up to Alaska since FS9.....or....many moons ago.


-------------


Posted By: c152flyboy
Date Posted: April-12-2009 at 10:12pm

Yes, this is great news Allen.  i have been waiting for some information on this and it looks and sounds even better then i expected.  i lok forward to some more information as it becomes available, the screen shots are great to see as well.  I have to start saing up for this.  Tongue

Kyle



Posted By: kenthansen
Date Posted: April-13-2009 at 7:11am
Allen that looks awesomeClap

Nice not having the fake roads, makes it look a lot cleaner. Some of the major roads in my home town are incomplete without residential turned on, this will eliminate the clutter.

Have a few questions:

You mentioned you eliminated the urban landclass, what lc textures will be used in place of them as a base? And how will they blend into the surroudings? Maybe youn could post a pic of the border area of your test area.

How will this effect 3rd party landclass? Will Poject X override other landclass in urban areas?

How will this effect 3rd party textures such a GEX?

I noticed a lack of vegetation autogen will this be corrected?

I Know this is still in the production stage but looks very promising.





-------------
Asus P9X79 Mobo, Intel i7 3820 Quad CPU @ 3.8 GHz, 16GB DDR3 Ram @ 2133 MHz, GTX 760 OC 2GB GDDR5, 850w PS, W7 64 Optimized


Posted By: JimC1702
Date Posted: April-13-2009 at 10:55am

This looks very exciting!  I'll be keeping an eye on it for sure.

Thanks Allen.

Jim

 

 



Posted By: vwest
Date Posted: April-13-2009 at 5:19pm
It does look great. Smile

-------------
Victor


Posted By: akriesman
Date Posted: April-13-2009 at 5:28pm

"You mentioned you eliminated the urban landclass, what lc textures will be used in place of them as a base?"

We will use the most appropriate vegetative landclass textures that available as background images.   In this particular area, the best textures are forest textures or forest/field textures without any structures embedded in them (like farms).

"And how will they blend into the surroudings? Maybe youn could post a pic of the border area of your test area."

This should blend in perfectly with UTX and pretty well with FSX (withoug UTX).  The roads with houses will directly replace the urban/residential roads in UTX (which most people turn off now anyway).   So, most roads that are tagged as urban/minor roads in UTX will have houses with this product.   Since the urban/rural classifications in the commercial road data are not perfect, there will be some roads that have houses when they should not and vice-versa.  But, the overall improvement in accuracy over landclass will be huge.

Some day, we might reclassify all roads by hand to be urban or rural.  But, this is a huge project.  So, for now we just have to rely on the commercial source data classifications.

"How will this effect 3rd party landclass? Will Poject X override other landclass in urban areas?"

Yes, it will override other landclass packages.   If you have UTX, then you just turn off the city landclass layer (no need for the extra overhead).

"How will this effect 3rd party textures such a GEX?"

It won't.

"I noticed a lack of vegetation autogen will this be corrected?"

The replacement landclass logic was just recently added.   I believe the texture that I chose for some areas was missing autogen definitions.   You will notice that the forested landclass textures used around the city have a lot of vegetation.

There is also a slight possibility that the maximum number of autogen objects in FSX are being displayed (6,000 ?).   But, I don't think this is the case.

We still have a lot of work planned for the texturing.    So, while this is pretty close to what the finished product will look like, there is a lot of polishing to do.



-------------
Allen



Posted By: kenthansen
Date Posted: April-13-2009 at 7:03pm
Thanx for you explanation Allen.

I was just trying to get a idea of how this would integrate with existing adds.

Can't wait to try it out.




-------------
Asus P9X79 Mobo, Intel i7 3820 Quad CPU @ 3.8 GHz, 16GB DDR3 Ram @ 2133 MHz, GTX 760 OC 2GB GDDR5, 850w PS, W7 64 Optimized


Posted By: TomG
Date Posted: April-13-2009 at 9:38pm
Wow!!
I just posted to AVSIM about this same problem - buildings all over the wrong places in FSX with my correct housing development vector streets over these tiles. One response pointed me to this post. I just installed UTX, GEX, and FSGenesis Landclass. Here is an image of what I get in FSX. Those diagonal streets and buildings don't exist in my area. The actual streets are the ones I pointed out in the picture. Hopefully your product will solve my problem. It can't come soon enough! Thanks for the good news. Tom


Posted By: c152flyboy
Date Posted: April-14-2009 at 8:35am

Tom, i have the same issue with the GEX textures showing some real nice looking but unrealisitc areas, especially if i enabled the rural roads with UTX. that is why i dont use them, i would love to be able to see my street, but i cant stand the look of photorealistic scenery, it's way too 2 dementional looking for my taste.

Product X looks to eliminate this issue.  I sent NickN a e-mail seeing if he thinks product X will play nice with GEX.  sounds like he has been in touch with Allen about it but it is way too early to know from what Nick said.  I hope it does because if it does not i would think i would have to go back to the default FSX textures and that means i would have no more use for GEX.  Cry

Allen, can you offer any insight into this?   looking forward to futue information.

kyle



Posted By: dmountford
Date Posted: April-14-2009 at 8:54am
Allen, that looks great!!! You are a genius!!! I can't wait to see what else can be accomplished with this approach :-)   


Posted By: NickN
Date Posted: April-14-2009 at 10:06am
Originally posted by c152flyboy c152flyboy wrote:

 

I sent NickN a e-mail seeing if he thinks product X will play nice with GEX.  sounds like he has been in touch with Allen about it but it is way too early to know from what Nick said. 

 

What I said is we have not evaluated the design and that since we have never looked this over I have no direct frame of reference to comment on it.

From the description this would overlay default and therefore overlay GEX because GEX simply replaces the default textures and autogen files for those textures. Since GEX is textures only which rely on landclass to call them and this replaces the UTX Urban LC then it would overlay GEX in the area it is used, I assume, by providing a base tile system or calling a series of veg tiles the roads and buildings will overlay so the GEX urban/suburb tile underneath would not be visible and therefore roads would not appear out of alignment.

Also, if I understand Allen correctly they are still in the design phase with this so how that all comes together is still in the works. What was presented here is an 'overview' demo with specifics still to come.

 

 

 

 

 

 



-------------


Posted By: dharris
Date Posted: April-14-2009 at 12:58pm
Just when you thought it could not get any better, you guys keep on bringing out more things to spend money on I don't have!  Oh well, I will have to get up earlier now , to walk along the roads and pick up extra bottles for the refund.  Keep up the good work, and with all the helpful folks on these forums and the great products you bring to the table, I can hardly wait for these to come out.  Thanks again for all your hard work.TongueTongue


Posted By: BaronKen
Date Posted: April-14-2009 at 3:07pm
This looks great Allen, will be watching for further developments.

-------------
Ken


Posted By: akriesman
Date Posted: April-14-2009 at 5:42pm

FWIW, I have no doubt that GEX will further enhance project 'X'.   And, as Nick says, we are still a long way from actually having a released product using this technology (except for maybe Alaska as a demo).

What we do know, is that there is interest in what we are doing, it looks good even in the current state, and will not slaughter frame rates.

I will keep you guys updated and try and provide a few new screenshots at times.   I would like to come up with an official name for the product ASAP, instead of calling it project 'X'.    One possible, but some boring name would be "Ultimate Cityscapes", which would play off our Ultimate Terrain product line.

 



-------------
Allen



Posted By: NickN
Date Posted: April-14-2009 at 6:57pm

 

 

You guys know we always work to try and keep things seamless.. As Allen and Jeff develop they will keep us in the loop as they always have.

So please dont jump the gun on what will work and wont..   Allen just wants everyone to see this which addresses a very primary restriction about how vector scenery works. Their solution looks very promising Smile



-------------


Posted By: barnicles
Date Posted: April-14-2009 at 7:41pm
Project X sounds like a good name. Im very bored with FSX scenery mainly because of the problems you show in your first default screenshot{generic objects just scattered everywhere} The ProjectX screens look very good, put me down for that , will it be out in under another 4 months?

-------------
C2D 6600

9600GT 512MB

W7 64



Posted By: jaya
Date Posted: April-14-2009 at 11:26pm
Looks like I'll be spending some money!!!!  Smile


Posted By: c152flyboy
Date Posted: April-15-2009 at 2:46pm
going to have to ask the wife for a raise in my allowance.  Wink


Posted By: empeck
Date Posted: April-15-2009 at 5:26pm
This is one of the best FSX projects I've seen lately.

Are you considering Europe version too?

-------------
http://www.simhangar.com">


Posted By: akriesman
Date Posted: April-17-2009 at 12:25am

Originally posted by empeck empeck wrote:

This is one of the best FSX projects I've seen lately.

Are you considering Europe version too?

Absolutely.    Europe is going to take some additional work because of the different textures and the diagonal "block buildings" in the larger cities (I don't know what they are called officially).

This new product is going to require us to expand our development team in order to get the products out in a reasonable time.    Even though we have highly automated the process, there is still a significant amount of manual work that must be done in each area (depending on how many semi-custom areas are present).

At this time, I am pretty sure that we are going to have to break the products up into smaller regions than UTX, if we provide a significant number of semi-custom objects.  The size of the product regions will also depend on the amount of urban areas present.   Since Canada is fairly rural, I can see it being a unique product.  But, the USA and Europe will probably have to be broken up a bit.    I will have more details as time goes on.  

What I am doing right now is creating test areas with the development system and trying to see how long it takes to work a region.   When the development system has been tweaked to the max for developer efficiency (speed of development), then it will be time to decide how the products will be packaged and how large the development staff will need to be.  At some point I will probably ask for some user feedback on the packaging.

 



-------------
Allen



Posted By: empeck
Date Posted: April-17-2009 at 12:45am
I understand.

I hope you won't forget Poland ;)

-------------
http://www.simhangar.com">


Posted By: pwheeler
Date Posted: April-17-2009 at 6:50am
This looks very promising!

One question - will you be adding a small area of concrete around each building? It might help them to blend in with the underlying vegitative texture. In your appartment complex example the buildings look a little stuck on. With a small surround of concrete they would blend much better.

Are the buildings attached to the road network by following the vectors, kind of like your encasements from UTX?

This really would be a great advance in the current technology. For me, it addresses one of the major issues with the current landclass system.


-------------
http://www.fswaterconfigurator.com">


Posted By: dmountford
Date Posted: April-17-2009 at 6:53am
Hey Allen,

I'll email you over the weekend :-)

Dean.


Posted By: daveEM
Date Posted: April-17-2009 at 8:40am
Well it's been a year since the Canada update and I was just about to spend the weekend compiling a whole raft of parking lot requests for Edmonton... and maybe a road extension, but I realize the road is too much work so I wouldn't have bothered with it.

Now this Clap.

Best you send me some tools and I'll compile Alberta for you this weekend. Big%20smile

A couple more guys like me and the product can be done before you have a name for it.

Any way the product sounds great and something I'd sure like.  The generic buildings look just fine as I've fixed up my city not bad with oil tanks, hospitals, my one shopping center complete with a requested parking lot, etc.

Looking forward to it.



-------------

Rreal traffic shot, yes red roads in the morning.
Edmonton 06:47 hrs. 10/19/07
dave
Canada


Posted By: larryisenor
Date Posted: April-17-2009 at 11:13am
Allen:

Perhaps in the future this could be applied to rural areas as well. Currently the farming areas of Western Canada have correctly placed roads but the underlying tiles also have roads as well as farm buildings. By using a generic field pasture scenery and attaching the farm buildings to the road grid it would provide a better representation of the area.

Larry


Posted By: kiwikat
Date Posted: April-17-2009 at 2:39pm
This is just incredible!!!  The magnitude of difference this will make is staggering.  Sounds like quite the project though.

Allen your PM inbox is full.  If you don't mind could you PM me with another means of contacting you?  Thanks. Smile

Keep it up guys.  This will be the biggest thing to happen to MSFS yet! Clap


-------------


Posted By: akriesman
Date Posted: April-18-2009 at 11:50pm



"One question - will you be adding a small area of concrete around each building?"

Paul. Yes, more recent versions already have this kind of thing as an option.  But, everything is a balancing act with framerates, etc.    I am trying not to go overboard with ground polys.   There are a lot of developer options when it comes to the underlying textures.   This demo is very basic.

"Are the buildings attached to the road network by following the vectors, kind of like your encasements from UTX?"

Only the houses are tied to vector roads.   The other buildngs are a complete different story and based on different logic.   The feature in our software that will probably be used most often is something called "dynamic object creation".   With this option, you select a large area using a ground polygon.   Then, the software adds buildings of the appropriate types inside the ground polygon, aligned with the roads.   As you can imagine, this piece has take a LONG time to develop and work out the kinks.   BTW, when I mean dynamic, I mean that the objects are created automatically at compile time, not runtime in FSX.  But, they are dynamic in that they are not manually positioned based user placement or set coordinate positions (like POI's).




-------------
Allen



Posted By: akriesman
Date Posted: April-18-2009 at 11:52pm

Originally posted by kiwikat kiwikat wrote:

This is just incredible!!!  The magnitude of difference this will make is staggering.  Sounds like quite the project though.

Allen your PM inbox is full.  If you don't mind could you PM me with another means of contacting you?  Thanks. Smile

Keep it up guys.  This will be the biggest thing to happen to MSFS yet! Clap

I cleared out my mailbox a bit.  So, you can PM me if you want....

"Sounds like quite the project though."

Yes, the software development has been very challenging.   A few of the base ideas came from a Terramodels development system that I did for Fly! several years ago.    But, this system probably has 20x the functionality.   The goal is to speed up object placement as fast as humanly possible, cutting corners where applicable.   For example, instead of modeling the entire downtown area for Tyler above, an area is marked off using satellite imagery, a category is selected using my software (medium sized city) and then the software populates all areas along the roads in the marked off area with appropriate objects (from the selected category).    So, the downtown area is the exact size as real-life.  But, the buildings placed are semi-random.   I don't know if that explaination made sense :)

I have been buried in code for at least the last 6 months on this, which is why I have not had much of a forum presence lately.   

I am hoping that this product can breath new life into FSX.  But, it is being developed with other simulator platforms in mind also just in case :)

 



-------------
Allen



Posted By: ldmax
Date Posted: April-19-2009 at 1:30pm
Allen, this is nothing short of revolutionary!  Another great leap forward in the history of flight simulation.  I am in awe at your dedication and hard work. 

-------------


Posted By: c152flyboy
Date Posted: April-19-2009 at 11:43pm

Allen, i thought your comment about breathing new life into FSX was dead on.  if FSX is the last of the breed, then developers like yourself are the ones who can make it a whole new sim again.  i think the project looks amazing and cant wait for more info.

kyle



Posted By: kiwikat
Date Posted: April-20-2009 at 12:03am
Thanks for the small insight Allen. I'm just fascinated by it all.  I can't say I have ever been this excited about flight simulator before! Embarrassed LOL

I sent you a PM too Tongue


-------------


Posted By: Thaellar
Date Posted: April-20-2009 at 10:31am
WOW...absolute must have. I love perusing the countryside. Thanks for the updates.

-------------


Posted By: alainneedle1
Date Posted: April-21-2009 at 8:51am

OK! so if I understand this new project will overwrite landclass?

I am very interrested about this new project, I will probably be first in line to get it but I need to see a picture or two of New York city before and after (with framerate counter on)

If you can do that I'll will prepaid my order if you take prepaid order.

 



-------------
Window 7 64-bit

Gigabyte UD-9

980x @4.50 24/7

PSU 1300W

2000MHz 7-7-7-20

Noctua NH-D14

Cricial SATA III C300 128GB

OCZ Z-Drive PCI-e 256GB for FSX

Zotac GTX 480 AMP<


Posted By: akriesman
Date Posted: April-21-2009 at 6:07pm
Originally posted by alainneedle1 alainneedle1 wrote:

OK! so if I understand this new project will overwrite landclass?

I am very interrested about this new project, I will probably be first in line to get it but I need to see a picture or two of New York city before and after (with framerate counter on)

If you can do that I'll will prepaid my order if you take prepaid order.

Does the picture of the frame rate counter have to be greater than 0 LOL   Just kidding.

NYC and other big cities will be challenging. The key will probably be to allow the users to fine-tune the scenery configurations in various ways to achieve an acceptable balance.    

 



-------------
Allen



Posted By: alainneedle1
Date Posted: April-21-2009 at 6:48pm

Does the picture of the frame rate counter have to be greater than 0 LOL   Just kidding.

NYC and other big cities will be challenging. The key will probably be to allow the users to fine-tune the scenery configurations in various ways to achieve an acceptable balance.    

 

I understand, that is why I am asking for a pic or two of New York city with the frame counter on (does not have to be higher than 60FPS) LOL. Any problem doing that? Where are the pics, I have my plastic in hands........



-------------
Window 7 64-bit

Gigabyte UD-9

980x @4.50 24/7

PSU 1300W

2000MHz 7-7-7-20

Noctua NH-D14

Cricial SATA III C300 128GB

OCZ Z-Drive PCI-e 256GB for FSX

Zotac GTX 480 AMP<


Posted By: dmountford
Date Posted: April-21-2009 at 7:18pm
Originally posted by alainneedle1 alainneedle1 wrote:

Any problem doing that? Where are the pics, I have my plastic in hands........



Alain, I believe the shots are early preview shots of the technology, i.e. pre beta...


Posted By: alainneedle1
Date Posted: April-21-2009 at 10:59pm
OK! will wait...........

-------------
Window 7 64-bit

Gigabyte UD-9

980x @4.50 24/7

PSU 1300W

2000MHz 7-7-7-20

Noctua NH-D14

Cricial SATA III C300 128GB

OCZ Z-Drive PCI-e 256GB for FSX

Zotac GTX 480 AMP<


Posted By: mhoffman50
Date Posted: April-22-2009 at 9:46pm

 

WOW! This is going to revolutionize flight simming.  Just for grins I took a screen shot of Tyler in google earth and your preliminary work is extremely close to the real thing!

 



Posted By: alainneedle1
Date Posted: April-23-2009 at 8:58am

I hope you did something with the highway's overpass, that's one turn off for me......any comments?

 

 



-------------
Window 7 64-bit

Gigabyte UD-9

980x @4.50 24/7

PSU 1300W

2000MHz 7-7-7-20

Noctua NH-D14

Cricial SATA III C300 128GB

OCZ Z-Drive PCI-e 256GB for FSX

Zotac GTX 480 AMP<


Posted By: quantumleap
Date Posted: April-23-2009 at 10:33am
Originally posted by alainneedle1 alainneedle1 wrote:

I hope you did something with the highway's overpass

As noted in the original message in this thread, this new product is related to building objects.

As for the technical nature of limitations in FSX extrusion bridge technology and use to build overpasses this was covered in this http://www.simforums.com/forums/forum_posts.asp?TID=30157 - thread just this week.

Jeff



-------------

Check out my aviation photography and digital art at http://www.photisify.com" rel="nofollow - Photisify


Posted By: NickN
Date Posted: April-23-2009 at 12:01pm

 

 

As Jeff mentioned there are FSX traffic limitation issues with making 3D overpasses work... one of the reasons I personally do not use the Highway Interchanges apron in UTX (disable them) is that to me it makes it less visually apparent if the apron is not present and much better to view from the air. Some places have those apron areas in real life and others do not.



-------------


Posted By: c152flyboy
Date Posted: April-23-2009 at 2:26pm
Allen, going by the following comment you made in an earlier post-
 The key will probably be to allow the users to fine-tune the scenery configurations in various ways to achieve an acceptable balance."
would i be correct that Project X will work in the same way that UTX allowed you to turn things on or off, like rural roads for example?
when you mentioned a place like NY city, it would be good if you have multiple levels of display for people.


Posted By: alainneedle1
Date Posted: April-23-2009 at 6:11pm

Hi! guys, I do not have any problem what so ever with having the choice of disabling anything to make the flight smoother or for any other reason but you have to make as much as possible (bridges, overpass or else) even if it is hard on FPS because a lot of people have high end PC and can maxout almost everything.

Not to compare other products because I like UTX a lot but FTX as a lot to offer as far as buildings roads and bridges with a natural feel.

Again don't get me wrong I love UTX but I am always looking to fly where it will look as real as possible.



-------------
Window 7 64-bit

Gigabyte UD-9

980x @4.50 24/7

PSU 1300W

2000MHz 7-7-7-20

Noctua NH-D14

Cricial SATA III C300 128GB

OCZ Z-Drive PCI-e 256GB for FSX

Zotac GTX 480 AMP<


Posted By: dmountford
Date Posted: April-23-2009 at 6:37pm
Originally posted by alainneedle1 alainneedle1 wrote:

Hi! guys, I do not have any problem what so ever with having the choice of disabling anything to make the flight smoother or for any other reason but you have to make as much as possible (bridges, overpass or else) even if it is hard on FPS because a lot of people have high end PC and can maxout almost everything.



I think there needs to be a realistic expectation set as to what is currently available in terms of dataset availability...

I wouldn't go expecting Allen to provide datasets that even the worlds top mapping companies don't have.

It's like photoscenery users for flight simulator... Some expect every aerial image to be perfectly color balanced, shot at high noon so there's no shadows from the mountains or buildings on the scenery, and they expect it to be continent or even global in coverage...

Expectations like those are simply unrealistic... First the datasets need to exist in reality, and large companies pay millions of dollars to generate such databases. Sometimes MSFS users can expect a little too much and don't always realize that the expectations go above and beyond even what military forces have...

I'm sure Allen's solution will have as much detail as is practical and available at this point in history... I can say though that in coming years more advanced and better datasets can be developed 'in house', but from what I've seen so far Allens work is far beyond anything anyone has done for FSX before...


Posted By: akriesman
Date Posted: April-23-2009 at 10:51pm

Originally posted by c152flyboy c152flyboy wrote:


would i be correct that Project X will work in the same way that UTX allowed you to turn things on or off, like rural roads for example?
when you mentioned a place like NY city, it would be good if you have multiple levels of display for people.

Yes.  There will be lots of ways to fine-tune things to improve frame rates (if necessary).    I will always try and squeeze every little bit of performance out of things if possible.

At this point, you can already toggle different layers off and on.  And, the autogen and scenery complexity sliders are already being used efficiently.    Since the very large cities are hardest on frame rates, I plan on balancing out the number of buildings using a weighted scheme.    For example, setting the scenery complexity slider to "Normal" will affect the largest cities more than the smaller cities.   But, at the maximum setting all objects will be displayed.



-------------
Allen



Posted By: NickN
Date Posted: April-24-2009 at 12:10am

 

 

How DARE you give us access to features for tuning Allen!

Shame on you!  Big%20smile



-------------


Posted By: alainneedle1
Date Posted: April-24-2009 at 9:24am
Originally posted by dmountford dmountford wrote:

Originally posted by alainneedle1 alainneedle1 wrote:

Hi! guys, I do not have any problem what so ever with having the choice of disabling anything to make the flight smoother or for any other reason but you have to make as much as possible (bridges, overpass or else) even if it is hard on FPS because a lot of people have high end PC and can maxout almost everything.



I think there needs to be a realistic expectation set as to what is currently available in terms of dataset availability...

I wouldn't go expecting Allen to provide datasets that even the worlds top mapping companies don't have.

It's like photoscenery users for flight simulator... Some expect every aerial image to be perfectly color balanced, shot at high noon so there's no shadows from the mountains or buildings on the scenery, and they expect it to be continent or even global in coverage...

Expectations like those are simply unrealistic... First the datasets need to exist in reality, and large companies pay millions of dollars to generate such databases. Sometimes MSFS users can expect a little too much and don't always realize that the expectations go above and beyond even what military forces have...

I'm sure Allen's solution will have as much detail as is practical and available at this point in history... I can say though that in coming years more advanced and better datasets can be developed 'in house', but from what I've seen so far Allens work is far beyond anything anyone has done for FSX before...

I understand what you are saying here, and I don't pretend knowing how to built a scenery because I don't have a clue of how to do that but if the builder can make fake houses, building or stadium can't they make fake overpass, shopping mall and stuff like that? 



-------------
Window 7 64-bit

Gigabyte UD-9

980x @4.50 24/7

PSU 1300W

2000MHz 7-7-7-20

Noctua NH-D14

Cricial SATA III C300 128GB

OCZ Z-Drive PCI-e 256GB for FSX

Zotac GTX 480 AMP<


Posted By: dmountford
Date Posted: April-24-2009 at 10:01am
Well yes that's possible to do on an individual level, but imagine doing that by hand for every single bridge for the entire nation of the United States, as well as Canada, Alaska and Europe...

Let me break it down... Say it takes 10 minutes to code a bridge for rendering for FSX... Now say there's 10 million bridges in the United States alone... Not all of which are shown in continent wide datasets...

To code by hand would mean 100 million minutes to code manually... That equals 1,666,667 hours, or 69,444 days, or 190 years... Not including time for sleep and eating... Which means the system needs to be automated...

Under automation what needs to happen is image analysis and vector analysis needs to be done across the entire dataset for the USA, Canada, Europe and Alaska... This is something that Allen has programmed for but it can only be done feasibly programatically...

The other possibility is to form business deals with commercial grade data, which still as of this time in history doesn't have every bridge, house and building coded into a database... There's a few companies that have a lot of data, but the data is definitely far from 100% coverage...

The other alternative is to invest in high end image analysis software, and prices starting in the thousands and tens of thousands of dollars... Not a real practical solution...

So for now, I think we should take those factors into consideration and understand that Allen has the best and most practical solution to these problems through his own technological development, Allen's technology is by far the best solution for any current home based flight simulator platform and even likely exceeds the databases that Level-D simulators have...

Basically he's giving us the world on a platter and it feels like we're asking him for more, and I think we're better to support him and understand the technical limitations and constraints of current 21st century technology and one day we'll see even more accurate datasets, but honestly Allen's done a spectacular job with this so far, all I'm encouraging us to do is ensure we're not expecting too much of the impossible, even though Allen has already achieved the impossible ;)...


Posted By: c152flyboy
Date Posted: April-24-2009 at 11:26am
i for one think that Project X is already far beyond what is currently offered for FSX and i am looking forward to it.  i understand the issues that Allen has to deal with here and i think he is offering a great product.  if you want to see every bridge and house, go buy photo scenery.


Posted By: alainneedle1
Date Posted: April-24-2009 at 11:27am

I understand what you'r saying and I have never asked for 100% accuracy on anything but it will be nice to have some overpass in a couple of big city like Miami, New York, San Francisco, Hawaii (Honolulu) and others, forget about Fargo.....Big%20smile

As far as the work Allen is doing I have no complain on it, I own all UTX + GEX + Scenerytech +++++ and I am very happy with these products, I was just making a suggestion to make all simmer's flight a little more real.

Best of luck on this new project......Thumbs%20Up

 

 

 

 



-------------
Window 7 64-bit

Gigabyte UD-9

980x @4.50 24/7

PSU 1300W

2000MHz 7-7-7-20

Noctua NH-D14

Cricial SATA III C300 128GB

OCZ Z-Drive PCI-e 256GB for FSX

Zotac GTX 480 AMP<


Posted By: akriesman
Date Posted: April-25-2009 at 1:56am

Originally posted by dmountford dmountford wrote:

Well yes that's possible to do on an individual level, but imagine doing that by hand for every single bridge for the entire nation of the United States, as well as Canada, Alaska and Europe...

Let me break it down... Say it takes 10 minutes to code a bridge for rendering for FSX... Now say there's 10 million bridges in the United States alone... Not all of which are shown in continent wide datasets...

To code by hand would mean 100 million minutes to code manually... That equals 1,666,667 hours, or 69,444 days, or 190 years... Not including time for sleep and eating... Which means the system needs to be automated...

Under automation what needs to happen is image analysis and vector analysis needs to be done across the entire dataset for the USA, Canada, Europe and Alaska... This is something that Allen has programmed for but it can only be done feasibly programatically...

The other possibility is to form business deals with commercial grade data, which still as of this time in history doesn't have every bridge, house and building coded into a database... There's a few companies that have a lot of data, but the data is definitely far from 100% coverage...

The other alternative is to invest in high end image analysis software, and prices starting in the thousands and tens of thousands of dollars... Not a real practical solution...

So for now, I think we should take those factors into consideration and understand that Allen has the best and most practical solution to these problems through his own technological development, Allen's technology is by far the best solution for any current home based flight simulator platform and even likely exceeds the databases that Level-D simulators have...

Basically he's giving us the world on a platter and it feels like we're asking him for more, and I think we're better to support him and understand the technical limitations and constraints of current 21st century technology and one day we'll see even more accurate datasets, but honestly Allen's done a spectacular job with this so far, all I'm encouraging us to do is ensure we're not expecting too much of the impossible, even though Allen has already achieved the impossible ;)...

Excellent analysis Dean !    And, thanks for the kind words.   Ideally, I wish that every object could be highly detailed and hand positioned.   But, this strategy is not even close to being feasible as you mentioned, unless we release individual cities as unique products (this has already been done by others).

In order to release a product that covers a much larger area, most of the features need to be placed using automation based on existing data (POI data and road layers in particular).  For those items of importance that should be created and placed by hand, we have really sped up the process (150 - 180 an hour).    But, even if you could create 180 buildings that are accurately sized and positioned, you could not do this for a very large area at this rate.   So, we are currently concentrating the manual parts on large features of particular importance like: universities, schools, hospitals, malls, large strip centers, etc.   Other features will be placed based on the POI data locations and object types.

Overall, I think most people will be blown away by the results.   Our goal with this project is to provide a realistic and accurate (where possible) environment at low level flight, without turning FSX into a slide-show.   



-------------
Allen



Posted By: katyp
Date Posted: April-25-2009 at 1:13pm
Allen,

"Our goal with this project is to provide a realistic and accurate (where possible) environment at low level flight, without turning FSX into a slide-show. "

That's the hardest part to have people understand sometimes... Better results almost always end with more expectations

-------------
Best
Katy


Posted By: dmountford
Date Posted: April-25-2009 at 7:47pm
lol that reminds me of two complaints with FSX...

A) it's a slideshow
B) not enough visual detail

I would just laugh and scratch my head and say well what do you want, high fps or every blade of grass?

The great thing is that new technologies are coming out to display imagery at high fps with low cpu/ram/gpu load, and at the same time GPU/CPU/RAM are all increasing in capacity...

i.e. smaller footprints computationally coupled with increased computing power will in the next few years lead to 3d rendering nirvana...   


Posted By: NickN
Date Posted: April-26-2009 at 2:22pm

 

 

That grass is a texture and bumpmap.. hardly anything that would effect perf at all

I do not use those 'global' replacement grass bump textures because they screw up snow and other surfaces.

They also make GEX blurry

Some scenery products use them where they only affect the scenery product area which is fine.

The issue of perf is based on rendering # of objects in a scene.

That is why trees are killer on perf, lighting too, and buildings are no where near as hard on a system as they are few in number compared to trees

 however when you start adding in more and more buildings it adds up fast.

There is no such thing as a free lunch with FSX.

# of objects rendered will always produce a perf impact and there is no way around that

 

 which is why the UTX boys give us settings to trim with their products.

 

 

 

 



-------------


Posted By: alainneedle1
Date Posted: April-26-2009 at 5:46pm

Hi! Nick, I agree with you on what you are saying but me I have no problem so far with screwed up snow and other surface, as far as GEX I don't have blurry....do you see them?

Mayby you will see them with a GTS 8800 512MB + CPU 2 core at 2.4GHz don't know.

I know nothing about making stuff for FSX but I can tell you right now than I will always buy what ever look good to me and so far I have all UTX + GEX + Sceneryteck and waiting for GEX Europe and I am very happy.

The grass bump texture is very nice and look very real compare to other products so if somebody come up with something better (visual effect) please let me know I have my plastic in hand.

Also waiting for project X.......

 

 



-------------
Window 7 64-bit

Gigabyte UD-9

980x @4.50 24/7

PSU 1300W

2000MHz 7-7-7-20

Noctua NH-D14

Cricial SATA III C300 128GB

OCZ Z-Drive PCI-e 256GB for FSX

Zotac GTX 480 AMP<


Posted By: NickN
Date Posted: April-26-2009 at 8:57pm

 

 

Then the addon you are showing in those images is using the texture for grass and it shows no where else

 

If you are seeing detailed grass eveywhere in the sim then snow, roads and all other surfaces have that same bump to them

and you simply cant tell the difference

 

2x800GTX is useless to FSX.. you are only seeing one card in use unless running above 2000x+ resolution and 16x+ AA 

 



-------------


Posted By: NickN
Date Posted: April-27-2009 at 2:10pm

 

 

I think this is getting off topic..

and none of the imgaes posted will show the grainy snow and roads from that height assuming its a global addon

(and no I dont want to see it as I know what it looks like and dont like it) LOL

 

My point was texture addons do not drop perf regardless of the detail they may show

Perf is related to # of object in the scene

and although higher res textures (above 1024) will also effect perf it's object rendering that drives it.

 

 

 



-------------


Posted By: alainneedle1
Date Posted: April-27-2009 at 2:25pm

The only products I used for the pics above is UTXCanada + Gex + Scenerytech and REX.

Pics are from Vancouver Canada.

Thank you for the good work.

 



-------------
Window 7 64-bit

Gigabyte UD-9

980x @4.50 24/7

PSU 1300W

2000MHz 7-7-7-20

Noctua NH-D14

Cricial SATA III C300 128GB

OCZ Z-Drive PCI-e 256GB for FSX

Zotac GTX 480 AMP<


Posted By: akriesman
Date Posted: April-29-2009 at 12:16am

I was working on the night textures today.  So, I thought that I would share a couple of shots :)

Remember, these textures are a work in progress.  But, they look pretty good right now.

Here is a good night shot of a suburban area.   Notice in the distance we have the default FSX suburban textures (the orange lighting).   You can see the big difference.   The UTX street lights were used here also for effect.  The suburban roads are accurate and houses line the streets.




-------------
Allen



Posted By: akriesman
Date Posted: April-29-2009 at 12:22am

Here is a dandy of a shot of downtown Tyler and the surrounding suburbs.  This shot has full autogen, the UTX lights and even light bloom.  So, you can imagine that it hits my older Dual Core machine pretty hard.  However, you can turn off only the houses at night (which are not very noticable anyway) and frame rates take off (no pun intended).



-------------
Allen



Posted By: Axelb9
Date Posted: April-29-2009 at 12:45am

Hello Allen,

Looks really good. Have you got any plans at this stage how large an area you would release in one package? I am thinking specifically of HawaiiSmile. Would a released area be the size of a state or like 4-5 enhanced areas for the US and Europe or something totally different?



Posted By: pwheeler
Date Posted: April-29-2009 at 3:53am
Great shots there Allen. Having everything line up with the main roads with the UTX lights looks awesome!

I'm not sure that the main road textures blend that well. Will they be upgraded at all? Are the main roads from UTX or 'X'? i.e. will people need both products for the look you are showing or will the new product include lights etc?

Will areas such as main shopping high streets get their own textures with a bit more colour?

-------------
http://www.fswaterconfigurator.com">


Posted By: c152flyboy
Date Posted: April-29-2009 at 8:51am
Allen,
Very impressed with that first photograph showing the compassion between the FSX default and Project X suburban textures.  lt is amazing how much more realistic it is looking.


Posted By: akriesman
Date Posted: April-29-2009 at 2:12pm

"Looks really good. Have you got any plans at this stage how large an area you would release in one package? I am thinking specifically of HawaiiSmile. Would a released area be the size of a state or like 4-5 enhanced areas for the US and Europe or something totally different?"

The size of the packages will completely depend on the amount of manual labor needed for each region.   We are going to try and automate as much as possible.   But, some things, like the semi-custom buildings will be hand drawn using our toolset.   While the latter is very fast compared to conventional methods, the labor time can really add up over a very large area.

Right now the goal is to figure out roughly (+/-20%) how long each part of the USA would take for the manual parts.   It is a balancing act to try and figure out which parts to automate and which parts to add a manual touch to.

For example, right now hospitals, schools, universities, apartments, stadiums, city centers and large strip centers have some element of manual work required.   Objects outside of these types are handled by automation that positions buildings based on POI (point of interest) data.   The commercial POI data positioning is done thru a process called address geocoding.   So, each object position will not be exact (but usually very close).    As a result, right now we are hand placing the most important objects (identifiable objects like schools, etc) and automating other types of object placement (like restauraunts, gas stations, etc).

We have several options available to use with this product line.  

Option 1:We could release a nearly-fully automated product, covering a larger area, which cuts out the semi-custom buildings (those placed by hand).   In this case, all buildings will be placed using POI data.   Then, we could release smaller addon packages for more detailed city areas.

Option 2:We could combine everything like we are doing now.   With this option, we would try and break the USA into something like 4 or 5 regions (hopefully).    The regions would be broken up by population centers, because the larger cities will take the longest to work.   So, the Northeast part of the USA would cover a smaller area than the Western USA.

Hope this helps answer some questions.   The protoype and development system are very nearly complete.    The question now is how we are going to package things that are best for both us and our customers.

Feedback will certainly beneficial to us on this matter.



-------------
Allen



Posted By: akriesman
Date Posted: April-29-2009 at 2:21pm

"Great shots there Allen. Having everything line up with the main roads with the UTX lights looks awesome!

I'm not sure that the main road textures blend that well. Will they be upgraded at all? Are the main roads from UTX or 'X'? i.e. will people need both products for the look you are showing or will the new product include lights etc?

Will areas such as main shopping high streets get their own textures with a bit more colour?"

The main roads and lighting will be part of UTX.   We might be create some alternate road textures for UTX users to blend in better if necessary.

In theory, this product should work pretty well with the default FSX roads also.  So, UTX is not required.   Of course UTX will give a much better overall look.

I am not completely clear on the comment regarding the high street textures.  maybe you could elaborate ?   Are you talking about major roads with a lot of commercial activity ?   If so, this would be kind of hard to identify because all major roads have no further classification available (i.e roads containing lots of businesses).    We could probably do this programatically with the POI data and road data together.   But, the street layers will be completely provided by UTX and not this product.   So, this would probably be more of a future UTX upgrade.



-------------
Allen



Posted By: c152flyboy
Date Posted: April-29-2009 at 7:59pm
Allen, let me see if i understand the two options you mentioned above.  option 1 is to cover more area with less accurate POI and less semi custom building and option 2 is better POI location and more custom buildings but for smaller areas?  if that is  the case i vote for option 2 myself.  the way i see it, if you are going to make a product that makes cities and towns look more realistic, though not exact, will be more closly done with option 2.  so say a north east, south east, north west, south west, and central section?


Posted By: akriesman
Date Posted: April-30-2009 at 5:06pm

Originally posted by c152flyboy c152flyboy wrote:

Allen, let me see if i understand the two options you mentioned above.  option 1 is to cover more area with less accurate POI and less semi custom building and option 2 is better POI location and more custom buildings but for smaller areas?  if that is  the case i vote for option 2 myself.  the way i see it, if you are going to make a product that makes cities and towns look more realistic, though not exact, will be more closly done with option 2.  so say a north east, south east, north west, south west, and central section?

Thanks for the feedback.   Yes, I am thinking the sections might be similar to what you mentioned.      FWIW, the POI data is automated, except for some exceptions that we kick out for manual work.   The big question here is how many types of POI objects do we kick out for exception work and which types ?  It is those exceptions that will add lots of manual labor and cause smaller regional products.

Here is my opinion on what we are trying to do.  We want to give you guys an environment so that when you approach a city, even one that you are familiar with, you can say "yeah, I recognize that area".   And, you can fly VFR based on the roads locations, houses and areas of buildings that you recognize.   Now, as you get really close to an area (almost on top of it), you might not recognize things as much.

One of the most important technological features in the custom software that we developed is something we call Dynamic Object Positioning.   With this feature, the developer only has to mark off a portion of the satellite imagery that contains buildings of a similar type.    For example, the entire downtown area for Tyler was done this way.     I marked off the downtown areas, told the software that the downtown area was type "Medium USA City", then the software starting creating random objects to fill the downtown area as appropriate.   A medium USA city might have 20% hi-rise, 25% offices, etc.   Each time you rerun the software, the downtown area might change somewhat.

Apartment complexes are created this way, plus storage buildings, warehouse districts, etc.   I can create the entire Tyler downtown area in a matter of minutes.  Doing the downtown area manually would take hours using our software or days/weeks using off-the-shelf products.

Downtown areas are very large.  So, this method works perfectly from an efficiency standpoint.   Unless you know a downtown area well, you won't know that the buildings are different, because they look realistic and are placed in the correct regional location.

Personally, I think this product became viable when the Dynamic Object Positioning code was completed.    If not for this piece of technology, the USA would be far too big of an area to tackle.

Ok, with that description out of the way, I can talk about semi-custom buildings and other buildings that are positioned by hand.    The dynamic logic previously described does not work for all areas.   Shopping malls, schools, universities and other areas probably need more work, because these objects are VFR markers for a lot of pilots (IMO).   We could just rely on POI data to place these objects.   But, my feeling is that these types of objects are important enough to warrant more attention to detail, even though they are not detailed objects per-se.

A very good product could be created with just the Dynamic Object Positioning, plus the POI data (converted to appropriate objects), plus the houses.   The dynamic object positioning takes a little bit of manual work.   But, it is much less than having to create the semi-custom objects (or placing library objects by hand).    At this point, I think that adding semi custom objects for a select group of building types will add 4x to 5x the manual labor.   So, they really need to be worth the effort.

The housing part is almost fully automated now that the software is complete.

Hope this helps.   By educating you guys on our processes, I can probably get the best feedback.     The world is a big place and I think this will be a popular product.  So, would rather not get bogged down providing too much detail, which will cause products to be released very slowly.   I am trying to find the perfect balance between value and feature sets.



-------------
Allen



Posted By: Axelb9
Date Posted: May-01-2009 at 2:36am

I know that this most probably out of question but I guess the world would be finished in a matter of a couple of months if there was some kind of a development kit released... Everybody would be most enthousiastic to create, customize and share their own area that they know as the back of their hands. So maybe the most efficient and fastest way would be to release large areas by you with the least amount of manual work involved and the most automation and then let the community to play around with the semicustom buildings, etc.

I know I would enjoy playing with this tremendously. And then of course the work of everybody could be shared.

Maybe even two versions: a cheaper basic one containing just the automated work and a more expensive one containing the possibility to customize and place the semi custom buildings...

I know I would opt for the more expensive one without hesitation.

The world would look fantastic in a matter of weeks including other continents.

Alex



Posted By: c152flyboy
Date Posted: May-01-2009 at 11:25am
which type of work should be custom work you ask?  well from your list i agree universities and colleges should be because they are unique looking and i can see malls as well.  warehouse and industrial could also use more custom buildings but not so much as colleges and universities.
   
as i was looking at the first picture you showed of Tyler with Product X, i thought sure maybe Tyler downtown does not have those exact buildings in the exact location but if it is more representative of reality, then all the better.  my hometown of New Bedford Massachusetts has a population of 100,000 but a very small condensed Downtown located on the far south end of the city along the waterfront.  (BTW, new bedford is the most productive seaport in $$$ in the world)  it also has 23 schools and a large industrial complex in the far north end.  KEWB is located between all of this.  the next town over has a large mall and a large university that both sit under the ILS RWY 5 approach and are used for visual approaches.  I would like these things to be represented and sure it would be great if they looked somewhat like the real thing and i will be willing to wait a bit longer, pay a bit more, or purchase a smaller section to get it.

like you said it wont look exact as your flying over it but from a distance it should at least give you the impression that what your approaching are the real items in the proper locations.


Posted By: akriesman
Date Posted: May-01-2009 at 4:32pm

Originally posted by c152flyboy c152flyboy wrote:

which type of work should be custom work you ask?  well from your list i agree universities and colleges should be because they are unique looking and i can see malls as well.  warehouse and industrial could also use more custom buildings but not so much as colleges and universities.
   
as i was looking at the first picture you showed of Tyler with Product X, i thought sure maybe Tyler downtown does not have those exact buildings in the exact location but if it is more representative of reality, then all the better.  my hometown of New Bedford Massachusetts has a population of 100,000 but a very small condensed Downtown located on the far south end of the city along the waterfront.  (BTW, new bedford is the most productive seaport in $$$ in the world)  it also has 23 schools and a large industrial complex in the far north end.  KEWB is located between all of this.  the next town over has a large mall and a large university that both sit under the ILS RWY 5 approach and are used for visual approaches.  I would like these things to be represented and sure it would be great if they looked somewhat like the real thing and i will be willing to wait a bit longer, pay a bit more, or purchase a smaller section to get it.

like you said it wont look exact as your flying over it but from a distance it should at least give you the impression that what your approaching are the real items in the proper locations.

This is good.  It sounds like we are thinking along the same lines.



-------------
Allen



Posted By: akriesman
Date Posted: May-01-2009 at 4:56pm
Originally posted by Axelb9 Axelb9 wrote:

I know that this most probably out of question but I guess the world would be finished in a matter of a couple of months if there was some kind of a development kit released... Everybody would be most enthousiastic to create, customize and share their own area that they know as the back of their hands. So maybe the most efficient and fastest way would be to release large areas by you with the least amount of manual work involved and the most automation and then let the community to play around with the semicustom buildings, etc.

I know I would enjoy playing with this tremendously. And then of course the work of everybody could be shared.

Maybe even two versions: a cheaper basic one containing just the automated work and a more expensive one containing the possibility to customize and place the semi custom buildings...

I know I would opt for the more expensive one without hesitation.

The world would look fantastic in a matter of weeks including other continents.

Alex

Alex,

Some of what you say is interesting.   So, I will give it some thought.   Perhaps, the semi-custom building portion could be released as a separate tool to our customers.    We could provide the downtown areas and buildings placed using the commercial POI data as a product, along with the tool to create the semi-custom objects.   While this sounds great, there would be obstacles though.

The first problem that would have to be overcome is the fact that during development, all 3 types of features are kind of related (semi-custom, POI buildings and houses).   The custom buildings are always done first, because they exclude the POI data buildings covering the same area.   Then, the custom and POI data buildings exclude the houses that might otherwise encroach on their territory.   So, there is an ordered development pattern that keeps things very neat.   Our software has a more efficient exclusion logic than what can be done by FSX at runtime (with exclusion files and layering, which might produce sloppy results IMO).

My other fear would be support for the tools.   I have written and distributed developer tools in the past.   And, the support can be very time consuming.   However, our users are very good when it comes to jumping in and helping others.   So, this might not end up being too big of a problem.

Another problem would be ownership rights.   Can we include user-created objects (using our tools) into future patches and versions of the product ?    What happens if we port the product to another platform (very possible future) ?

The last problem that I can think of would be the danger of the developer tools falling into the hands of those outside the UTX community.   Or, the tools being cracked by other developers.   This is only a slight concern.

Off the top of my head, those would be my concerns with distributing some of the development tools.   The biggest problem would be encorporating the user enhancements into the product so that they produce a clean result, without a lot of confusing clutter and excessive layering.

I think that a lot of people like UTX and products like it, because they don't have to contstantly download lots of small freeware packages to improve their environment.    UTX gives them a lot of functionality in one package.    I would like to keep this same mentality for future products.

Your idea does have merit, despite the potential hurdles.   I wonder how many users would actually share their work.    Or, maybe they don't share their work at all.  But, having the tool makes the product much more valuable to them.

One more potential problem....We cannot release any raw commercial data to our customers.  So, any distributed resource data would have to come from freely available sources.    This may not be a problem though from what I am thinking right now.   

Good discussion !



-------------
Allen



Posted By: dmountford
Date Posted: May-01-2009 at 7:10pm
I like the idea of end users being able to develop areas, one thing I've thought of before is having an online site where users could upload data and then on the server it generates the datasets for everyone to access...

I have long been thinking about something called Project GeoCode... I have some tools that can help with people logging POI's in Google Earth and then it runs a process and generates BGL's based off that...

The thing is that worldwide most locations don't change much in 10-20 years, some houses and buildings exist for a good 30-50 years...

Similar with vegetation, there's not usually a lot of change over the years. I've always thought a good central database for scenery design worldwide would be great... Kinda like openstreetmap but for scenery development...


Posted By: c152flyboy
Date Posted: May-01-2009 at 11:33pm

Allen,

looking at Alex's respnse about having user end people doing building and POI additions is like using ADE for making changes in FSX.  there may not be alot of peopl ewho do it ( i am one of them ) but they get shared over avsim or other sites all the time.  i could see the same thing happening for project X.  but from your side of the fence, i could see it would be a lot easier to develop and release project X with out the additions like that.  you could always release a basic and pro version with the pro side having the extra tools for peopl eto do changes thenselves.



Posted By: kenthansen
Date Posted: May-02-2009 at 1:44am
OK
I've been waiting for some feed back before replying.

What I would like to see is a fully automated product with the POI and DOP (Dynamic Object positioning) as a base product.

This will enhance UTX to a new level and give us the new vector autogen technology.

Then you can focus on detailed regional updates, with custom buildings.
I'm sure like myself, I will buy the region I live in...
And like most I want the most detail I can get, so I will update all the new regions as they come out.

For releasing tools, this is something to put a lot of thought into.
Even though I would like to be able to customize my home town.
I can see this leading to a lot of tech support.
Maybe something to add in the future.
There are already a lot of tools to out there to do this.






-------------
Asus P9X79 Mobo, Intel i7 3820 Quad CPU @ 3.8 GHz, 16GB DDR3 Ram @ 2133 MHz, GTX 760 OC 2GB GDDR5, 850w PS, W7 64 Optimized


Posted By: spesimen
Date Posted: May-02-2009 at 8:31pm
looks awesome!

as soon as i saw it i couldn't help but think "i'd love to get my hands on that tool to do my hometown." i agree that a lot of people who made such areas would freely contribute them, the sheer amount of freeware afcads and repaints on avsim is evidence of this.

although kent suggests there are other tools out there for this, i'm not sure i've seen any that would make it easy to annotate large buildings from a satellite image and plop them down short of doing a full scenery package, which sounds extremely time consuming. i like the idea of doing a large area quickly using existing vector scenery as a template.

i can understand that creating a toolset and releasing that would invite a lot of support nightmares as evidenced by the shenanigans some people have even with getting even basic stuff like utx setup correctly.

anyway it's an exciting idea, keep up the good work!
cheers,
-andy crosby


Posted By: alainneedle1
Date Posted: June-12-2009 at 2:01pm

Will this add tall building? What I'm getting at is this, if I fly with UTX as it is right now over FT-Lauderdale West Palm beach or Miami there is no tall building as it should be on the beach or the city.

Now everybody will say "If you add tall building you'll get a hit on your FPS", I already know that but for peoples who have a very good and fast PC it will be nice to have something (special slider?) to add building or not with this new product "x".

I live in FT-Lauderdale and I like when I see something more real as the pic. you posted at the begining of this tread so with the building been where they should be it will be fantastic. I hope you understand what I'm asking.

Thank you for all the good work.

PS: Anything new? More pics.?

 



-------------
Window 7 64-bit

Gigabyte UD-9

980x @4.50 24/7

PSU 1300W

2000MHz 7-7-7-20

Noctua NH-D14

Cricial SATA III C300 128GB

OCZ Z-Drive PCI-e 256GB for FSX

Zotac GTX 480 AMP<


Posted By: akriesman
Date Posted: June-12-2009 at 11:54pm
Originally posted by alainneedle1 alainneedle1 wrote:

Will this add tall building? What I'm getting at is this, if I fly with UTX as it is right now over FT-Lauderdale West Palm beach or Miami there is no tall building as it should be on the beach or the city.

Now everybody will say "If you add tall building you'll get a hit on your FPS", I already know that but for peoples who have a very good and fast PC it will be nice to have something (special slider?) to add building or not with this new product "x".

I live in FT-Lauderdale and I like when I see something more real as the pic. you posted at the begining of this tread so with the building been where they should be it will be fantastic. I hope you understand what I'm asking.

Thank you for all the good work.

PS: Anything new? More pics.?

I do understand your question.  But, it is a little difficult to answer without getting technical.   I don't know how well this will go over.  But, I will try :)

As I mentioned earlier, there are 3 parts to "project X".  

  1. Housing subdivisions.
  2. POI building translations
  3. Semi-Custom buildings

I think most people understand the housing subdivision part of it.  So, I will skip that and focus on the other two.

The other 2 parts of "project X" will provide your larger buildings.  

You heard me mention before the POI data (point of interest data).   This is the same data that you get with your GPS units.   POI data is used to direct you to restaurants, shops, etc using your GPS.   Each individual point in the data set has a location, name, genre (restaurant, shop, etc).   The locations are not exact, but they are generally close enough to their real life locations.  The reason they are not exact locations, comes from the fact that the lat/long positions are obtained from geocoding algorithms.  Geocoding is somewhat of a GIS industry standard that attempts to place building locations using address ranges to estimate lat/long locations.   The image below shows a small sample area of Tyler with the POI locations displayed on top of the roads.    There are a TON of POI's available (over 13 million for the entire USA).

Geocoding will place the POI's near their appropriate place on a street segment.  In a nutshell, our software takes the POI and moves it off the street (on the correct side).   It then groups POI's that are at the same address, so that a larger building can be placed for a group of POI's occupying the same address.  The "Project X" software always rotates the objects towards the street and makes sure objects don't overlap.   Lastly, it chooses an appropriate generic building from a library of hundreds of objects that have been created, plus some useful objects in the FSX library.   For example, a church gets a church building.   A BBQ restaurant gets a generic BBQ restaurant building object.   The results are pretty stunning, considering that the POI is not really designed to be used in this manner.  And, POI data tells nothing about the building structure or footprint shape.  But, each POI data point is very detailed as to what type of activity takes place at that location.  So, that is a big help.

The better our algorithms for automating the placement of buildings using the POI data, the less hand-drawing has to be done (the semi-custom building part).   And, more automation means larger product regions that can be sold as a single product in a shorter period of time.

Some buildings are more difficult than others to get correct using POI data.   Very tall buildings are one example.   For certain types of buildings, like those with more than x number of businesses within the same building, we kick the POI data out as exceptions.   And, we then work these by hand as semi-custom buildings (the 3rd part).   These semi-custom buildings are exactly the same size, shape and position as the real world buildings they represent.   The textures are generic and the building heights may not be exact, depending on the availability of quality satellite and ground images.

Our dilemma right now has to do with packaging the products and splitting them into regions.   We have some ideas.  But, nothing is definite.     Our goal is to automate as much as possible, with few exceptions.   The more exceptions that have to be hand-drawn, the longer it takes to develop and the smaller the product regions will be.    Even though we can hand draw/place about 150 objects an hour from satellite images, this can take a long time (up to many months) if you have too many objects kicked out as exceptions.

I hope this helps clarify some things.   We have a lot of intelligent users on this board that can probably understand exactly what we are trying to do from this discussion.     The more educated our users are, the better we can deliver what you guys want.   And, the better chance we have of meeting your expectations.

Cheers !



-------------
Allen



Posted By: alainneedle1
Date Posted: June-13-2009 at 9:19am

Thank you for the explaination, I now understand better.

 



-------------
Window 7 64-bit

Gigabyte UD-9

980x @4.50 24/7

PSU 1300W

2000MHz 7-7-7-20

Noctua NH-D14

Cricial SATA III C300 128GB

OCZ Z-Drive PCI-e 256GB for FSX

Zotac GTX 480 AMP<


Posted By: John K
Date Posted: June-14-2009 at 3:47pm
Originally posted by akriesman akriesman wrote:

Originally posted by Axelb9 Axelb9 wrote:

Finally also a tool that could solve the issue of autogenless photoscenery. BTW how are the green areas addressed inside a city such as parks, greens and trees alongside the roads?

Thanks Alex.   I am glad that you understand what it is we are trying to accomplish.

I actually started looking into this type of thing several years ago.   But, it has taken that long to work out a system that looks good, can be accomplished with reasonable amount of manual labor, and won't just kill frame rates.

City parks will mostly be provided by UTX, as they are now.   As we work on individual areas, we will add more detailed ground polygons as necessary (like we do with the parking lots here and other underlying terrain areas).    More or less trees could be added by changing a vector autogen entry associated with the houses.

Yes, this should work very well with photoscenery also.  Dean Mountford (FS Dreamscapes) and I are in touch often.   So, I can see a lot of our stuff working well with photoscenery.   Exept for the semi-custom drawn buildings, most other buildings won't overlay the photoscenery perfectly in most cases, because the precision of what we are providing won't be exact (but it is a big improvement over urban landclass).

This system is highly configurable from the users end.   So, you can do a lot of tuning for frame rates if necessary.

It will be great if this will help clean up some of the autogen problems when using photoscenery (I use some MegaSceneryX). Hopefully it will provide a better 3d look. And being that most of the UTX roads should already be alined with photoscenery roads, then so much the better.

John K



Posted By: alainneedle1
Date Posted: June-14-2009 at 5:13pm

Is it possible to get a picture of Miami Beach or FT-Lauderdale, I just would like to see it myself with all the good stuff on.

If this is like I think it will be I will buy it as soon as this is for sale, I already have my plastic in hand.

What is holding me back flying over these area is the lack of tall or medium building coverage around the cost and city.

Maybe I'm asking to much right now but one pic or two would be nice....

Thank you very much and keep the good work. 



-------------
Window 7 64-bit

Gigabyte UD-9

980x @4.50 24/7

PSU 1300W

2000MHz 7-7-7-20

Noctua NH-D14

Cricial SATA III C300 128GB

OCZ Z-Drive PCI-e 256GB for FSX

Zotac GTX 480 AMP<


Posted By: akriesman
Date Posted: June-15-2009 at 1:48am
Originally posted by alainneedle1 alainneedle1 wrote:

Is it possible to get a picture of Miami Beach or FT-Lauderdale, I just would like to see it myself with all the good stuff on.

If this is like I think it will be I will buy it as soon as this is for sale, I already have my plastic in hand.

What is holding me back flying over these area is the lack of tall or medium building coverage around the cost and city.

Maybe I'm asking to much right now but one pic or two would be nice....

Thank you very much and keep the good work. 

We have only worked on a couple of small demo areas using the "Project X" technology.   So, I don't have a way of giving any custom screenshots yet.

Alaska will be the first product that makes use of this new technology.  After that, we will begin working on the rest of the USA.

It is going to be a while before this product is available for the entire USA.   At this point, I don't even want to take any guesses as to any release dates.



-------------
Allen



Posted By: akriesman
Date Posted: June-15-2009 at 2:06am
Originally posted by John K John K wrote:

It will be great if this will help clean up some of the autogen problems when using photoscenery (I use some MegaSceneryX). Hopefully it will provide a better 3d look. And being that most of the UTX roads should already be alined with photoscenery roads, then so much the better.

John K

John,

Yes, the photoscenery support will be interesting with this product.     The houses and larger buildings should follow the photoscenery roads good enough.   However, they will not align with the building imprints in the photoscenery, except for cases where semi-custom buildings are drawn.  I don't think having the objects align perfectly with the building imprints in the photoscenery will be critical, as long as the appropriate objects are in the general vicinity and follow the roads.

The footprint data that Dean mentioned earlier is something that I have a lot of interest in also.   Footprint data actually has the building bases drawn already.   The "Project X" software that we have written can take any footprint and extract a 3D building model from it.    When we create semi-custom buildings, we basically draw our own footprints using satellite imagery and then run it thru our software.    Having the footprint data already available would eliminate 99% of the manual work needed for the semi-custom objects.

Building footprint data is probably the key to the future when it comes to developing accurate flight sim cities.   However, I don't believe that footprint data is available for many areas just yet, outside a few major USA cities.   It could take a while to get full coverage in the USA and Europe.   Also, having a custom object for each building footprint would probably kill perfomance right now in FSX.   We get around this performance issue with "Project X" by using libraries of predefined objects and vector autogen for a majority of the objects.

Hope this explanation helps.  



-------------
Allen



Posted By: John K
Date Posted: June-15-2009 at 4:22pm
Originally posted by akriesman akriesman wrote:

Originally posted by John K John K wrote:

It will be great if this will help clean up some of the autogen problems when using photoscenery (I use some MegaSceneryX). Hopefully it will provide a better 3d look. And being that most of the UTX roads should already be alined with photoscenery roads, then so much the better.

John K

John,

Yes, the photoscenery support will be interesting with this product.     The houses and larger buildings should follow the photoscenery roads good enough.   However, they will not align with the building imprints in the photoscenery, except for cases where semi-custom buildings are drawn.  I don't think having the objects align perfectly with the building imprints in the photoscenery will be critical, as long as the appropriate objects are in the general vicinity and follow the roads.

The footprint data that Dean mentioned earlier is something that I have a lot of interest in also.   Footprint data actually has the building bases drawn already.   The "Project X" software that we have written can take any footprint and extract a 3D building model from it.    When we create semi-custom buildings, we basically draw our own footprints using satellite imagery and then run it thru our software.    Having the footprint data already available would eliminate 99% of the manual work needed for the semi-custom objects.

Building footprint data is probably the key to the future when it comes to developing accurate flight sim cities.   However, I don't believe that footprint data is available for many areas just yet, outside a few major USA cities.   It could take a while to get full coverage in the USA and Europe.   Also, having a custom object for each building footprint would probably kill perfomance right now in FSX.   We get around this performance issue with "Project X" by using libraries of predefined objects and vector autogen for a majority of the objects.

Hope this explanation helps.  

Yes I understand that you won't be able to place autogen buildings in their proper 'footprint' like GEX does with the texture tiles. Besides making the UTX products look even better, I also want it to make photoscenery look better, which it looks like your trying to do. Keep up the good work, speaking of which, please get back to it!!! Wink

John K



Posted By: boleyd
Date Posted: June-17-2009 at 8:33am
The general landclass, as I understand it, is 1km x 1km. This can distort the reality of some elements such as smaller towns or cities. Also, can extend a city across a river where there is nothing on the other side. Cracking the 1km x 1km landclass barrier would be a fine accomplishment. Probably hard coded but who knows.....Star



-------------
Dick   near airport 5G8


Posted By: akriesman
Date Posted: June-17-2009 at 11:02pm

Originally posted by boleyd boleyd wrote:

The general landclass, as I understand it, is 1km x 1km. This can distort the reality of some elements such as smaller towns or cities. Also, can extend a city across a river where there is nothing on the other side. Cracking the 1km x 1km landclass barrier would be a fine accomplishment. Probably hard coded but who knows.....Star

Jeff and I (and I think Holger) have discussed the 1km x 1km landclass limitation with the Aces team at a previous developers conference.

I had hoped that the landclass precision could have at least been improved to double what it is now (500m) in FS11 (before FS11 was cancelled).   But, what I heard was this was going to be very difficult with the way the current FS engine was designed.

Project 'X' completely eliminates the need for landclass, except for the underlying vegetation that shows thru (grass, forests, etc).    So, the only limitation in accuracy with this system is in the data that is used.

 

 



-------------
Allen



Posted By: c152flyboy
Date Posted: July-10-2009 at 4:18pm
"Project 'X' completely eliminates the need for landclass, except for the underlying vegetation that shows thru (grass, forests, etc)"

Allen, if that is the case, if are already using UTX and it's landclass, would project X eliminate that landclass or at least some of it.  I know you mentioned that UTX and project X would play nice but it seems like project X would remove the need for a portion of what UTX does/


Posted By: akriesman
Date Posted: July-12-2009 at 12:30am

Project 'X' would eliminate the NEED for the urban landclass portion of UTX.    But, all other parts of UTX (roads, water, ground polys) will enhance project 'X'.

For Alaska, where we will combine both UTX and Project 'X' features, we will still provide urban landclass for those that want it.   But, it will be turned off by default.

Alaska is the only product that will combine features from 2 products (UTX and Project X).   This is because Alaska has a limited number of urban areas.   In Alaska, we will even be hand working the very small remote airport villages and outposts.

We really need to decide on a more permanent name for Project X.   'Ultimate Cityscapes' is the frontrunner right now.

Hey, we have to put the word 'Ultimate' in the title somewhere Wink



-------------
Allen



Posted By: kiwikat
Date Posted: July-13-2009 at 9:29am
Don't forget the "X" either...

Because we REALLY need another 3 letter acronym ending with X.

I've got a quick question.  Will there be any trees in neighborhoods or just in areas determined by the vegetation landclasses?  I see there are SOME trees in the previews but how are they implemented and how will they work in a larger city where only residential areas have trees?

I can't wait to see some more screenshots! Star


-------------


Posted By: alainneedle1
Date Posted: July-13-2009 at 11:00am

Originally posted by kiwikat kiwikat wrote:

Don't forget the "X" either...

Because we REALLY need another 3 letter acronym ending with X.

I've got a quick question.  Will there be any trees in neighborhoods or just in areas determined by the vegetation landclasses?  I see there are SOME trees in the previews but how are they implemented and how will they work in a larger city where only residential areas have trees?

I can't wait to see some more screenshots! Star

Yeah! more eyes candy..........



-------------
Window 7 64-bit

Gigabyte UD-9

980x @4.50 24/7

PSU 1300W

2000MHz 7-7-7-20

Noctua NH-D14

Cricial SATA III C300 128GB

OCZ Z-Drive PCI-e 256GB for FSX

Zotac GTX 480 AMP<


Posted By: akriesman
Date Posted: July-13-2009 at 10:40pm

Originally posted by kiwikat kiwikat wrote:

Don't forget the "X" either...

Because we REALLY need another 3 letter acronym ending with X.

I've got a quick question.  Will there be any trees in neighborhoods or just in areas determined by the vegetation landclasses?  I see there are SOME trees in the previews but how are they implemented and how will they work in a larger city where only residential areas have trees?

I can't wait to see some more screenshots! Star

Trees can be defined in both the landclass and along with the housing objects (using vector autogen).   Although, there are some limits as to how the trees can be defined using the vector autogen method.

The next best chance for screenshots will take place when we get to the "project X" part of Alaska.     Anchorage, Fairbanks and a couple other large Alaskan cities will use the vector-housing methods you have seen discussed in this thread.   The smaller villages, towns and seaports will have hand-placed houses using the semi-custom object part of project X.

Working on the "Project X" parts of Alaska are going to be a lot of fun.   Right now, we are still working on the UTX parts of Alaska.  



-------------
Allen



Posted By: steve_1979
Date Posted: July-22-2009 at 2:13pm

This all looks incredibly impressive - I can't wait to try out UTX Alaska.

What does it look like on a lower powered computer where you have to reduce the number of autogen objects to run it at a decent frame rate?




Print Page | Close Window

Forum Software by Web Wiz Forums® version 12.01 - http://www.webwizforums.com
Copyright ©2001-2018 Web Wiz Ltd. - https://www.webwiz.net