Jump to content
The simFlight Network Forums

GaryGB

Members
  • Posts

    3,123
  • Joined

  • Last visited

Posts posted by GaryGB

  1. Hi Pups:

    Gas Bar?!!! :shock:

    Spoken like a true Emmaite! They serve E-Rations that far north? :lol:

    Good hygiene for your aircraft with your EFFC bucket and squeegee set should help prevent a case of the "Bird Flew"! :twisted:

    PS: Cool ultralight and livery; where'd you get that one? :roll:

    GaryGB

  2. Hi All:

    I was studying up on FSX hardware vs. performance issues while awaiting a reply from tdragger or someone from Microsoft on the issue of whether FSX is definitively and efficiently written to benefit from multi-core CPUs, or if it merely has more than one thread which the operating system may, at it's sole discretion, choose to run on a second core at a given point in handling its total workload.

    This, I believe is a process which users cannot control unless the software is specifically written to allow user control over what thread goes to which core, and I'm not so sure Microsoft would want to even allow that to take place because if problems occur, Windows might initially look bad for not being able to cope with too much user control. Anybody remember the collisions we used to encounter with older and poorly written apps (mostly 16 bit) which wouldn't obey the traffic management that Windows was trying to promote for the 'efficient throughput of all tasks'? In short, multi-core "process or task scheduling" as it apparently is called is otherwise totally under the control of the operating system.

    ( See: http://www.devx.com/amd/Article/29117 for its initial explanation of this )

    Multi-threaded application control in a multi-core setting seems to be quite different than merely assigning a priority level for a process in the Windows XP Task Manager with an extra option to somehow pick a core and "tell it where to go" for operation; we might currently be able to sit there and angrily 'tell Windows where to go', but not a specific application thread!

    Pondering this issue in regards to FSX and the desire multi-core system owners would have to maximize FSX peformance by having threads critical to perceived smoothness of operation and framerates running in the most efficient multi-core scenario, I started thinking about all the problems the computing industry has had getting application program software companies to write for specific use of multiple discrete neigboring processors per mobo in the past, and currently now for multi-core per die CPU chips.

    I remember the controversial discussions about sophisticated compilers which always seemed unlikely to ever be able to sort out complex program code and semi-automate the process to make such programming more cost effective to allow benefit from use of more than one CPU chip regardless of its location (whether on a single chip die or elsewhere on the mobo). Few companies ever found it cost effective to attempt to do this unless the performance of their product would otherwise suffer and affect sales, and/or the consumer response to the sale price of the product within its specialty niche market would still allow the costs to be recouped.

    I heard rumors in the past about plans to put a kind of artificial intelligence into CPU chips themselves that would allow only very slightly modified programs to run threads on other CPUs and increase application speeds, but nothing ever seemed to materialize. But now, I believe there may be technology already in some multi-core CPUs to allow single-threaded applications to still benefit from multi-core processors independent of any Windows DX10 api!

    I bought the US September 2006 issue of Maximum PC magazine, and on page 12, there was a news item which I quote below:

    "Reverse-Hyper-Threading to Arrive Sooner than Expected?"

    "In July, we speculated that one way AMD might be able to compete with Intel's Conroe CPU is to implement reverse-Hyper-Threading. By having two cores appear to applications as one super-core, multicore CPUs could deliver more performance to single-threaded applications. The latest rumors hint that not only is the tech real, but it's already implemented in AMD's new socket AM2-based CPUs. Even more surprising, it seems Intel might have similar technology embedded in Conroe. An image popped up on the Internet of an Intel 975x mobo with a 'Core Multiplexing Technology' option available in the BIOS. We contacted both companies about this issue, but neither Intel nor AMD would comment on the rumor."

    That's exciting 'news'... even if a rumor! It kinda' makes one do a 'double take' on the parallels between "RAID 0" and "RING 0" doesn't it?

    Perhaps all our games, sims, applications, utilities etc. which are not written to take advantage of multi-threaded operations on multiple CPU cores could benefit. Like FSX as well, perhaps, if it proves to not yet be written to effectively derive major benefit from multi-core computers? (...Still waiting for a reply here, folks!)

    But wait... would Microsoft really want this to happen? One might reasonably assume that they always want to prevent loss of market share to competing operating systems, some of which are already out in use, and others are rumored to be under construction somewhere in Asia to be bundled with new offshore PCs as I write this. Wherever Microsoft can put something proprietary into Windows that provides incentive for consumers to buy and use it (and the software programming kits that talk to the exclusive Windows APIs, and the training courses and publications to educate one on how to better use the exclusive Windows APIs), they likely will do so. Which might merit taking another look at what's going on with FSX, DX10 and Windows Vista.

    Ever wonder what FS flight dynamics performance would be like if one could shunt the physics calculations off away from the main CPU to another processor? Sure, we all probably have, after all, we've been wondering for a while now what FS graphics performance would be like if one could shunt the graphics computations off away from the CPU to another processor (like the GPU on our expensive new video card ...please!).

    Currently, a lot of what we see in FS9 apparently is done via DirectX in conjunction with the FS rendering engine operating in memory. Most of us by now are running DX9.0c, and have heard tales of what DX10 might do for us in FSX when it is released in Windows Vista.

    Below is a glimpse (by artist's rendition) of what Microsoft theorizes FSX will look like under DX10 compared to DX9. Gorgeous isn't it? Well, it reportedly will only be available through Windows Vista, and everyone will have to buy a new DX10 capable videocard to see it because supposedly no one has such hardware out for anyone, even Microsoft, to test, much less to buy.

    < Edit: I found an online article with even better detailed images which I substituted below; the following is an excerpt from a series of articles entitled: The People Behind DirectX 10: Part 4--Game Developers. >

    See: http://www.extremetech.com/article2/0,1817,00.asp >

    Flight Simulator X—Carl Edlund

    Carl Edlund is a Software Development Manager at ACES Studio, the internal Microsoft Games studio producing the big tenth major installment of Flight Simulator. Vista was originally scheduled to ship late this year, around the same time as Flight Simulator X so the title was always planned to be somewhat of a showcase for the new operating system's focus on games, and that includes DirectX 10.

    ExtremeTech: What is DX10 allowing you to do in Flight Simulator X that you couldn't do simply with a more powerful DX9 card?

    Edlund: DirectX10 not only allows you to get even more of the advanced graphics features, at an even better performance, but as part of the Unified Driver Model, the Memory Manager and Scheduler on Windows Vista create a far superior user experience. Games will be able to provide multiple 3D view windows at the same performance of a single DirectX 9 window. By shifting more of the graphics processing tasks to the graphics hardware, the CPU is freed up to focus on other gameplay related tasks like better AI, more in-scene entities and deeper physics, thus enabling entirely new gameplay experiences.

    ExtremeTech: It seems like nobody has actual DX10 hardware yet. Are you doing most of your work in the DX10 reference rasterizer? Simply making more complex shaders along a DX9 code path? "Render to spec" work in a program like 3ds Max or Maya? Exactly what methods are you using to do DX10 development at this point?

    Edlund: At this point, we are ensuring that the abstraction layer between the game and the graphics calls are clean and that we can swap in the DirectX 10 pipeline with minimal impact to the game engine itself. We have worked extensively with the DirectX 10 SDK, the DirectX 10 team and the hardware manufacturers to create a detailed plan of what features we will be taking advantage of and how we will integrate the code and art assets into the product pipeline.

    ExtremeTech: Last we heard from you, Flight Simulator X would ship with merely DX9 support, and then a DX10 patch would be offered later. How soon after the January launch of Vista do you hope to release this patch?

    Edlund: Yes, this is our plan. We are currently polishing and tuning our DirectX 9 pipeline to provide as detailed and smooth an experience as possible. When DirectX 10 hardware becomes available, we will create a parallel effort to port onto DirectX 10 and implement the feature extensions. We still plan on having this available at the time of Windows Vista launch.

    ExtremeTech: DX10 is coming really late in the project for your game, so there's only so much you can do with it. Could you discuss some of your plans to possibly further exploit DX10 in future expansions or the next version of Flight Simulator X?

    Edlund: Integrating DirectX 10 into our application will allow a much richer and more detailed world to be rendered at equivalent frame rates. Over the next releases, we will be continuing to increase the level of realism, detail in the world and world objects, dynamic lighting and subtleties, weather effects, and extending the capabilities of user interaction, exploration, and camera control in our world-spanning rendering environment.

    ExtremeTech: In your own words, why should gamers care about DX10 and fork over the money for a hot new DX10 card when they come out?

    Edlund: The performance and increased visual detail alone should be enough to make you run screaming to get the new cards. DirectX 10 will allow developers to take advantage of more of the DirectX 9 functionality, in addition to the new DirectX 10 features, to create more realistic scenes, not just showcased objects, then ever before possible. Supporting this is a more stable and coherent driver architecture that will end the days of driver crashes interrupting your game experience. By freeing the CPU up to do more game-related processing, you will see a quantum leap in the depth of AI, physics, character interaction and realism that becomes inherent in the next generation of games.

    An article on page 8 of the US September 2006 issue of Maximum PC magazine entitled: "Get ready for Direct X 10" also makes an important point that "One feature that won't be included in DirectX 10—at least not at its initial release—is a physics API. "We are working with the companies that currently make physics engines," said Chris Donahue, Microsoft's director of business development for Games for Windows, "but we have nothing to announce at this time."

    Interestingly, on Page 10 of the same issue, there is also a news article entitled: "Microsoft Wades into Physics Battle" containing some intriguing information.

    It reads: "We ran a rumor mill story in July (2006) about Microsoft's plan to introduce its very own physics API-dubbed Direct Physics. Now, two months later, there's a job opening on Microsoft's website for a "Direct Physics" programmer. The ad states that the company is looking for an engineer, "to join a growing team responsible for developing Direct Physics."

    It makes perfect sense for Microsoft to wade into the physics scrum because, ultimately, having a single, unified physics API will benefit consumers and developers (as long as it's as good as, or better than, what Havok and Ageia have to offer). Rich Wickham from Microsoft's Games for Windows group told Maximum PC he wrote the ad almost a year ago, but insisted that the company has no plans for a physics API. Wickham stated that clearly physics are up-and-coming, and it simply wants to be on top of the situation."

    But in case there is a delay in release of a DX10 Direct Physics api, hardware manufacturers may come to our rescue again. In an article at: http://www.xbitlabs.com/news/video/disp02331.html entitled: "Graphics Processors Capable of Physics Processing – ATI GPUs Will Process Physics, ATI Says" there is a description of plans ATI may have for using its graphics cards as physics processors.

    The article states: "In addition to cost advantage over dedicated physics processors, GPUs also have higher bandwidth to send vertex data to itself after physics is processed. According to ATI, in case mainstream physics engines, such as Havoc, support physics calculations, the company will not need any special application programming interface in order to allow game developers to take advantage of capabilities".

    Hmmm... one might wonder if Aces Studio actually did "create a parallel effort" in the form of other "beta threads" of FSX which are "scheduled" "to implement the feature extensions" for "all" the DX10 API's still being finished for Windows Vista and whether they "still plan on having this available at the time of Windows Vista launch" which will allow revolutionary changes in how physics, graphics and overall performance is rendered to the screen in Flight Simulator on DX10 enabled hardware. :idea:

    Perhaps one of the new FSX "missions" at release time will revolve around a flying carrot with an FSX logo on it which we must follow to a hangar painted with a Windows Vista logo, via a control tower with DX10 on it in high visibility colors! :twisted:

    We can only hope that we might get our hands on a DX10 enabled "Vista-matic FSX" around the time that Windows Vista launches, instead of having to spend the next 2 to 3 years saddled with DX9 class single core CPU-rendered physics and graphics technology, and our costly computers built for a truly screamin' FSX experience could be outdated by that time! If Aces Studio keeps its commitment above, we can then all see how "DirectX 10 will allow developers to take advantage of more of the DirectX 9 functionality, in addition to the new DirectX 10 features, to create more realistic scenes, not just showcased objects, then ever before possible"; and if we get excited enough by that, we can grab our wallets and "run screaming to get the new (DX10) cards". :roll:

    PS: ...Still waiting for a reply here, folks! Oh, and see:

    http://download.microsoft.com/download/_Cores.ppt

    GaryGB

    post-14010-128689423288_thumb.jpg

    post-14010-128689423355_thumb.jpg

  3. (Pups who periodically reverts to the 3rd person, starts musing to himself upon hearing that a 32 MB card and a 1 GHz processor will cut the mustard in this day and age of modern sanitary pedestals.)

    Pups (cub reporter for the Emporium’s Daily Bugle)

    Hey Pups, you wouldn't be alluding to the purported hardware minimums for FSX by any chance , would ya'? :lol: :lol: :lol: :lol: :lol:

    GaryGB

  4. Hi Justin:

    Thanks for clarifying those important points! :D

    I was concerned that the machines were unable to do justice to the (then available) FSX demo build you had installed to run under the "hellish" ambient temperatures and humidity which you all had to endure for so long! :?

    Talk about a trial by fire... you guys deserve a medal for your efforts! :lol:

    I have edited the original post regarding my "understanding" of the Beta demo build running that day in an effort to minimize confusion. :oops:

    To explain the reference to 2 GB of RAM: I heard one of the other reps state to a party who was pressing for bottom line details on a hardware shopping list to make FSX run with 'maximal efficiency', that although 1 GB of RAM would be adequate to run FSX, 2 GB would be advisable along with the most powerful CPU and video card available to one's budget to be better positioned for the next 2 to 3 years life cycle of FSX. I did not personally happen to hear what you might have said to anyone regarding anticipated "official" RAM requirements, and I had not remembered to ask you or any of the other reps what the installed RAM was in the demo computers that day (or what the hard drive configuration was, if you could clarify that too please).

    Thank you so much for taking the time to show us glimpses of the San Juan Islands, and in particular Orcas Island, where you had been flying in the real world for Kenmore and Horizon Air. It was really great to hear of your experiences there, and to learn of your personal acquaintance with Richard Goldstein; it brought the FS world to a new height of excitement and anticipation as we all dream of flying there soon in FSX!

    I understand that you are involved in some way with an upcoming scenery for Portland Oregon, but I forgot to write down what the publisher's website address was; could you post it here so we can see what you're up to in the future? :roll:

    Here's hoping to hear from you more often at the EFFC forum! :P

    GaryGB

  5. Francois:

    I think you gave some very good advice to would-be FSX users who want to be doubly sure that they are not rushing an expensive hardware upgrade for something that may or may not significantly benefit from multi-core CPU and/or multi-GPU video card array.

    However, I hope that if Microsoft sees this as a growing sentiment in the FS community prior to October, it doesn't rest on its laurels and assume that we are all going to just sit back and wait until after the launch date to plan and buy our computers. I believe that statistical studies have shown that games and sims (including FS) drive the hardware market, and that Windows upgrades drive the software market. For the $50 and $70 USD that the new FSX versions are reportedly going to cost, I don't think Microsoft is worried we are going to run out of money to buy FSX after buying our expensive hardware. And a Windows Vista expense won't be in people's budgets until 2007 anyway, so I don't see any point in Microsoft not telling us NOW what we might wish to know about realistic FSX hardware requirements and multi-core capabilities, and the advisability of 3rd party add-ons being programmed as multi-threaded apps to take the load off the main CPU core where the bulk of the FS rendering is likely to still be taking place (assuming FSX is ultimately disclosed to be definitively AND efficiently multi-threaded!) Any further reticence on this issue is likely to further stall the hardware market and pi$$ off the hardware manufacturers, who, last time I checked, were the largest market for (indirect) sales of (bundled) Windows... Microsoft's biggest cash cow.

    Hmmm... one might wonder if reticence on the real world FSX hardware requirements could be related to a desire for deferred hardware sales so that the new Windows Vista operating system will be available for all the new computers being built after the FSX release and after the holidays!

    And I'd hate to think that we are going to see another ridiculous Microsoft Logo Certification 'do-wop' wherein 3rd party FS developers are compelled to submit their creations to an "FSX-compatibility certification test program" for fear that consumers might mistrust their product's impact on FSX as the CPU's already demanding main squeeze. I could almost see Tom Allensworth wondering if he should allow "non-FSX logo certified" freeware onto the AVSIM servers if things got too far out on this idea! :mrgreen:

    Pups:

    Heck, I thought every app we run as EFFC members was 'Elitist' software! :twisted: :lol:

    I am confident that Microsoft has already adjusted the scale of the FSX slider controls for the rendering engine to be able to throw out even more of its potential gorgeous details than FS 8 & 9 do already in order for FSX to run at least 12 - 18 FPS on older computers; nothing kills sales like a "Fright Stimulator slideshow".

    The question is, how much detail can one see in the textures of the aircraft and scenery, and how much terrain mesh contours will be visible after "pulling back" on those sliders to accomodate an older computer even before loading various add-ons and utilities that run in the background?

    Perhaps in equitable fairness to FSX, how many 3rd party FS add-on publishers can or will tell us now that their utility apps are multi-threaded so that they are even capable of being routed to a second CPU core assuming "Windows DuJour" was able (and in the mood... under a given set of workload parameters) to send it away from the main CPU core to run?

    This is why I am concerned that someone from Microsoft disclose something substantial and authoritative on this issue now so we can ALL start planning our purchases and allow for 3rd party add-on re-tooling to be done early. Thank you, Microsoft, for including the SDK with the DVD on the shipping date, but the rest of us out here who aren't partnering with you (yet!) needed to know at least some basic things in advance so we can be better prepared for the FSX launch date.

    Personally, I'd run FSX as a slideshow for its new textures like I did FS2002 years ago until I got a new computer, then I ran FS2004 as a 'pausing, blurry non-slideshow' until I got a new videocard, because I loved its new terrain textures.

    As of this last week, I already know that Ultimate Terrain will provide me with more detailed coastline and landmass terrain rendering than that included in FSX.

    ( see: http://www.simforums.com/forums/forum_p17884&PN=1 )

    And although I haven't personally tested Flight Environment and Ground Environment, from what I see in FSX, I will likely be getting sky, ground, and especially water textures that will keep me happy without an additional purchase to enhance them.

    Bird's Eye View would have to improve on the extensive autogen details provided in FSX to earn my upgrade purchase, but I thank them for making it possible to fly FS9 with their enhanced autogen as I do now.

    I will be seeing more details with Steve Greenwood's excellent FS-Traveler 10 and 30 meter high LOD ('sharp edged') terrain mesh for the entire US which I already have, so the FSG-derived 38.2 meter USA terrain mesh "enhancements" (with the characteristic FSG 'blunted edges') apparently included in FSX will not likely matter to me.

    I also think that we will be able to run FSX on Windows XP, but the question is, how much of the enhanced performance besides DX10 will require Windows Vista as an incentive to compel upgrades for a subsequent new cash infusion for Microsoft and ultimately for 3rd party application software publishers?

    I agree with Don that you should not have a problem re-installing from your Windows XP 'restore disk' image (assuming it is not actually a Symantec "Ghost" image), but I would do it onto a new hard drive, preferably a 2-drive SATA RAID 0 for faster (2X) throughput on an add-in hard drive controller card, and keep your existing drive as a backup still connected to the motherboard hard drive controller along with your CD/DVD drive; that way when you move to a new motherboard, you can take the controller card and attached RAID hard drive(s) with you intact, since RAID striping compatibility is unlikey even among a given manufacturer's embedded RAID controller chips on various mobo's and external add-in cards. :idea:

    GaryGB

  6. I could be mistaken, but I seem to recall a task manager screenshot that tdragger showed on a dual core machine back in 2005. On that shot both CPUs were being used, which he later revealed was FSX.

    Yeah. I don't know where the dual-core rant came from...

    Greetings tdragger! :lol:

    I thought that post might get some attention! Besides the fact that I usually write my "Rants" tongue-in-cheek as a satire of an issue for some imposed levity, I wasn't so sure I followed what the bottom line was in regards to whether FSX will, or will not, benefit from running on a multi-core CPU machine. :)

    I could not get any clarity or commitment from what I heard at the Oshkosh demo, and the posts I have read thus far in SimFlight and AVSIM have not clearly stated whether FSX will have smoother flight dynamics and run more efficiently with less pauses, blurries, stutters, "sequential" rendering of aircraft components, and low framerates in high autogen object and BGL scenery object areas if a second core is used in the appropriate machine. :?

    If FSX will benefit in some or all of these areas, will it do so in Windows XP, or will it require Windows Vista to implement the ability to utilize a second CPU core? :?:

    And if FSX benefits from use of a second CPU core, how much is it likely to benefit regardless of optimizations being done on unrelated issues between now and October's release? :?:

    This is critcally important to us buyers who have personal rather than corporate budgets funding their purchase of a new computer bought largely for FSX! I don't think FSX will sell less if the percent improvement is small, since current cache memory, bus speeds, disk I/O, overclocking, 2-GPU SLI or CrossFire etc. all still tend to yield small incremental speed enhancements to a system, but the commitment by Microsoft to begin using at least a second CPU core NOW over the next 2 to 3 years in FSX would ensure happy users and good word-of-mouth PR which should in theory contribute to better sales. Yeah, we'd all love to think we could double our speeds by adding a second "whatever" into our PC, but that rarely has happened over the years, and people have learned to be happy with what they were able to achieve as long as "measurable" performance increases could be "detected", or better yet "experienced" during simming, while they await further technology improvements. :idea:

    Thanks very much for any detailed, clear explanantion you might be able to disclose to us here at EFFC, one of the largest groups of vocal FS consumers on the internet! Any links you could post for us to view other and prior discussions and disclosures on this issue would be very helpful as well. :wink:

    PS: The "" was from my subconscious, where I have some apprehension about whether FS enthusiasts (read: junkies) will be frustrated with FSX for 2 to 3 years if it does not benefit adequately, if at all from their hard-earned money being spent on a multi-core CPU computer upgrade with a multi-GPU video card. There's big money at stake here for a lot of us when it comes to buying and/or building a genuinely "powerful" new computer, and as stated above, it will be more money than is going to be spent on Microsoft's software for FSX and Vista combined, so we need something definitive from Microsoft about whether and how much FSX will or will not benefit from running on multi-core CPUs, so we can budget, plan, and schedule our computer hardware purchases accordingly! :roll:

    GaryGB

  7. Hi Don (Fragmentum):

    Sorry I couldn't catch up with you at Oshkosh this year, my schedule was up for grabs and generally too limited for a full day at EAA AirVenture this time; perhaps in a coming year? :wink:

    Thanks for the link to your post in the FSX Forum pages and your nice sharp pix; from your PM I know you are busy with your transition right now, but I was wondering if you get some time, could you post some of your FSX demo experiences into this thread for the other Emmaites to expand the FSX first-hand knowledge base? :roll:

    To all those otherwise inquiring:

    Regrettably I was only able to check out a small subset of the features in FSX due to time limitations at Oshkosh, and in consideration of others eager to try out the FSX demo at the limited number of available stations. Weather wasn't one of the things I checked out, as personally I tend to always fly in the summer with modest, fixed weather settings while primarily enjoying the scenery and doing amphibious touch and go's.

    However, I did watch a "FS 10 Preview Video" yesterday linked from the AVSIM-FlightsimTV page: http://www.airsidetv.com/flightsim/default.asp wherein Microsoft's Shawn Firminger, manager of Aces Studio, "explains the new features of FS 10 and brings us up to date with the latest graphics performance on the way with Windows Vista"; in that interview he mentions the weather engine and frequency of updates, shows numerous interesting aspects of FSX and some ground clips of animals. Choose "FS X Preview Video" from the pull-down list box below the helo image on the top left of the page, and your connection will begin cueing with prompts.

    I didn't see any animals where I was flying (over an area I know like the back of my hand: Lake Wisconsin, 20 miles north of Madison, WI), but I did see the brief segments on the tail end of the "helo rescue" demo video, and they looked adequately realistic with smooth movement viewed from both the air and the ground nearby, although TDragger said elsewhere that video was "speeded up slightly" when sampled to PC video.

    The autogen trees seemed to blend quite well with the forest textures; I didn't fly in an area with buildings and other objects, so I can't comment on them. The FSX publicity screen shots posted elsewhere suggest they will be adequately textured, abundant in number, and fairly accurate in size and position due to the NavTeq data for streets being incorporated into FSX.

    See also these news articles:

    http://www.simflight.com/modules.php?nae&sid=8167

    http://www.simflight.com/modules.php?nae&sid=8177

    http://www.avsim.com/pages/robertw/FSX/FS-X.htm

    http://en.wikipedia.org/wiki/Flight_Simulator_X

    See also these pix:

    http://fsinsider.com/screenshots/default.htm

    http://www.simflight.com/modules.php?nae&sid=7675 ( includes the "ultimate" Beaver shot! )

    http://www.simflight.com/modules.php?nae&sid=7503

    http://www.avsim.com/pages/0106/FSX/fsx_thumbs.html

    http://www.simviation.com/cgi-bin/yabb/1154316551

    http://www.flightsim.com/cgi/kds?$=mainsx0104.htm

    http://www.fs2004.com/gallery/thumbnails.php?album=7

    http://www.sim-outhouse.com/sohforums/shp?t=14412

    http://www.gamespot.com/pc/sim/microsoftml?page=1

    [EDIT - New URLs I discovered, showing some European areas:]

    http://guillaume.pages.free.fr/screen1.htm

    http://guillaume.pages.free.fr/screen2.htm

    See also these videos:

    http://www.simflight.com/modules.php?nae&sid=8002

    http://www.flightsimx.co.uk/

    http://www.gamespot.com/pc/sim/microsofmedia.html

    [EDIT - New URLs I discovered:]

    http://guillaume.pages.free.fr/videos.htm

    http://guillaume.pages.free.fr/videos1.htm

    Enjoy! 8)

    GaryGB

  8. Hi All:

    Here are some details of the FSX demo I flew hands on several times on Friday afternoon. :!:

    First, some important disclaimers: this is based on what I understood to have been described as a "beta-2" or possibly later build of FSX, and what is released in October will likely be different and much improved. It is also an informal review of my own observations, and a composite of what I saw and heard during the demo by both myself and others at the FSX demo stations, conducted with no special notification given to the demo team. Regrettably, I had not noted the specific slider settings in use as I eagerly began using FSX after waiting my turn in lines that were 3 to 4 people deep at each of the 2 available functioning demo stations.

    Also, the 3.X GHz Intel computers running ATI 1900 PCI express video cards (with 2 GB RAM?) were reportedly suffering from overheating issues due to the unusually hot weather at Oshkosh which was running upwards of 98° Fahrenheit with high humidity. Only half of the 4 machines were in operation for purposes of demo by users, one was completely turned off and presumably not working, the other was "being checked" extensively for stability via "presentation mode flying" by one of the Microsoft-authorized FSX demo representatives; I won't mention the name of the one who was doing all that "error-detection" flying! The FSX display booth console cases in which the computers were mounted had nearby large fans blowing into the open ends of the cases with some incidental benefit in airflow to the users standing at the stations performing their interactive demos.

    The 3 FSX demo representatives present at the time of my demo were generally courteous, well-informed, and somewhat more subdued in their interactions with attendees than I have seen in previous years, and in my opinion were discreet, if not cautious or even "guarded" in their responses to probing questions by users about hardware requirements needed to make FSX run smoothly. A consensus of responses I heard while present generally was that FSX would be demanding on hardware as far as CPU speed and graphics display card capability, that 2 GB of RAM would be advisable as a minimum, and that one should buy the most powerful CPU chip and graphics card available and accessible to one's budget at a given point of purchase, since FSX is designed to "push the envelope" of computer performance for a period of up to 3 years per FS release life cycle.

    It was not clear as to whether there would be, by release time, any significant concessions to accommodate multi core CPU chips, that the feature set for FSX was pretty well fixed by now, that special programming is required to allow applications to properly utilize multi core technology, and that FSX performance optimization would be the primary focus between now and October. However, "future releases" of FS reportedly would begin seeking to implement offloading of an increased portion of rendering tasks to GPU chips rather than through intensive reliance on a single CPU chip as is currently done in FS.

    The "helicopter oil rig rescue" video was playing on two of the multiple approximately 24" diagonal LCD screens in the FSX demo area; as rendered media files, they did not have the same degree of resolution that was seen on the screens for the FSX interactive demos, and this point should be kept in mind as one plays those demo videos on one's own machines: FSX definitely looks even better live!

    The user interface screens presented a bright, clean, easy to use pick list for the many aircraft and repaints included with the demo; those screens wrote and scrolled quickly and smoothly. Other aspects of the FS user interface were changed (I think for the better), and were easily adapted to.

    After picking a plane, livery, and parameters for the flight, the load time began... I would say an average of 15 to 20 seconds was required to render the initial screen scenery, followed in another 5 seconds by the shell of the cockpit, followed in another 5 seconds by the remainder of the aircraft, followed by a screen blank of another 5 seconds, followed by the fully rendered scene for an average of about 30 seconds to start one's flight. It is unclear whether the Intel machines were using RAID drives; I'm inclined to think that they were not (based on my own personal experience with running FS9 on various 7200 rpm parallel Ultra ATA 2-drive RAID 0 machines, and a 4-drive RAID 0 machine). If one crashed the aircraft into the ground, a blank screen of approximately 10 to 15 seconds would eventually occcur before getting to the next screen.

    Aloft, the the Beaver and the Goose handled well with overly conservative flight dynamics apparently having realism set for novice demo users; they had nicely detailed cockpits and showed simplistic default water landing effects. The sound was turned down so far I could barely hear anything of the engines, and there was apparently presentation style fanfare music playing in the background during all flights. I was tempted to think that the sound files were not finished yet, but it may simply be that Microsoft was trying to be considerate of other FSX demonstration stations and other vendors nearby in the hangar. When I asked one of the representatives if all of the sounds were in surround 5.1, or merely a combination of DirectX multi-positional "mono" sound and some ambient and/or surround sound in external views, he could only say (guardedly?) that "sound counts" when choosing a sound system for one's machine to run FSX. [EDIT]

    The new quasi-photorealistic terrain and 'data-derived' water textures are excellent, as are the gorgeous ultra-ultra high detail aircraft textures. The terrain textures were like a hybrid of existing FS9 data texture tiles with large areas of newer landclass textures well blended in a somewhat cooler color balance. There were still some blurries and pauses when changing spot view orientation with the hat switch.

    The USA has a 38.2 meter terrain mesh reportedly for "everywhere", and the geopositioning of landmasses and coastlines, and landmarks is likely to now be very close to the real world, however the number of data points for coastlines appeared to be lesser than that seen in Ultimate Terrain. But overall this will revolutionise the future for commercial and shareware add-on sceneries as nearly everyone's scenery should be relatively properly positioned geographically with the FSX default. The bad news is when FS2004 came out and more people began using 3rd party terrain mesh, scenery makers should have seen the future and positioned according to 30 meter mesh; now they will have to "correct their FS2002 and FS2004 corrections" since they set their goals then on volume sales based on people using the "default geographically incorrect out of the box". Volume commercial scenery sales and shareware scenery popularity must now be anticipated on a 30 meter FS mesh default; ...it's about time!

    The shoreline/texture blending had not yet been fixed and there was still water bleedthru on the land side. When flying low and slow, the brush and treetop textures if seen from directly above look somewhat odd when distributed over the flat plane surface of a land polygon; in these situations, a higher mesh setting will be imperative to maintain realism, and this may have been due to mesh detail being purposely turned down to achieve decent performance during the FSX demos running on a very hot day.

    In the above "nearly top-down view" scenario, the impression came to my mind that a new algorithm was being used in the rendering engine to throw out mesh details differently than in FS9 with the anticipation that fliers would not see the lack of terrain mesh detail in areas with relatively similar elevation data points due to the suggestibility factor provided by the new vegetation textures. Consequently, my impression was that areas with sparse numbers of trees and bushes may suffer from the "flat look" artifacts described above if more terrain contour shadowing textures (which we had almost too much of in FS 8 & 9) are not re-implemented in some limited measure for those areas.

    The terrain texturing realism was generally good overall for flying in a manner other than low and slow, as the terrain textures when distributed over hillsides (with a somewhat increased number of autogen trees compared to FS9) blends so well, that with individual treetops projecting above the horizon line on hilltops at a distance, the effect has very pleasant 'real world' appearance in heavily forested areas. The land class look up information does appear to be more detailed and accurate, and autogen may already be tied into the NavTeq data as recently announced on the FS news sites.

    I had the pleasure of meeting Justin Lamb, now based in Portland Oregon, who was there on behalf of the FSX demo team as a "special beta tester" in his capacity as a considerably well trained and experienced pilot in the past for Kenmore Air and Horizon Air. Justin was kind enough to discreetly share some of his experiences flying in the Seattle and San Juan Islands area, and disclosed that he was the pilot who flew Richard Goldstein around the area when he was preparing for production of his Orcas Island scenery. Justin was flying a twin-engine aircraft around Orcas Island in FSX, and it looked properly shaped and wonderfully textured as he took his nostalgic tour around the area, and the nearby ocean water textures were out of this world! He indicated that there was likely to be some future changes to make the watercolors more accurate to the real world. However, I personally absolutely loved the bluish tint of the ocean water, as it reminded me of the (static) water seen off Half Moon Bay west of San Francisco international in Flight Unlimited 3, with nice high contrast highlights on the wave crests. He also indicated that some future effort was likely to be made to tone down the highlights for calmer waters, and to also implement wave directionality in keeping with prevailing winds.

    Overall the demonstration was an enjoyable experience, and I look forward to seeing the optimized, finished product in October.

    I must say that I have some apprehension about the efforts made by the development team to implement use of multicore CPU chips which are already becoming well established in the marketplace, and this could prove to have a serious and detrimental effect on the market success of FSX, as users become frustrated with both existing hardware and their newly purchased, expensive, multicore hardware not being able to run FSX as they had anticipated.

    Your attention please, Microsoft: as you know, FS is and always has been phenomenally CPU-intensive; in FSX you MUST divert some of the work to the 2nd core to get better performance, and let us get our money's worth out of our hardware (CPU's and GPUs) which are "far more expensive than your software". Quite frankly, if you claim that you're releasing software which is designed to run on the machines to be available in the next 2 to 3 years, and multicore CPUs are already in the marketplace, you'd better damn well start talking to those multicore CPUs now or you're guaranteed to piss off your existing and prospective user bases for the next 2 to 3 years with a program as capable and as demanding as FSX! And no, we don't want you to dumb down the FSX features to accomodate single-CPU-core-only application programming!

    Unless an initial press release and a service pack addressing this issue is made available within weeks or months of the initial release, I would be very concerned about the success that this product will meet in the marketplace as users anticipate waiting 2 to 3 years for Microsoft to address product limitations and/or problems "in the next release" as has often been the case with the majority of its products in the past. Please remember what happened with CFS3 and notice what happened with Lock On Modern Air Combat, Far Cry and Need for Speed Most Wanted to mention but a few: their sales and retail prices plummeted as users found that their systems could not run them adequately. Thankfully, however, this general Microsoft pattern of "wait for the next version for your fix" was changed with the release of the FS 9.1 patch, and if they have not implemented multi core CPU utilization in the October release, perhaps we will be given "the gift that keeps on giving" by seeing one shortly after the October 2006 FSX release. "Sooner" will be far more advisable than "later" on the multi-core issue... never ask a junkie to wait for his "fix"! :twisted:

    I would like to thank Microsoft for making the demo available to us all to try out, in a rather pleasantly surprising change towards "openness" with its users, as we saw starting with the early FS franchise continuation announcements, screenshots, corporate and development team representative interviews and blogs. I believe this trend of open communication will prove very beneficial to the development process, and user enjoyment of Flight Simulator. I can hardly wait for October's release! 8)

    GaryGB

  9. Hi Madaz:

    I did get a very few pics aside from FSX shots, as I spent most of my limited time at the Microsoft FSX demo display this year.

    It will be a few days until I have them developed, and the negatives scanned (not using a digital still camera; still too heavily invested in a professional 35mm SLR film 'system'!)

    In prior years, I did take a good number of photos, and only briefly visited the Microsoft 'booth' at the entrance to the exhibit hangar; this year it was the other way around!

    :idea: To enjoy the sights of the EAA Airventure, go to: http://www.airventure.org/2006/gallery/

    GaryGB

  10. Can't understand those efforts to sell graphic cards by the use of beutiful girls :?:

    Most geeks would'n recognize one even if they sat on them :wink:

    /Bqt

    Hi Catflyer:

    If a geek reached the point that he no longer recognized a beautiful girl because he had been occupied with his computer for so long, he would probably benefit from a very prolonged power outage! :lol:

    GaryGB

  11. I just got back from Oshkosh a few minutes ago. I am excited to share with you all the details of the FSX demo I flew hands on several times on Friday afternoon. :!:

    First, the really good news: it actually is scheduled to be out in October! :D

    I will be sharing my review, and some video and photos taken off the LCD screens shortly. :idea:

    Overall, an exciting upgrade beyond what we have now in FS2004! :shock:

    Here's what was on the cards given out by Microsoft at the demo setup location:

    GaryGB

    post-14010-128689420933_thumb.jpg

    post-14010-128689420946_thumb.jpg

  12. Hi Shamir:

    You previously referenced some us as "nut heads" for our outrageous humor which we sometimes post in the EFFC forum. :idea:

    I made a joke about us being OK with the term "nuts", with the clarification that we'd like to be thought of as 'nuts about FS' using the preferred term "wing nuts" (the picture of hardware items I posted above shows several "wingnuts" as they are called in the US and most english speaking countries (wingnuts are on the left column with your name at the top!) As you can see, Francois gets the largest one that is, of course, GOLD! :wink:

    And Fritz Deter (read your Emma Field Manual) is just a 'plain' big nut! :lol:

    Do a search on the EFFC forum for Officer Barbie, Mindy and Katy and you'll discover their histories here! :roll:

    I also made a joke about being a 'wing nut' here: http://forums.simflight.com/viewtopic.php?t=54081

    Unfortunately I was in a rush getting ready to depart for the Experimental Aircraft Association (EAA) Airventure held in Oshkosh, Wisconsin USA, so I didn't communicate too well. Fly around Wittman Regional in FS2002 (with the MSFS addons from MS' FS sites, or AVSIM, or it is partially included in FS 2004 ) and you'll see what I mean!

    I hope this helps explain my occasional obscure humor!

    ( See: http://www.airventure.org/ )

    Con Cuidado, Shamir! :D

    GaryGB

  13. Hi Pups:

    Thanks for the reassuring feedback. I've been under the gun for 2 days rebuilding one of the more complex of our mutiple computers in our office, so I'm not at my ususal peak of exuberance (not at the 'acme of my pinnacle').

    I'm frantically finishing up some local responsibilities this morning before I start the 3-1/2 hours drive to Oshkosh for an early and extended weekend.

    I won't be able to fly in to the seaplane base this year as originally hoped, due to a recent health crisis on the part of my colleague.

    I understand that FSX will be on display (and interactive demo?) at the Microsoft booth there in one of the 4 big hangars as is usual each year. I can't wait to see to it, and if I can, I'll get some pix and digital video for you all upon my return!

    Take Care, All... 8)

    GaryGB

  14. Helping new pilots? Here’s a story that will warm the coldest of hearts!

    One must have numerous degrees (in any acceptable discipline) in order to even be allowed an entrance interview with Gary, the ubiquitous master of flight!

    The entrance requirements to one of Gary’s classes are rigorous to say the least. One must be well-versed in aerodynamics, the physics of flight, the requisite knowledge of physics and math, plus Gary will only accept mature individuals that are determined in their maturity, to work towards the lofty goals that Gary himself has set.

    Because of this extremely difficult entrance requirement, the university has had to scour the countryside looking for suitable applicants.

    I never actually completed the 'Flying Training' in FS,

    Does that make me less capable as a Virtual Aviator?

    Whoops... looks like my 'visualization' attempt at humor crashed and burned on that one! :oops:

    Actually, in my haste, I was attempting post a little comedy about efforts needed to join the fictional "EFFC Elite Cadre". :roll:

    Anyone's approach to flying in FS is certainly fine with me as long as they are... enjoying themselves! :P

    I was merely carrying forth on some hilarity from the posts above, and from some "wingnut" posts at: http://forums.simflight.com/viewtopic.p3&start=60 :idea:

    Sorry if I lost or alienated anyone with my hasty posts above in this thread; its hard to avoid occasional mishaps when trying to participate in the forum with snippets of limited time one finds during the course of a busy day! :cry:

    I never completed the Virtual Pilot program either, Madaz; and when I fly in FS, I give VFR a new definition as 'completely free form anarchy' sometimes as I zoom aound enjoying scenery! :lol:

    Its probably the indulgence in those bad habits while in FS that have kept me from trying out multiplayer flying thus far, because I'd also have to behave myself in the virtual world like I do in the real world of flying! :twisted:

    Anyway, I hope all are enjoying their own individual style of FS aviation to their heart's delight like I do! 8)

    PS: I wish they'd had TV programs like "Jay Jay The Jet Plane" when I was a Kid!

    Happy (not 'hapless') Flying Everyone!

    GaryGB

×
×
  • Create New...

Important Information

By using this site, you agree to our Terms of Use. Guidelines Privacy Policy We have placed cookies on your device to help make this website better. You can adjust your cookie settings, otherwise we'll assume you're okay to continue.