Blurries and contention (round 2)

[PhilTaylor 45]

Sorry, it must be me because I still cannot parse your original reply. It doesnt matter now though as the discussion has moved on.Both of those cases you quote directly above are interesting.Changing a monitor is changing the hw config. Somehow that made a difference. Is it an 18-bit color monitor and that is the issue or is it something else? Its hard to say, but how is the sw at fault there? Having 2 identical PCs is another interesting case. And why I replied when it appeared we had just such a case. So thinking that case through, if the engine was at fault, you would expect it to fail everywhere if the engine clearly had a global issue.The fact that it works on 1 config and not on the other indicates there is something, as yet undiscovered, that is different betweeen the 2 configs and causing the issue. That feels config related and local not global.That isn't to say that once we discover the issue it might not be something in the engine, just that the data doesnt automatically scream that out.Again, if there is a global issue in the engine - how does it ever work anywhere? Since it does, that is an existence proof there isnt a simple global issue causing this problem. It most likely is some interaction between a local config-specific issue and the sw. I say "most likely" since this sort of "remote debugging by thinking aloud" is an exercise in deductive logic. And Occam's razor serves to direct our thinking to the simplest answer given a choice between 2 competing answers.

[Guest S77th-GOYA]

>Both of those cases you quote directly above are interesting.>>Changing a monitor is changing the hw config. Somehow that>made a difference. Is it an 18-bit color monitor and that is>the issue or is it something else? Its hard to say, but how is>the sw at fault there? Unfortunately for you and I and many others, I cannot say where the fault lies. That would be a huge step in resolving the problem. However, seeing the spectrum of hardware configs that are experiencing the problem it becomes just as hard to place blame on hardware just as much as software. It certainly is an interaction between the two. To me, that puts more of a focus on the software that must be made compatible with a wide array of hardware, drivers, etc. Granted, an unenviable task. The bottom line is that FSX IS compatible with that wide array but still has problems with systems that seem to fall into that array. And it can't be ignored that the majority of users reporting the problem associate the origin with installing SP1.>Having 2 identical PCs is another interesting case. And why I>replied when it appeared we had just such a case. >>So thinking that case through, if the engine was at fault, you>would expect it to fail everywhere if the engine clearly had a>global issue.Yes, and it obviously isn't a global problem. (Pardon the paraphrase but I balk at the use of issue when the best word is problem. I'm becoming crotchety in my middle age.) But I hope you'll agree that a developer simply can't ignore a problem that isn't global but exists on a substantial amount of hardware configs. That being said, I would be very interested to find the percentage of users here that have this and/or other problems. Perhaps we are in the minority enough to make it seem more of an aberration, but we're vocal enough to make it seem more widespread. Which begs questions along the lines of your beta-testing process, but I'll forego those.>The fact that it works on 1 config and not on the other>indicates there is something, as yet undiscovered, that is>different between the 2 configs and causing the issue. That>feels config related and local not global.>>That isn't to say that once we discover the issue it might not>be something in the engine, just that the data doesn't>automatically scream that out.>>Again, if there is a global issue in the engine - how does it>ever work anywhere? Again, the problem is obviously not global and I don't believe anyone is saying it is. There is too much evidence that many users do not have this problem.>Since it does, that is an existence proof there isn't a simple>global issue causing this problem. It most likely is some>interaction between a local config-specific issue and the sw.>>>I say "most likely" since this sort of "remote debugging by>thinking aloud" is an exercise in deductive logic. And Occam's>razor serves to direct our thinking to the simplest answer>given a choice between 2 competing answers. Occam's razor is a wise axiom but for the purpose of this problem and its elusive answer it seems to be inapplicable. There simply are no possible answers from which to choose. Perhaps a Gordian knot might be a better corollary.

[Guest The Wizard of Oz]

Phil,Youv'e posted text files. You need to press JPG.

[Guest]

Well it looks like I'm building a new rig this week in an attempt to cure my blurries, and if it doesn't cure my blurries then I get a 50% RTM peformance boost and I'm ready to play Crysis. I'm swapping for an EVGA 680i, with an E6750, and 2GB of DDR2 1066. I'll then format the partition, install XP, NIC driver, XP-SP2, the latest video driver, and then FSX followed directly by Acceleration. I'll let FSX install which ever version of DirectX it comes with, and let Acceleration update that as needed. I'll then test for blurries using the settings outlined above in my last test, with no other changes.Anybody have any other suggestions to help me aviod the illusive "issue with your config"?

[soaring_penguin 428]

Hey Phil, thanks for the correction ...I think this is very important, because not everybody (me included) know the inner workings of the sliders as you (ACEs) do.Jan

[Guest jcmckeown]

Been offline for a few days and pleased to come back & see Phil has jumped in. Seems we are confusing blurries and blurries...As I have mentioned before this is not a texture filtering issue. It is a problem with the higher level mip maps just not loading.Phil, on my config I tested one HD with a vanilla XP SP2 install and another HD with Vista. I had the blurries issues on both configs with SP1 and when I uninstalled and reverted back to RTM blurries were solved on both configs.The testing I have performed apart from being vanilla installs with up to date drivers on XP and Vista (+ all OS patches up to date) had FSX and I was testing on the FranceVFR scenery with small FS9 style tiles. RTM fine, SP1 blurries.In the original thread you asked me to test with stock FS10 textures under certain conditions. I posted the results of the tests and as you know I didn't have the issue with stock FS10 format textures.It seems to me that when many small files need to be loaded SP1/2 gets itself into a bit of a "state" and can't load textures any more. RTM doesn't have these issues.Why don't you try and load the "FlightAlpes Nord" scene from FranceVFR (www.francevfr.com) and see what happens. This obviously pushes the FSX engine to it's limits but most of us that have been affected by the blurries get the repros either with this type of scenery or with TileProxy that uses the same format for the scenery textures.If I can be of further assistance please send me a personal message via avsim.Kind regards,JC

[sightseer 895]

SolarEagle - get as much RAM as you can. Personally, I would wait a little longer to see what else is new in the hardware department.JC - Do you run with autogen? how much?I have come to a semi-uneducated conclusion that the problems originated with SP1 because that is when autogen batching was started.Those who have systems which could handle the shift of the burden of work never noticed any decrease. In fact they got an increase in performance.Those of us with systems closer to the margin line did notice the shift and we had to do something about it.I'd be willing to bet that people either used lower resolution textures or little or no autogen or the smallpartrejectradius tweak.(or some combo).I recently lowered the resolution of my FEX textures and improved my situation. Its not as good as it was pre SP2 though.I think its the autogen batching that is causing this 'problem' (in quotes because it only seems to affect those of us with systems that are less capable in some way).I would really, really, really like to see a way to force the smallpartrejectradius switch to work. Otherwise I need to lower my autogen settings just to get what I had before SP2. Note that I havent tried lowering my autogen settings yet becuase I dont want to.I would rather just uninstall Acceleration.I shouldnt have to lower my settings to compensate for an 'update which improves performance'.

[Guest jcmckeown]

I do not have the autogen packs for francevfr scenery. So there is no autogen (which is good for frame rates, particularly in RTM). I do not have autogen turned off in the slider though. There is just none in the scenery. Sorry to break the theory...JC

[PhilTaylor 45]

The discussion we are having is referring to the "blurries all the time" condition and not the "when using photo-scenery blurries". Trying to discuss and debug all the different types at once just doesnt work, separation of the parts is valuable here.

[Mace 1,928]

>>Anybody have any other suggestions to help me aviod the>illusive "issue with your config"?Yes.Did you do any research about chipsets? X38, P35, 680i.If it were me I would be doing homework on these chipset motherboards and what kind of performance their users get in the sim, with particular attn being paid to blurries or lack thereof.RhettAMD 3700+ (@2585 mhz), eVGA 7800GT 256 (Guru3D 93.71), ASUS A8N-E, PC Power 510 SLI, 2gb Corsair XMS 3-3-3-8 (1T), WD 150 gig 10000rpm Raptor, WD 250gig 7200rpm SATA2, Seagate 120gb 5400 rpm external HD, CoolerMaster Praetorian

[PhilTaylor 45]

Sorry about posting these as txt.Here are the jpgshttp://forums.avsim.net/user_files/179958.jpghttp://forums.avsim.net/user_files/179959.jpghttp://forums.avsim.net/user_files/179960.jpghttp://forums.avsim.net/user_files/179961.jpg

[Guest]

>Did you do any research about chipsets? X38, P35, 680i.>>If it were me I would be doing homework on these chipset>motherboards and what kind of performance their users get in>the sim, with particular attn being paid to blurries or lack>thereof.>>Rhett>I have been though all the blurry threads, and I have not picked up any discernable pattern between hardware and blurries. The hardware is not my concern, as the blurries have to be software related, otherwise they would not be cured by uninstalling SP1. Besides, 680i performs as good as or better than anything out there.

[PhilTaylor 45]

which blurries do you mean?for "blurries when using photo-scenery" that is one fix. and the changes in SP1 did have an affect there. note another fix for that type of blurries would be to re-author the photo-scenery using the advice in the Acceleration/SP2 SDK.for other types of blurries, it isnt that simple.

[Mace 1,928]

You know I was thinking about your system build and it occurred to me, that you might want to emulate GSalden and/or GHN's systems when you build yours. By that, I mean those two guys are getting really good results, even with DX10, with their setups. It might pay off to find out exactly what components they used and what sw they are running, etc.RhettAMD 3700+ (@2585 mhz), eVGA 7800GT 256 (Guru3D 93.71), ASUS A8N-E, PC Power 510 SLI, 2gb Corsair XMS 3-3-3-8 (1T), WD 150 gig 10000rpm Raptor, WD 250gig 7200rpm SATA2, Seagate 120gb 5400 rpm external HD, CoolerMaster Praetorian

[Guest Dave65]

Hi guys,I haven't had a chance to switch monitors with my friend,but I did hook up my old 19 inch Samsung CRT monitor,and my blurries got at least 50 percent better,above 5000 ft It seems crisp with no blurries,im not sure what to make of this though..