Stutter Experiment - Can I ask you to join in?

[Guest babua]

but thats what i have, frame rate @20 and Texture b/w @40.

[brucets 94]

Mike -I tried the test and I found that in *locked spot view* and with all views zoomed out to 70% I get very minor stuttering and no surging in any view. I think the key for me is to make sure spot view is locked. My system is dedicated to FS only (not online):P4-3.46 Ghz extremeRadeon ATI X850XT pcie Platinum editon2Gb ram2 WDR Raptor 10,000 rpm HDs (one for just FSX)I have only a few tweaks in place. Here's my settings(which do not eliminate blurred ground textures BTW):Mike's Test FSX.config settings:[sOUND]AmbientUIMusicVolume=-6.000000SOUND_FADER1=0.340000SOUND_FADER2=0.500000SOUND_FADER3=0.500000SOUND_FADER4=0.500000AmbientUIMusic=FSX01SOUND=1SOUND_QUALITY=2SOUND_LOD=0UISound=1AmbientUI=0[Display]ChangeTime=4.000000TransitionTime=4.000000ActiveWindowTitleTextColor=255,255,255ActiveWindowTitleBackGroundColor=0,28,140,64NonActiveWindowTitleTextColor=255,255,255NonActiveWindowTitleBackGroundColor=24,33,87,64InfoUpperRightTextColor=255,0,0InfoUpperRightBackGroundColor=0,0,0,0InfoLowerLeftTextColor=255,255,255InfoLowerLeftBackGroundColor=255,0,0,128InfoLowerRightTextColor=255,255,255InfoLowerRightBackGroundColor=255,0,0,128InfoBrakesEnable=TrueInfoParkingBrakesEnable=TrueInfoPauseEnable=TrueInfoSlewEnable=TrueInfoStallEnable=TrueInfoOverspeedEnable=TrueBLOOM_EFFECTS=0SKINNED_ANIMATIONS=0TEXTURE_BANDWIDTH_MULT=40UPPER_FRAMERATE_LIMIT=0WideViewAspect=False[Main]User Objects=Airplane, HelicopterSimObjectPaths.0=SimObjectsAirplanesSimObjectPaths.1=SimObjectsRotorcraftSimObjectPaths.2=SimObjectsGroundVehiclesSimObjectPaths.3=SimObjectsBoatsSimObjectPaths.4=SimObjectsAnimalsSimObjectPaths.5=SimObjectsMiscProcSpeed=4564PerfBucket=5Maximized=1Location=284,190,1316,964,.DISPLAY1HideMenuNormal=0HideMenuFullscreen=1FIBER_FRAME_TIME FRACTION=0.33[PANELS]IMAGE_QUALITY=1DEFAULT_VIEW=0QUICKTIPS=0PANEL_OPACITY=100PANEL_MASKING=1PANEL_STRETCHING=1UNITS_OF_MEASURE=0[Weather]WindshieldPrecipitationEffects=1MinGustTime=10MaxGustTime=500MinGustRampSpeed=1MaxGustRampSpeed=200MinVarTime=5MaxVarTime=50MinVarRampSpeed=10MaxVarRampSpeed=75TurbulenceScale=1.000000WeatherServerAddress=fs2k.zone.comWeatherServerPort=80WeatherGraphDataInDialog=0AdjustForMagVarInDialog=1DynamicWeather=2DownloadWindsAloft=0DisableTurbulence=0CLOUD_DRAW_DISTANCE=3DETAILED_CLOUDS=1CLOUD_COVERAGE_DENSITY=5THERMAL_VISUALS=0[DISPLAY.Device.RADEON X850 Series.0]Mode=1360x1024x32Anisotropic=1[CONTROLS]Controls_Default=StandardControls_Current=StandardKBDAIL=64KBDELEV=64KBDRUD=64[TextInfo.1]Latitude=1,1Longitude=1,2Altitude=1,3Heading=1,4AirSpeed=1,5WindDirectionAndSpeed=1,6[TextInfo.2]FrameRate=1,1LockedFrameRate=1,2GForce=1,3FuelPercentage=1,4[TextInfo.3]Latitude=1,1Longitude=1,2Altitude=1,3Heading=1,4AirSpeed=1,5WindDirectionAndSpeed=1,6FrameRate=2,1LockedFrameRate=2,2GForce=2,3FuelPercentage=2,4[slewTextInfo.1]Latitude=1,1Longitude=1,2Altitude=1,3Heading=1,4AirSpeed=1,5[slewTextInfo.2]FrameRate=1,1LockedFrameRate=1,2[slewTextInfo.3]Latitude=1,1Longitude=1,2Altitude=1,3Heading=1,4AirSpeed=1,5FrameRate=2,1LockedFrameRate=2,2[DynamicHeadMovement]LonAccelOnHeadLon=-0.020000LonAccelOnHeadPitch=-0.010000RollAccelOnHeadLat=0.010000YawAccelOnHeadLat=-0.100000RollAccelOnHeadRoll=0.100000MaxHeadAngle=5.000000MaxHeadOffset=0.300000HeadMoveTimeConstant=1.000000[VirtualCopilot]VirtualCopilotActive=0[GRAPHICS]SHADER_CACHE_PRIMED=1693428480TEXTURE_MAX_LOAD=1024NUM_LIGHTS=8AIRCRAFT_SHADOWS=1AIRCRAFT_REFLECTIONS=1COCKPIT_HIGH_LOD=1LANDING_LIGHTS=1AC_SELF_SHADOW=0EFFECTS_QUALITY=0GROUND_SHADOWS=0TEXTURE_QUALITY=3IMAGE_QUALITY=0See_Self=1Text_Scroll=1[uSERINTERFACE]ShowAllACPaintSchemes=1SelectAircraftManufacturer=AllSelectAircraftPublisher=AllSelectAircraftType=AllOpenATCOnCreate=0DisplayFuelAsWeight=1PageID=1SHOW_MISSION_CAPTIONS=0PAUSE_ON_LOST_FOCUS=1PROMPT_ON_EXIT=1SITUATION=C:Documents and SettingsBruce ScottMy DocumentsFlight Simulator X FilesCO 738 DEN GJT IFR Push BackMap_Orientation=2[FlightPlanMap]LineWidth=2SHOW_AIRPORTS=1SHOW_VORS=1SHOW_NDBS=1SHOW_APPROACHES=1SHOW_INTERSECTIONS=1SHOW_VICTOR=0SHOW_JET=0SHOW_AIRSPACE=1SHOW_FLIGHTPLAN=1SHOW_WEATHERSTATIONS=1SHOW_WEATHERSYSTEMS=1SHOW_DATATAGS=1SHOW_TERRAIN=1show_waypoints=1show_airways=1show_markers=1show_volume_boundaries=1show_ac_twr=1[MAPVIEW_MAP]SHOW_AIRPORTS=1SHOW_VORS=1SHOW_NDBS=1SHOW_APPROACHES=1SHOW_INTERSECTIONS=0SHOW_VICTOR=0SHOW_JET=0SHOW_AIRSPACE=1SHOW_FLIGHTPLAN=1SHOW_WEATHERSTATIONS=1SHOW_WEATHERSYSTEMS=1SHOW_DATATAGS=1SHOW_TERRAIN=1show_flight_history=1[ATC]ShowATCText=1COMM_MSG_NONE_COLOR=FFFFFFFFCOMM_MSG_ATC_USER_COLOR=FFB6FFB6COMM_MSG_USER_ATC_COLOR=FFFFD21BCOMM_MSG_ATC_AI_COLOR=FF00FF00COMM_MSG_AI_ATC_COLOR=FFFF7840AutoOpenAirTrafficWindow=1UsePilotVoice=1PilotVoice=0[PointOfInterestSystem]CycleSetting=0[sCENERY]LENSFLARE=0DAWN_DUSK_SMOOTHING=1IMAGE_COMPLEXITY=4[TrafficManager]AirlineDensity=30GADensity=0FreewayDensity=0ShipsAndFerriesDensity=0LeisureBoatsDensity=2IFROnly=0AIRPORT_SCENERY_DENSITY=2[TERRAIN]LOD_RADIUS=4.500000MESH_COMPLEXITY=70MESH_RESOLUTION=20TEXTURE_RESOLUTION=24AUTOGEN_DENSITY=0DETAIL_TEXTURE=1WATER_EFFECTS=2[AContain]ShowLabels=0ShowUserLabel=0ShowLabelManufacturer=1ShowLabelModel=1ShowLabelTailNumber=0ShowLabelDistance=1ShowLabelAltitude=1ShowLabelAirline=0ShowLabelAirlineAndFlightNumber=0ShowLabelFlightPlan=0ShowLabelContainerId=0ShowLabelAirspeed=0ShowLabelHeading=0LabelDelay=1000LabelColor=FFFF0000[iNTERNATIONAL]ASLAT=2ASLON=1MEASURE=0[REALISM]PFactor=1.000000Torque=1.000000GyroEffect=1.000000CrashTolerance=1.000000General=1.000000UnlimitedFuel=FalseTrueAirspeed=FalseAutoCoord=FalseRealMixture=FalseStressDamage=TrueGEffect=TrueManualLights=TrueGyroDrift=TrueCrashWithDyn=FalseCrashDetection=TrueAutoTrim=False[sIM]SYSCLOCK=1[sTARTUP]DEMO=0SHOW_OPENING_SCREEN=1STARTUP_DEMO=LoadWindow=1[FACILITIES]COUNTRY=United StatesSTATE=CaliforniaCITY=Los AngelesGTL_BUTTON=1501[Misc]Com_Rate=7[JOBSCHEDULER]affinity mask=0[bufferPools]poolSize=5000000Thanks for the test opportunity!B. Scott

[Cruachan 705]

Hi,Thanks very much for your input. Here are my observations and suggestions:"I have only a few tweaks in place. Here's my settings(which do not eliminate blurred ground textures BTW)"TEXTURE_RESOLUTION=24 <--- have you tried upping this to 25 (makes a significant difference on my setup - I see blurries at 24)?affinity mask=0 <--- I note you have a single core CPU. This may be the key and it's possible single core users are seeing less stuttering/surging than multiple core users. As we know SP1 saw the introduction of significant changes/sharing of core use in multiple cored CPUs.LOD_RADIUS=4.500000 <--- have you tried reducing this value to 3.5 (medium) and reducing your visibility distance in the sim. To my mind this gives a much more realistic representation of the real world, although I do admit it doesn't look quite so 'pretty'. Should give you a few more fps, however.FIBER_FRAME_TIME FRACTION=0.33 <-- Can you try the effect of commenting this out (place a semicolon in front of the line) and re-run the test? I believe the value 0.33 is the default value so this may not be having any influence on your performance unless you change it. poolSize=5000000 <--- can you comment on whether you see any performance benefit by including this line? Since SP1 (and indeed before) I could not convince myself that it was helping much if at all. Again, try commenting it out and run the test again.UPPER_FRAMERATE_LIMIT=0 <--- does the sim stutter if you peg the target frame rate to, say, 20 or 25? I can't run at unlimited as the frame rate fluctuations are quite large making the sim run badly.I note you have FSX installed on a 10,000rpm Raptor. I'm sure this must help to reduce any pausing (stuttering?) as new textures are being loaded on-the-fly. Just how much benefit as compared to, say, a 7,200rpm SATA II well defragmented drive is perhaps open to debate, but it certainly can't hurt.Thanks again for taking the trouble.Now, where are all those multiple core stutter-free simmers?!Mike

[Cruachan 705]

Hi Martin,Thanks for contributing to this topic."You SHOULD see slight stutters on any system doing 360 degree turns. FS9 had the same issue"I agree. The difference, however, between the two sims is that the stuttering seems to persist in FSX even after several 360's over the same ground area.You may be correct about the CPU bottleneck/bus issue. I wonder whether anyone else is able to comment on this?Mike

[Mace 1,928]

>>I agree. The difference, however, between the two sims is that>the stuttering seems to persist in FSX even after several>360's over the same ground area.Mike I am not seeing that at all on my setup...in very limited testing (1 long flight, 2 or 3 short hops). For me to say I have no stutters would be premature at best. I did notice one on replay once, but that's nothing new.I will grab your flight and see if I can replicate.I will not be changing my fsx.cfg, however.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

[oktorn777 0]

HI Mike,I'm running a c2d 4300@2.7Ghz, FSB1200, 2gigs Corsair DDR2-800, radeon x1900 Pro 512mb und am getting some stutters when panning the view or making sharp turns. I build this setup just a few weeks ago and also replaced my old trusty sony 19' CRT with a Asus LCD.I don't have a lot of experience with lcd's and gaming, but for now I get the least stutters when locking the framerate to 30 (half the refresh rate of the display).I also hoped for stutter free sim after sp1, because stutters kill the immersion for me, no matter how small and short they are.Looking forward to any solutions,cheers,:-beerchugC.G.

[Cruachan 705]

Hi C.G."I get the least stutters when locking the framerate to 30"Try reducing your Target Frame Rate to 25 or even 20 and then try the stutter test using my provided flight situation (see download link to the zip file in my original post). Please let us know how you get on. Is your HD well defragmented? If you are relying on the Windows Defragmenter (I am assuming you are on Windows XP) it's a good idea to defrag 3x in succession. The first will take the longest, but the following 2 should complete relatively quickly. I have no experience of the Vista defragmenter but I imagine it is similar. This may reduce the texture loading time and slight hesitations noted while panning.Mike

[Cruachan 705]

Hi Rhett,Glad to see you are joining in :)"I will grab your flight and see if I can replicate."Yes, I would appreciate it. If we all use the same flight situation to test then we should get a better idea who are the real 'winners' and 'losers'. Please, if you're experience is favourable, it would be helpful if you would attach your current FSX.cfg to your next post. We do need to see what settings each of us is using. Best to attach the complete file rather than just the selected tweaks."I will not be changing my fsx.cfg, however."Perfectly okay. We can see how your sim is set up in relation to your hardware's capabilities from your FSX.cfg.Mike

[Guest abulaafia]

I can contribute RE Bufferpools: which I have experimented with ad nauseam: I see absolutely no improvement when the parameter is added and increased stutters with values over 2000000. My sim runs smoothest without it added to the fsx.cfg. Apparently some cards manage cache better without any interfering. ==Marten:Weber==P-D920 oc3.64MHz 2GB RAM WinXP ATI X1900+1024MB 2xSATA250 2xVIEWSONIC 19"Best AI available today: www.world-of-ai.com

[Guest abulaafia]

Try 10/20 and check if there is any difference. Also your CPU+bus speed would be helpful. There are MANY PCs out there with cheaper motherboards (LE, SE versions etc.) which are overburdened with the new double-core CPUs. I have regular business contacts with motherboard manufacturer Asus and they tell me that board bottlenecks are a big headache for them since Intel released the duals. Load testing of MB with multiple cores working has only become standard in the last few weeks! Only in September will we see new MBs which can really handle two or four cores working full speed. Dell is currently selling a lot of OEM stuff that doesn't match.When you have a MB with a bottleneck, it doesn't matter would your processor SHOULD be able to do , but only what the bottleneck can deliver. In the fsx.cfg you find an entry called ProcSpeed: SP1 calculates the data load based on the processor speed. You can try set it much lower to see if it changes anything. Finally, try setting the processor/bus to different speeds. Overclocking often has different effects. On my machine I get massive stutters when OC 10 or 20 percent, but none when OC to 30%. RAM speed is also an issue. Cheap or defect RAM chips sometimes create bottlenecks. There is software out there to test your entire PC and locate bottlenecks, try www.download.com. ==Marten:Weber==P-D920 oc3.64MHz 2GB RAM WinXP ATI X1900+1024MB 2xSATA250 2xVIEWSONIC 19"Best AI available today: www.world-of-ai.com

[Cruachan 705]

Hi Marten,This is interesting stuff - thanks!"On my machine I get massive stutters when OC 10 or 20 percent, but none when OC to 30%."I confess I can't get my head around this seemingly paradoxical observation. If the MoBo bus is the bottleneck then why are you not seeing increasing problems at higher CPU clocks?"In the fsx.cfg you find an entry called ProcSpeed: SP1 calculates the data load based on the processor speed. You can try set it much lower to see if it changes anything."Yes, I wondered about that as well. You may recall from an earlier post of mine in this thread, before I realised that the "Cool 'n' Quiet" technology was interfering with the correct detection of my CPU speed, that I was seeing much lower values for ProcSpeed and PerfBucket. These values (1885 and 2 respectively) produced very low initial settings in the sim. The image quality was dreadful but, on the plus side, stutters and 'surging' were both absent. I found I was able to gradually increase the in-sim settings to a certain point when the stutters started to reappear. So, why am I now running with the (?)optimum values (4629 and 6)? Simply because at the moment I am assuming (possibly incorrectly) that the problem lies elsewhere. To be fair, once I had established 'optimum' settings with minimal stuttering, I did try experimenting by changing these values in the hope that lower figures would tip the scales in favour of a stutter-free experience, but it didn't seem to have any noticeable effect.I may be wrong, but I have a feeling that the ProcSpeed and PerfBucket are only used during the initial creation of FSX.cfg and produce the base in-sim settings as determined by your system's capabilities. Thereafter, these values seem to be being ignored:4629 and 6http://forums.avsim.net/user_files/172909.jpg2315 and 3http://forums.avsim.net/user_files/172910.jpg1885 and 2http://forums.avsim.net/user_files/172911.jpgTry it yourself.Mike

[fbass 1]

Mike,You indicated that you "turned off" AVG. Can you define "turned off?" The reason I ask this question is because I also have AVG and I HAD these lurches you described. I disabled the processes using the Windows function and I thought I really had turned both AVG processes off. However, when I ran EndItAll, both processes were still there. When I "killed" both of these processes my "lurching" completely went away.What I do now is run EndItAll, kill the two AVG processes, run FSX, quit FSX, and reinstate AVG.Frank

[Cruachan 705]

Hi Frank,My normal practice prior to simming and gaming, software installation or defragging (? whether really necessary with defragging) is to right click the AVG system tray icon and select "Quit AVG Control Center". Avgcc.exe then closes and is removed from the Processes list in Task Manager.I don't use EndItAll these days but checked it just now and I see that a couple of AVG related items are locked, viz. avgamsvr.exe (Alert Manager) and avgupsvc.exe (Update Service). The 2 items that can be closed/stopped are avgcc.exe (Control Center) and avgemc.exe (E-Mail Scanner).When I watch these 4 Processes in Task Manager (sort the list so they appear together at the top), without quitting AVG Control Center, the only activity I see in the CPU column is avgcc.exe. It was very brief, showed 1% and occurred while loading a flight. I watched for a few minutes longer and nothing occurred. That's not to say nothing would occur over a longer period, of course.My feeling is that I only need to quit the AVG Control Center. Nevertheless your experience is interesting and should certainly be noted.Mike

[Cruachan 705]

Here's something else for everyone to try:I came across this posting by 'Cavokk':http://forums.avsim.net/dcboard.php?az=sho...id=401630&page=You will note he/she is noticing a difference in performance at relatively high slider settings for Target Frame Rate.I am a little sceptical about this one. At first, I thought I was seeing a much more smooth experience using my flight situation in Stutter_Test.zip, but subsequent tests left me feeling less than certain.It would be interesting to explore this further. I would suggest starting at 'Unlimited' and working back in, say, 5fps decrements. If this works then it may be we will all have different 'sweet' spots. Worth a try.While testing I did note that my GPU temp went up from 62 degreesC at 20fps to 67-68 degreesC at 75-Unlimited fps with a corresponding increase in fan noise, so it's working much harder. My actual frame rate varied between 20-40+ at the higher settings. My CPU temp didn't budge from 50.50 - 51.00 degreesC at either setting.Mike

[Carabou 0]

Hi Mike, First let me thank you for your efforts to establish the test an so on. I tried your test flights as well as your FSX.cfg.I get stutters with your cfg and mine in spot and cockpit especially when I started to view the airport. I reduced autogen to zero and it improved a bit but still not great.Another thought: I've had Friday Harbor as a default flight for some time (before SP1 was installed) and it's important to note that I had the stutters and minor jerkiness when SP1 wasn't installed! Therefore, the problem may not be entirely due to SP1. It would help if someone who doesn't have SP1 installed trys your test to see what results they have with the original FSX. Hope this helps. Once again, thanks for your efforts,