Ground textures gets more blurry the further I fly

[AnkH 2,266]

28 minutes ago, gboz said:

Why are we limiting in NI.?

Short answer: because "we" obviously do not understand the difference between FPS and monitor refresh rate (Hz)

[gboz 157]

16 hours ago, SteveW said:

No.

 

Here's a graph of around 20fps so it shows up better:

 

As you can see pushing an extra frame or two simply induces stutter and takes up process time that is wasted.

Steve, no offence, but why are you showing FSX graphs in a P3D thread?
I think FSX and P3D have grown far enough apart by now that projecting FSX performance onto P3D is not necessarily valid, especially in this case where you use the fractional VSync NI setting that does not work with P3D anyway. P3D for me is smoother at TFR 31fps compared to TFR 30fps (internal), App Vsync and 30HZ.

gb.

[Adamski_NZ 952]

11 minutes ago, gboz said:

P3D for me is smoother at TFR 31fps compared to TFR 30fps (internal), App Vsync and 30HZ.

How/where are you setting 30Hz (genuine question)?

Adam.

[AnkH 2,266]

Genuine question, yes. Especially for one like me, using a 144Hz monitor with G-Sync not really capable of 30Hz. But, this discussion is going on for months now and no one ever could tell me how I should adapt this Hz-logic to my monitor. For the moment, I also have my FPS limited to 30, I leave all monitor settings as they are (just forced G-Sync to off for P3D) and I am pretty happy. Could be a notch smoother, but then...

[Tomaz Drnovsek 440]

Against my better judgment I just made a switch from v3 to v4. I just finished installing all my addons and I didn't made any full flight yet. Than I see this thread and now I'm worried. With my GTX970 there's really no way I can fly without TFR setting to Unlimited. Is this a general issue and everybody experience it with mentioned TFR setting? How quickly do ground textures become blurry and autogen disappears?

I really hope that after spending 400€ for all my addon upgrades I didn't just go from nicely working v3 to an unflyable v4.

[SteveW 1,792]

3 hours ago, gboz said:

Steve, no offence, but why are you showing FSX graphs in a P3D thread?
I think FSX and P3D have grown far enough apart by now that projecting FSX performance onto P3D is not necessarily valid, especially in this case where you use the fractional VSync NI setting that does not work with P3D anyway. P3D for me is smoother at TFR 31fps compared to TFR 30fps (internal), App Vsync and 30HZ.

gb.

I'll get around to posting some graphs for the naysayers one day. The top line shows a fixed fps of 21 against 20 - on a 60hz monitor - it's a longer line than at 20 and 20 is also 5% less throughput than 21 - that's the point and it's the same with P3D and any D3D apps.

You know it makes sense.

[Adamski_NZ 952]

1 hour ago, 0Artur0 said:

I really hope that after spending 400€ for all my addon upgrades I didn't just go from nicely working v3 to an unflyable v4.

You'll be fine. I have a GTX970 too. What CPU do you have (mine's an i7 @ 3.40GHz)?

Actual frame rates will depend (as ever) on your autogen sliders and detail settings - but setting the TFR to something sensible will definitely cure the "disappearing autogen" problem - without too much of an impact on your overall frame rate. Believe me - I just switched from "unlimited" myself and it *definitely* helps. My scenery stays wonderfully updated all the time - from the beginning of a flight right to the very end - regardless of my detail and weather (clouds) settings.

BTW: If people here would like to create FPS graphs, I've [only] just discovered the free/donationware SimLauncherX (by Maarten Boelens) to work well. It's an excellent piece of programming all round. After only using it for an hour, I already feel moved to sending him a donation.

Edited August 13, 2018 by Adamski_NZ

[Tomaz Drnovsek 440]

1 minute ago, Adamski_NZ said:

You'll be fine. I have a GTX970 too. What CPU do you have (mine's an i7 @ 3.40GHz)?

Actual frame rates will depend (as ever) on your autogen sliders and detail settings - but setting the TFR to something sensible will definitely cure the "disappearing autogen" problem - without too much of an impact on your overall frame rate. Blieve me - I just switched from "unlimited" myself and it definitely helps. My scenery stays wonderfully updated all the time - from the beginning of a flight right to the very end - regardless of my detail and weather (clouds) settings.

I have I7 4790k@4.5GHz.

I'm afraid limiting TFR is not an option.. I loose ~10 FPS by doing that and my TrackIR head movememnts are extremely choppy and stuttery when frames are limited. And by loosing 10FPS on payware airport with payware aircraft my frames go from 25FPS to 15FPS. I don't consider 15FPS flyable.

I really hope LM won't just ignore the issue or say that that's what's to be expected because in v3 this issue did not exist and by reading the posts here people didn't experience it before v4.2.

[Adamski_NZ 952]

14 minutes ago, 0Artur0 said:

I'm afraid limiting TFR is not an option.. I loose ~10 FPS by doing that and my TrackIR head movements are extremely choppy and stuttery when frames are limited.

Oh dear! I don't run anything like TrackIR, so poor frame rates (I get around 15!!!!) don't actually interfere to much with my eye-candy-rich simming style..

I understand that this problem may have actually got worse in P3Dv4 (and LM should fix it), but is reducing some of your autogen settings an option at all? I do agree, though ... you shouldn't *have to* do that.

[Tomaz Drnovsek 440]

2 minutes ago, Adamski_NZ said:

I understand that this problem may have actually got worse in P3Dv4 (and LM should fix it), but is reducing some of your autogen settings an option at all? I do agree, though ... you shouldn't *have to* do that.

Well, I guess I'll have to reduce it if that's the case and if this will prevent the problem to occur with TFR set to Unlimited.

[SteveW 1,792]

4 hours ago, gboz said:

Steve, no offence, but why are you showing FSX graphs in a P3D thread?
I think FSX and P3D have grown far enough apart by now that projecting FSX performance onto P3D is not necessarily valid, especially in this case where you use the fractional VSync NI setting that does not work with P3D anyway. P3D for me is smoother at TFR 31fps compared to TFR 30fps (internal), App Vsync and 30HZ.

gb.

Run a coupe of tests at 20 and 21, 20 has the shortest line each time. Not surprising. At 21 the fps is being pulled between 21 and 20, whereas at 20 there's a constant line. I want to arrange a better test to show up the fact but can't at the moment.

I suspect your 31 is simply because your monitor is running slightly faster than my 57-59Hz and so on mine 29 is better - on a slightly faster monitor 62-63Hz then 30-31 is better.

 

It is still the same point I made before - do not push the extra frame or it is simply lost processing and heat gained.

 

[SteveW 1,792]

...so as you can tell it's a test YOU have to do - (not me) 🤣

[grof 0]

10 minutes ago, SteveW said:

Run a coupe of tests at 20 and 21, 20 has the shortest line each time. Not surprising. At 21 the fps is being pulled between 21 and 20, whereas at 20 there's a constant line. I want to arrange a better test to show up the fact but can't at the moment

...

ot push the extra frame or it is simply lost processing and heat gained.

 

Hi Steve,

Just one question. Why should we use 20/30 tfr on a 60 Hz monitor? It has to be half/third of the monitor refresh rate?

My monitor set to 75Hz. Then I should set TFR to 24?

Thanks

 

Zoltan

Edited August 13, 2018 by grof

[SteveW 1,792]

17 minutes ago, grof said:

Hi Steve,

Just one question. Why should you use 20/30 tfr on a 60 Hz monitor? It has to be half/third of the monitor refresh rate?

My monitor set to 75Hz. Then I should set TFR to 24?

Thanks

 

Zoltan

Simply because someone has better performance at 31 or whatever is not helping - why - because each monitor and cable and GPU setup is different. The exact time is important - we might say 20 or 30 but that's just for simplicity.

I showed a graph of 20 because it's magnified compared with data from 30 or 60 - that's all there is to it. Also only a few frames are needed since with too many the data becomes too crammed to inspect.

Choosing 10,15,20,30,60 all coincide more often with 60 or 120 or 240 Hz Monitors.

In tests it can be seen quite clearly that the GPU is pulled between any closest multiple and the set output frequency of the GPU/Game. And so as I keep saying your closest match or below is better than the closest match or one above - it's that simple.

So with a 75 Hz monitor 25,50,75 are all good - so yes you should check out "24". if you understand it that might be 25 as apposed to 26.

 

Look in the specifications of the monitor a 75Hz can be anything between 70 and 80 or worse.

Guys, don't come posting stuff like "Well - mines better at N fps", that's not helpful in this context - each system has to be adjusted to suit and put simply an extra frame is a big mistake.

 

Edited August 13, 2018 by SteveW

[SteveW 1,792]

...so recapping:

Adam said he noticed long frames (stutter) at 31fps.

So I suggested try 29. He can try 30 and even 32 if I am understood correctly - it depends ultimately on the timing of his monitor and setup.

 

My point was to go lower rather than higher because of a greater success rate:

a) 1fps less at 31 is 3% extra throughput or 3% less heat - there's tools around that people pay for to yield 3 or 4%.

b) The system does not store frames in hand - they are lost.

 

Edited August 13, 2018 by SteveW