Can you see the difference between 30fps and 60fps?

[RJC68 823]

Just now, sonny147 said:

oh it does thanks a lot RJC

Sorry I forgot to say if you use the overlay functions in RTSS there is a nice FTV graph that shows the timing in milliseconds between each frame. The less jumping around in fps the smoother the sim will be. Couple that with the higher fps that FG produces it really is something to see.

[sonny147 586]

48 minutes ago, bobcat999 said:

Lucky B!    Which one?  55 inch?

FTV = frame time variance, or roughly speaking the amount of time it takes for your PC to render a complete frame, which can vary depending on complexity of course.
Some think the holy grail is to get them paced out equally to get the best in smoothness even if you have lower frame rates.   A bit obvious really, but not that important if you can just power through it.

So I am not too concerned with it.  While it is a useful measure for some struggling for smoothness, if you want to be locked at say 60 fps, and you have the headroom to hit 75 / 80 fps without the 60 lock or vsync, it matters not one iota. 

It all goes out the window at a complex airport with high traffic anyway!  

This one

 

Edited April 3 by sonny147

[sonny147 586]

1 minute ago, RJC68 said:

Sorry I forgot to say if you use the overlay functions in RTSS there is a nice FTV graph that shows the timing in milliseconds between each frame. The less jumping around in fps the smoother the sim will be. Couple that with the higher fps that FG produces it really is something to see.

lets see tomorrow brings when i install the new tv

[sonny147 586]

43 minutes ago, sonny147 said:

This one

50TM8B 

was going to get the 55 but i use a 43" now and think the extra 7"😉 is the sweet spot for me 

Edited April 3 by sonny147
removed link

[cobalt 801]

I stopped worrying about FPS a long time ago. Most of the time I experience smooth-as-silk performance with just an average computer and video card. It's interesting that this topic rears its head periodically despite continuing advances in hardware and software that one would think would have rendered this subject moot by now. Yet I have no doubt that some day soon we will see arguments about whether the difference between 150 and 170 FPS is detectable. Count on it. 

[turbomax 3,307]

50 minutes ago, sonny147 said:

what are they on about FTV ?,

https://www.tomshardware.com/reviews/gaming-processor-frame-rate-performance,3427-2.html

micro-stuttering occurs when the amount of time that passes between frames appearing on-screen is inconsistent, resulting in what appears to be choppy gameplay, even in spite of high average frame rates. For example, two different PCs, both generating an average of 24 frames per second, may convey different experiences, one stuttery and one smooth, if the amount of time between each frame is less regular on one and more regular on the other.

In the chart below, System A sees a consistent amount of time between those 24 frames, while System B does not. Therefore, you may notice a stuttering effect on System B, even though both machines still average 24 FPS.

 

https://techreport.com/review/inside-the-second-a-new-look-at-game-benchmarking/

Edited April 3 by turbomax

[sonny147 586]

1 minute ago, cobalt said:

I stopped worrying about FPS a long time ago. Most of the time I experience smooth-as-silk performance with just an average computer and video card. It's interesting that this topic rears its head periodically despite continuing advances in hardware and software that one would think would have rendered this subject moot by now. Yet I have no doubt that some day soon we will see arguments about whether the difference between 150 and 170 FPS is detectable. Count on it. 

yes Cobalt i am the same as you, iam  getting new tv for smoothness not FPS , i have a 4090 and a 4k 60 hz is holding it back , my hope is for better smoothness with the VRR etc  malarky !

[RJC68 823]

34 minutes ago, sonny147 said:

lets see tomorrow brings when i install the new tv

Here is a perfect example, I captured this 15 minute segment with CapFrameX from the threshold of EGGP in the Fenix with a good amount of FSLTL traffic and ultra settings with 6K shadows and a TLOD of 450 in the air.

Now Do I expect to see that somewhere like InI Builds Heathrow, of course not, I have the headroom where by I can drop my fps and GSYNC on my Sony X90K takes care of things but it never gets out of control.

Hmm I should actually try this test with the excellent AutoFPS at a larger hub and see how it performs.

Edited April 3 by RJC68

[s0cks 135]

Some people in this thread have such high standards that I can see now why they spend so much time tweaking and buying expensive hardware. Each to their own of course. My FTV is terrible at times, but 90% of my flight is smooth so I just shrug past the occasional stutters and get on with my flight. I am excited to see what MSFS2024 brings to the table though in terms of performance. It will be so nice if we can get the sim using some more CPU cores. It is kinda ludicrous that we buy these expensive multi-core CPUs and MSFS is still running primarily on a single thread, lol.

Edited April 3 by s0cks

[Cpt_Piett 4,830]

2 hours ago, Noel said:

It's what I've stated all along.

I just changed my opinion based on new input, including yours in this thread (re: FTV in relation to FPS), and admitted that my previous statement was inaccurate. I have no problem doing that. 

2 hours ago, Noel said:

Which, "on this"?  I think the "artifacts" described *should* resolve w/ the new CPU.  What I'm dobutful about is 2-screen problem.  Near as I can tell no way that will resolve.   And as I've said if I can maintain a rate of 45 or greater, w/o FG, I'll be more than happy w/ the 7800X3D.  With perfect FTV and a rate of 45, there's not a whole lot of need to increase rate.   I'm certain I'll easily hit 45 in any plane anywhere.

There are also other sources of lack of perfection and I'm never sure what to attribute those to.  I can have major CPU main thread, GPU and VRAM headroom and yet get the odd stutter here and there.   And this varies greatly by flight.  Sometimes never a single stutter, other times some occasional ones.  Hopefully the better CPU reduces these but TBH they're already minimal.

Re: the benefit of running at 60FPS vs 30FPS. I'm sure you'll have no problem maintaining a constant 70FPS, even at major hubs, even with a near flat frametime graph. Although probably not with AI traffic.

Anyway, I look forward to hearing about your experience with the 4090+7800X3D combo (probably the best that exists at the moment) and sincerely hope that you'll be happy with your new rig. 

12 minutes ago, s0cks said:

Some people in this thread have such high standards that I can see now why they spend so much time tweaking and buying expensive hardware. 

it's because we've taken the red pill. Some of us are probably so deep down that rabbit hole that I doubt we'll ever be able to resurface 😆

Edited April 3 by Cpt_Piett

[Noel 4,563]

4 minutes ago, Cpt_Piett said:

I'm sure you'll have no problem maintaining a constant 70FPS, even at major hubs, even with a near flat frametime graph. Although probably not with AI traffic.

Yes I did 70 fps out of several major hubs with 9900K when I was trying FG and really no problems there, and that is with limited FSLTL traffic.  I will increase traffic some, but only by about 30% tops if that.  It really comes down I think to the 2-screen problem.  There are workarounds for that, by simply not looking at it when the mouse click happens on the main screen.  As a SLC user I'm frequently going between the two screens so we'll see if it's worth it, i.e. going from perfection at 45-50fps, to the artifacts FG brings to get to 70fps:  I'll be shocked if is!

Here's a recent full gate to gate flight out of a major hub in the PMDG 738:  nary a stutter anywhere and that was with RS of 160%.  I can imagine this ought to get to 50fps w/ the 7800X3D.  The latter has improved a couple of annoyances:  no longer can see the voxel grid on the edges of some clouds in certain type of lighting--the grid is much smaller now at that render scale.  Plus, and I really like this:  runway edges are much more well defined now during initial approach and on in.

[Benjamin H 14]

7 minutes ago, Cpt_Piett said:

4090+7800X3D

When my 2080Ti+7800X3D combo delivers a constant 60FPS with vSync (locked, but unlimited I have seen 112 in the VC of the Fenix in Boston if I remember correctly when using DX12+FG Mod) I am certain that this combination will provide a simulator that can be enjoyed for years to come (and all games to be released as well as MSFS 2024 of course) so a lot to look forward to is all I have to say to the one who is waiting for this beast of a system. 

[Cpt_Piett 4,830]

1 hour ago, turbomax said:

micro-stuttering occurs when the amount of time that passes between frames appearing on-screen is inconsistent, resulting in what appears to be choppy gameplay, even in spite of high average frame rates. For example, two different PCs, both generating an average of 24 frames per second, may convey different experiences, one stuttery and one smooth, if the amount of time between each frame is less regular on one and more regular on the other.

2 hours ago, RJC68 said:

It means "Frame Time Variance" it is basically the timing between each frame. The more consistent it is the smoother the experience in the sim will be. For any given hardware setup if you make adjustments to the settings accordingly you can get a consistent FTV even at lower fps although having a low fps with a consistent FTV may help to eliminate stutters it is not so good on a large display such as a TV as I described in a post above.

Great explanations! Even though my FPS is locked at 70 (and stays there nearly all the time thanks to AutoFPS), every now and then there will be spikes on the frame time graph. This is always consistent with stutters - sometimes micro-stutters, other times hard stutters. Most common cause of these stutters nowadays on my system is changing between views or panning too quickly. Since I'm never GPU-limited, this is always due to being mainthread-limited (when the load on the mainthread is too high to produce consistent frame times). 

For the last few days I've been flying around Europe in the A300 without AI traffic and I swear the FPS has not deviated from 70, and the frame time graph has been near flat. LFMN, EDDK, EHAM, EGLL etc. Again, AutoFPS plays a HUGE part in accomplishing this. It's not that hard too achieve with decent hardware (a 4090 is not needed!) - there's plenty of information on how to do this on Avsim.

If I throw AI traffic into the mix - my CPU is simply not good enough for the task at hand. Or put in another way: MSFS in its current iteration is too poorly optimized for multithreading, not being able to take advantage of the latest and greatest in CPU technology. 

2 hours ago, RJC68 said:

Even FTV is important at any FPS setting, I have shown it many times on here before. Locked 80fps with FG and zero artifacts = a sublime experience. You need to have evenly matched hardware for it to happen though.

Indeed my friend! It needs to be experienced to be appreciated fully. And like you've mentioned - the bigger the monitor/TV size, the more it matters.

Sadly the experience cannot be reproduced fully in YT videos due to the 60FPS limit as well as the codec limitations with all but the most expensive video editing software. Another aspect is that Chrome doesn't support G-Sync, hence there will be screen tearing when watching YT-videos. 

1 hour ago, RJC68 said:

Here is a perfect example, I captured this 15 minute segment with CapFrameX from the threshold of EGGP in the Fenix with a good amount of FSLTL traffic and ultra settings with 6K shadows and a TLOD of 450 in the air.

Amazing result! I'm very happy with FTV at the moment, but still struggling with 1% and 0.1% low FPS. Can you remember from our conversations how I was able to achieve this in the past - I seem to have forgotten 😅

1 hour ago, Noel said:

Here's a recent full gate to gate flight out of a major hub in the PMDG 738:  nary a stutter anywhere and that was with RS of 160%.  I can imagine this ought to get to 50fps w/ the 7800X3D.  

Again an awesome result. You'll be able to reach 70 FPS at the major hubs (if you want to) with the same excellent FTV result. 

EDIT: 

Panning and buzzing (at default speed) around iniBuild's EGLL terminals in drone mode with no AI traffic (iniBuild's A300-600RF parked at cargo terminal): 

1 hour ago, RJC68 said:

Now Do I expect to see that somewhere like InI Builds Heathrow, of course not, I have the headroom where by I can drop my fps and GSYNC on my Sony X90K takes care of things but it never gets out of control.

Hmm I should actually try this test with the excellent AutoFPS at a larger hub and see how it performs.

The latest release build in AutoLOD mode performs exceptionally well!

I'll redo the test with AIG 50% traffic once it's up and running (does take a while to start as I've got 50+ GB installed). 

Edited April 3 by Cpt_Piett

[Cpt_Piett 4,830]

I made a little video to compare performance with and without AI traffic (the difference is pretty huge). It seems to have turned into a mini-tutorial on how to use the RTSS overlay combined with CapFrameX, as well as the significance of frame time variances (FTV) on smoothness. 

Hopefully it's somewhat useful, as many Bothans died in order to make the video. Sadly, Jar Jar Binks survived...

Edited April 3 by Cpt_Piett

[turbomax 3,307]

1 hour ago, Cpt_Piett said:

Hopefully it's somewhat useful,

incredibly useful. that's what I like about Captain Piett's contributions. he provides pure facts and figures for an ever more complex "game", the age of enlightenment with professional benchmarking of MSFS has begun. you are the Rob Ainscough of MSFS.

Edited April 4 by turbomax