5 hours ago, Bill A said:

Going back to my last video.. anyone know what could be causing those little skips? Since some of you gave mentioned that they get none of those frame issues. 

Hi,

I would think the tiny flickers are due to the AI aircraft. Have you tried removing the traffic and see if the flickers disappear? 

22 minutes ago, Mike_CFII_MEL said:

Hi,

I would think the tiny flickers are due to the AI aircraft. Have you tried removing the traffic and see if the flickers disappear? 

Hi Mike, I watch your videos and have commented on your videos many times during your live feed. I was wondering if the AI might be causing that, even though I don't have anything live I just have static plain set with FSLTL set to 3. I think for a test I did turn off all traffic, and I still ended up getting a couple of those little hiccups here and there. Pretty frustrating

10 hours ago, Virtual-Chris said:

A couple of things… TV judder reduction or motion smoothing or whatever it’s called does a crude form of frame generation, but it has significant latency. You won’t find this option in game mode on your TV for that reason. It works differently from Nvidia frame generation in that it waits to have two consecutive frames to work with that makes it easy to interpolate a few frames in between. Hence the massive latency. Of course, when you’re watching a movie, you don’t care if there’s several frames of latency, but if you’re gaming, you want what you see n the screen to respond instantly to your inputs.  What’s more, TV frame interpolation or judder reduction on most modern TVs will actually generate 4 frames for each real frame to effectively turn 24FPS film content into 120HZ video - which is why almost all TVs have a 120Hz refresh rate… it’s a common multiple for 24, 30, and 60FPS content. 
 

Nvidia’s FG is very different… it is building a new frame on the fly without having the next frame, it uses only the previous frames and uses AI (machine learning) to construct a new frame. No latency what so ever.

Now you could argue that a flight sim doesn’t really need ultra low latency, and you may find it works fine for you. Just be aware of the trade off you’re making. As always, YMMV. 

So nothing in this post has anything to do with my suggestion which was to use a 4000 series card and a TV with Judder reduction to give 100fps on screen for 25 pre FG and pre TV judder reduction and report back what the experience was. I'd love to hear from someone who can actually try it. Even so, I'm sure your wrong about a number of things in your post. Which is it, "significant latency" or "massive latency"?. Truth is it is neither, even in titles like Codemasters F1 which I play at 50hz for 100fps on screen with judder reduction on. The difference between the gameplay smoothness experience at 60fps with the judder reduction off and 100fps with it on is notable. In flight sim the difference between the raw 30fps and the 60fps with Judder reduction active is huge and there simply is no perceivable lag. I've had this conversation before and I've tested the lag with a shooting game. A purely subjective test. No fancy software for timing or anything like that. I just simply set myself up to take a shot using the trigger on XBox controller at 60hz for 60fps on screen and pulled the trigger and asked myself could I discern a delay between pulling the trigger and hearing the shot. I figured I could but I wasn't 100% certain of if. Did this for a while and concluded that I could barely discern a practically non existent lag between trigger pull and sound of shoot. I then did the same thing at 30hz for 60fps on screen using Lag inducing Judder reduction. I'll admit that I could discern a delay between pulling the trigger and hearing the shot but only barely but I also found that there was virtually no difference in both instances i.e. in either case I found that I could barely notice any lag and that it was only discernible as a result of taking a very hard look for it. I don't play shooters but I do play Codemasters F1. With 50 real fps coming from the GPU and another 50 interpolated frames per second provided by the TV. That game involves split second timing and I do very well. Just as well if not better than I do at 60fps raw with the no so called massive lag.

I'm pretty sure TV frame interpolation only generates one frame in between each real frame. I think your interpretation of how that works is wrong. Like PC monitors TV are compatible with a number of refresh rates. i.e. that can only handle signals with particular frequency's. Monitors used to be something like 30hz, 60hz, 75hz, 90hz. We used to hear stern warning not to send a signal to your display that exceeded its maximum refresh rate otherwise you might brake it. This may be still true. I don't know.

My TV which I am sure is typical according to NCP is compatible with 4K @ 23hz, 24hz, 25hz, 29hz, 30hz, 50hz, 59hz and 60hz. So if I set the refresh rate in NCP to 23hz, I get 23fps on screen. If I turn the TV judder reduction on I get 46fps on screen. If your 4KTV is labelled 120hz chances are its actually 60hz with frame interpolation and compatible with other refresh rates commonly used in TV shows, Sports and Movies. i.e. the ones above. Frame interpolation only generates one frame in between each real frame.

 

Edited September 1, 20232 yr by FBW737

1 hour ago, FBW737 said:

So nothing in this post has anything to do with my suggestion which was to use a 4000 series card and a TV with Judder reduction to give 100fps on screen for 25 pre FG and pre TV judder reduction and report back what the experience was. I'd love to hear from someone who can actually try it. Even so, I'm sure your wrong about a number of things in your post. Which is it, "significant latency" or "massive latency"?. Truth is it is neither, even in titles like Codemasters F1 which I play at 50hz for 100fps on screen with judder reduction on. The difference between the gameplay smoothness experience at 60fps with the judder reduction off and 100fps with it on is notable. In flight sim the difference between the raw 30fps and the 60fps with Judder reduction active is huge and there simply is no perceivable lag. I've had this conversation before and I've tested the lag with a shooting game. A purely subjective test. No fancy software for timing or anything like that. I just simply set myself up to take a shot using the trigger on XBox controller at 60hz for 60fps on screen and pulled the trigger and asked myself could I discern a delay between pulling the trigger and hearing the shot. I figured I could but I wasn't 100% certain of if. Did this for a while and concluded that I could barely discern a practically non existent lag between trigger pull and sound of shoot. I then did the same thing at 30hz for 60fps on screen using Lag inducing Judder reduction. I'll admit that I could discern a delay between pulling the trigger and hearing the shot but only barely but I also found that there was virtually no difference in both instances i.e. in either case I found that I could barely notice any lag and that it was only discernible as a result of taking a very hard look for it. I don't play shooters but I do play Codemasters F1. With 50 real fps coming from the GPU and another 50 interpolated frames per second provided by the TV. That game involves split second timing and I do very well. Just as well if not better than I do at 60fps raw with the no so called massive lag.

I'm pretty sure TV frame interpolation only generates one frame in between each real frame. I think your interpretation of how that works is wrong. Like PC monitors TV are compatible with a number of refresh rates. i.e. that can only handle signals with particular frequency's. Monitors used to be something like 30hz, 60hz, 75hz, 90hz. We used to hear stern warning not to send a signal to your display that exceeded its maximum refresh rate otherwise you might brake it. This may be still true. I don't know.

My TV which I am sure is typical according to NCP is compatible with 4K @ 23hz, 24hz, 25hz, 29hz, 30hz, 50hz, 59hz and 60hz. So if I set the refresh rate in NCP to 23hz, I get 23fps on screen. If I turn the TV judder reduction on I get 46fps on screen. If your 4KTV is labelled 120hz chances are its actually 60hz with frame interpolation and compatible with other refresh rates commonly used in TV shows, Sports and Movies. i.e. the ones above. Frame interpolation only generates one frame in between each real frame.

 

There's lots of good articles online about motion interpolation in TVs. There's no doubt it introduces latency, there are sets which introduce multiple frames, and there are sets that run at a true 120HZ refresh rate (my LG C1 included). And I don't think there's any debating whether it's a possible substitute or augmentation for Nvidia's frame generation technology. I'm not going to test it, as motion interpolation is not an option on my LG in game mode where G-Sync and low latency are key benfits... but maybe someone else will. 

42 minutes ago, Virtual-Chris said:

There's lots of good articles online about motion interpolation in TVs. There's no doubt it introduces latency, there are sets which introduce multiple frames, and there are sets that run at a true 120HZ refresh rate (my LG C1 included). And I don't think there's any debating whether it's a possible substitute or augmentation for Nvidia's frame generation technology. I'm not going to test it, as motion interpolation is not an option on my LG in game mode where G-Sync and low latency are key benfits... but maybe someone else will. 

Good. I doubt if you would be a credible witness.🤪

23 hours ago, AnkH said:

BTW: you have a 9 too much in your signature 😉 

and an "e" missing. he can have one of mine. 😊

4 hours ago, FBW737 said:

warning not to send a signal to your display that exceeded its maximum refresh rate otherwise you might brake it.

that was an issue with old analog picture tube TVs and monitors, where an increase in horizontal deflection frequency (15.625 Hz on PAL TVs or pc monitors @ 1024×768 pixel @ 85 Hz fps ca. 68.700 Hz) would increase the high voltage required to drive the picture tube because the high voltage (20.000 - 30.000 volt) was derived as a direct multiple of its horizontal deflection unit. picture initially got brighter with fps increase and then black = kaputt. those were the days. 🤣

 

Edited September 1, 20232 yr by turbomax

10 hours ago, cappy42 said:

I have an RTX 4070 and my only issue with Frame Generation is the constant screen flickering.  Someone else said to turn off Flare or Bloom or something, and I did that, but to no avail.  And yes, Gsync is enabled.  I run TAA because I can't stand the blurry gauges in DLSS.

What refresh rate does your monitor have? if fps goes higher than your monitor's refresh rate you will get flicker because VSync doesn't work currently with FG, you can try to cap your framerate 3 fps lower than your monitors refresh rate in NCP.

Edited September 1, 20232 yr by Ixoye

I have a 4080 and leave FG off. FG still doesn’t come quite close to the level of continuous smoothness (you can see that in turns , taxiing on ground while looking sideways)  I can achieve while locking frames at 30 and then on top of that using my Sony’a 4K TV frame generation (motion smoothing).

There is no latency problem to me eyes (I guess Flight simming is too slow for this to be a real problem…)  

 

 

27 minutes ago, anthony500 said:

I have a 4080 and leave FG off. FG still doesn’t come quite close to the level of continuous smoothness (you can see that in turns , taxiing on ground while looking sideways)  I can achieve while locking frames at 30 and then on top of that using my Sony’a 4K TV frame generation (motion smoothing).

There is no latency problem to me eyes (I guess Flight simming is too slow for this to be a real problem…)  

 

 

Get a g-sync and change your mind after you’ve used it:-) 

And if you use trackIR, latency becomes an issue. 

2 hours ago, anthony500 said:

I have a 4080 and leave FG off. FG still doesn’t come quite close to the level of continuous smoothness (you can see that in turns , taxiing on ground while looking sideways)  I can achieve while locking frames at 30 and then on top of that using my Sony’a 4K TV frame generation (motion smoothing).

There is no latency problem to me eyes (I guess Flight simming is too slow for this to be a real problem…)  

 

 

Just for the sake of curiosity you should try Nvidia FG and the TV motion smoothing at the same time per my post above. Set your refresh rate in NCP to 50, turn on your Nvidia FG and turn on your TV motion smoothing. By my reckoning in a heavy test scenario you should be able to cripple your system to less then 30 fps, right?  Nvidia FG will turn that into 50 something FPS. NCP will cap that to 50fps with Vsync. The TV wont distinguish between the real frames and the generated frames, it will just do its job of interpolating a frame between each frame it gets so we go from 50fps at 50hz to 100fps on screen. That gets us from 25fps to 100fps. Give it a try and see what happens. If I had a 4000 series GPU it'd be the first thing I'd do!🤪

Edited September 1, 20232 yr by FBW737

2 hours ago, anthony500 said:

I have a 4080 and leave FG off. FG still doesn’t come quite close to the level of continuous smoothness (you can see that in turns , taxiing on ground while looking sideways)  I can achieve while locking frames at 30 and then on top of that using my Sony’a 4K TV frame generation (motion smoothing).

There is no latency problem to me eyes (I guess Flight simming is too slow for this to be a real problem…)  

 

 

I guess thats why I feel like FG use comes to personal preference.  Locking at 30fps is all good, but it gives no room for that random slight drop thats the difference between smooth and stutter.  Also, i thought 30fps was smooth until i upgraded to the 4070 and turned on FG and was then was getting 50fps+ during take off roll and switching camera views.  The smoothness is completely different for me.  I can't go back to 30fps, its just not the same experience as 60fps+ with FG.  I think i have my settings good to a point where i get complete smoothness with FG. 

19 hours ago, Bill A said:

I followed all those settings. I do use riva tuner but just to show the frames only. Maybe I'll try to just not have it running at all. Did you notice the jitters towards the landing of my last video? I'm on a 60mgz monitor. The monitor can run 75 megahertz, but I choose to keep it at 60 since my monitor on the side can only run at 60., It's a different brand and not the same monitor. . I don't know if that matters or not.

Hey Bill,

I watched your video and what you saw when the buildings came into view looks the same as what I was seeing when I was messing with RTSS. Have you tried a test with no afterburner/RTSS running? I would be curious to see if there’s something going on in the background. Also have you tried your main monitor on its own?

With the settings Chris posted my sim Is the most fluid I’ve ever seen in all the years I’ve been a flight sim enthusiast. Your rig is not that dissimilar to mine, if anything your CPU is probably a little stronger than my 12700K

Hope you can figure this out

Just now, RJC68 said:

Hey Bill,

I watched your video and what you saw when the buildings came into view looks the same as what I was seeing when I was messing with RTSS. Have you tried a test with no afterburner/RTSS running? I would be curious to see if there’s something going on in the background. Also have you tried your main monitor on its own?

With the settings Chris posted my sim Is the most fluid I’ve ever seen in all the years I’ve been a flight sim enthusiast. Your rig is not that dissimilar to mine, if anything your CPU is probably a little stronger than my 12700K

Hope you can figure this out

I can uninstall afterburner, even though I don't use it and it's not running. I tried running the MSFS without Riva tuner for showing the frame rates only and the same thing happened. The other monitor I only use for navagraph charts v pilot, pac X. 

Should I not use the other monitor and have nothing there for a test? 

19 minutes ago, Bill A said:

I can uninstall afterburner, even though I don't use it and it's not running. I tried running the MSFS without Riva tuner for showing the frame rates only and the same thing happened. The other monitor I only use for navagraph charts v pilot, pac X. 

Should I not use the other monitor and have nothing there for a test? 

Frame generation is a bit like gsync, it won’t work unless the sim Window is in main focus. So, if you switch to another monitor and use Charts etc… or even open them up on the same monitor, you’ll disengage FG, which can also cause a momentarily stutter.