Blur Busters Forums

Hi Blur Busters, I knew this would be the best place to post a monitor-related question so I gladly made an account to do so!

Hope all of you are doing well with what's going on in the world.

Here's my current predicament: I scratched my Dell S2716DG (1440p 144Hz G-Sync) panel last night...

It's not a bad scratch per-se, but it bothers me personally, so I'm trying to decide between two options: Go with another S2716DG and stay with G-Sync or go with a FreeSync 1 or 2 display. These FreeSync displays specs call out to me. I wouldn't mind going with something ultrawide, 144Hz+, etc

I love my current monitor, so I have no issue getting the exact same one, however I had a few questions about FreeSync 1 or 2 in regards to my current use-cases of my G-Sync monitor

My current GPU is a GTX 1080, which works perfectly with G-Sync, but I was reading a bit about how LFC works best/only with AMD GPUs?

Also, I REALLY like how my G-Sync range starts at 30Hz, as I play Red Dead Redemption (and a few other) games locked to 30fps on PC, but I've read how most FreeSync ranges seem to start at like 40Hz+, would this mean my games locked to 30fps won't benefit from FreeSync?

I'm a lock-to-30fps-for-a-cinematic-experience guy at times, so VRR range would be very important to me

Thanks for the help, and if you guys need any additional info, please let me know, would be glad to share anything I can

MarshallHurtado wrote: ↑

28 Mar 2020, 15:04

Here's my current predicament: I scratched my Dell S2716DG (1440p 144Hz G-Sync) panel last night...

Ouch!

MarshallHurtado wrote: ↑

28 Mar 2020, 15:04

My current GPU is a GTX 1080, which works perfectly with G-Sync, but I was reading a bit about how LFC works best/only with AMD GPUs?

Perfect LFC is indistinguishable from perfect non-LFC. LFC looks bad only when the new refresh cycles collides with a monitor-busy-still-refreshing. That's how LFC stutters occur. So LFC behaves flawless for perfect framepacing, while LFC adds stutter if the frame pacing is imperfect.

In some cases, LFC can look better than non-LFC for consistent-framerate situations (e.g. watching a movie with VRR-compatible video player). This is because you're avoiding the flickering caused by low-Hz LCD deay and/or low-frequency inversion artifact flickering, by instead running at a higher Hz. However, the lack of variable overdrive built into cheaper FreeSync panels, can make motion look worse at some framerates than others. If you get FreeSync, get the most premium kind of FreeSync you can afford, and preferably one that is also NVIDIA certified too (G-SYNC Certified). Those are better quality VRR.

MarshallHurtado wrote: ↑

28 Mar 2020, 15:04

Also, I REALLY like how my G-Sync range starts at 30Hz, as I play Red Dead Redemption (and a few other) games locked to 30fps on PC, but I've read how most FreeSync ranges seem to start at like 40Hz+, would this mean my games locked to 30fps won't benefit from FreeSync?

If you have perfect framepacing (like a movie), 30fps looks perfect as if it had native Hz. LFC vs nonLFC can look identical in an ideal situation.

And sometimes LFC is occasionally superior looking on some models, when it can manage to avoid those slowly-scrolling chess board texture appearing in solid colors (inversion artifact in solid backgrounds during slow turns / horizontal panning / scrolling) and/or flickering effects. But it can be much worse. It all depends on the variables.

However, if you let your framerate vary a bit, the new frames might start colliding with those predictively-timed repeat-refresh-cycles (causing the frame to wait for the monitor) -- creating stutter.

LFC simply tries to predict a repeat-refresh to occur between two frames. And as you see in high speed video of LCD refreshing, www.blurbusters.com/scanout -- it takes 1/240sec to refresh all pixels (in a top-to-bottom fadesweep) = a monitor busy for 4.2 milliseconds refreshing a single 240Hz refresh cycle. This will remain constant a lower refresh rates on a VRR monitor, so even at 50 frames per second, the monitor still only needs 1/240sec to refresh.

LFC algorithms are very reliable with steady low frame rates, because it's easy to predict a repeat-refresh right in between. A repeat-refresh in an ideal situation is a no-operation (you see no visible effect on the screen because an image is being replaced by a duplicate image, so you can't tell LFC from non-LFC)

But LFC fails when frametimes vary a lot, so sometimes the repeat refresh starts, then the game finishes rendering a frame, and then suddenly the game is waiting for the monitor to finish repeat-refreshing (an old frame) before it can display the new frame. Thus, stutter. The good news is that this becomes less at higher Hz.

The LFC collision window is always max-Hz. (A frame-finish-rendering being forced to wait for a monitor-still-busy-repeat-refreshing). So the higher the VRR Hz, the smaller the LFC collision window is. On a 48Hz-240Hz VRR monitor, the LFC frame-vs-rerefresh collision window creates a maximum of 4.2ms (1/240sec) stutter in the worst-case scenario. The average LFC collision will be the halfpoint of that since the stutter error will be between [0...4.2ms]. Now if you got a lower maximum Hz such as 144Hz, your LFC collision window would be 6.9ms (1/144sec), so LFC stutters are worse on a 144Hz monitor than 240Hz monitor. So, if you're so worried about LFC stutter, make sure your max-Hz is higher to compensate.

Now if you buy that new 360 Hz monitor (future model, not sure of VRR range), and if it uses LFC algorithms (both NVIDIA and AMD use similar algorithms now), stutters from LFC algorithms on 360Hz will be at most, a 2.8ms stutter (2.8 pixel stutterjump at 1000 pixels/sec motion) at worst case, but a random number between [0...2.9ms] would be only 1.4ms stutter average (1.4 pixel stutterjump at 1000 pixesls/sec motion). At this point, without a strobe backlight, this begins to become hidden in low-framerate stutter, since 48fps at 1000 pixels/sec creates (1000/48) = 20.8333 pixels of motion blurring, or 20.8333 pixels of objectjump.

So with a HUMONGOUS variable refresh rate range, e.g. 48...360, then the LFC stutters completely fall into the noisefloor of low-framerate stutter! Y'know (with proper drivers & proper LFC algorithm) even 1.4ms stutter error being completely lost in 20.8 pixel stutter at 48 frames per second. Big whooooop-deeee-do. The virtue of a massive VRR range works in the favour of LFC!

To play it safe, please stick to high-rated VRR. There are artifacts of cheap uncertified VRR (generic adaptive sync with no AMD or NVIDIA certifications) that can look worse than LFC artifacts.

TL;DR:
- LFC doesn't add any stutter if you have consistent low framerate (like a perfect 30fps movie)
- LFC can worsen stutter for volatile low framerates (frametimes varying frequently across refreshtime of min-Hz).
- LFC stutter error is directly proportional to max-Hz
- LFC stutter error (in milliseconds) averages out to equalling half the duration of a max-Hz (e.g. 2.1ms for 240Hz).
- Thusly, LFC becomes unnoticeable with wide VRR ranges like "48Hz-360Hz" instead of "48Hz-120Hz".
- Thusly, if worried about LFC stutter
.....Framepace your low framerates well to help LFC work better
.....And get the biggest VRR range you can afford
- Premium VRR (G-SYNC certification and higher-end FreeSync) is worth it for other reasons than LFC too, but depends on goals.

Now assuming you stuck to 1440p 144Hz and min Hz was 48Hz:

Assuming a random single-refresh-cycle LFC "fail":

Max LFC Stutter Error Margin: 1/144sec = 6.94ms
Average LFC Stutter Error Margin: 0.5/144sec = 3.47ms
we already know your sttuter error margin for any LFC of any vendor, will be 1/144sec (6.9ms).
This is equal to an average 3.47 pixel stutter-jump during 1000 pixels/second motion.

Stutter of 48Hz = 1/48sec = 20.8ms
At 48Hz refresh rate, your objects will move 20.8 pixels per 1000 pixels second (1000/48).

LFC Stutter Injection During LFC Collision = 3.47ms/20.8ms = 17% randomization added to stutter.
If an LFC collision occurs, your average stutter error at 48fps will be a maximum of 17%. This is very small and usually unnoticeable. But it depends on how sensitive you are. But remember, not all refresh cycles have an LFC collision event.

LFC Collision Likelihood at a specific framerate = hard to calculate
Theoretically, (1/144) / (1/48) = roughly one third of the time
This is only a rough guideline that will never be an exact science. This is harder to reliably calculate because LFC collisions don't always happen every refresh cycle. Framerate volatility patterns will have a major effect on this. There are harmonic and beat frequency effects that may cause LFC to almost never happen, versus LFC always happening. However, the faster the scanout, and the lower the LFC Hz, the less likely for an LFC collision to happen. Someone should write a complex algebra/calculus formula, that includes beta/deviation/volatility variables, and it would vary from LFC algorithm to algorithm because there's some predictivity (And potentially AI) involved in timing the LFC. And LFC algorithms are getting better and better even at volatile framerates.

Firstly, I can easily understand why they call you Chief Blur Buster... your response was EXTREMELY fast and EXTREMELY detailed, so thank you VERY much

Secondly, I apologize for my reply format as I tried to reply to your quotes but failed as I don't usually post in forums so I'm unfortunately keeping it simple

In reference to your LFC response, essentially as long as I stay with a high-rated VRR/Freesync 1 or 2 monitor, one that is preferably Nvidia-certified, I should have an on-par experience with my current G-Sync monitor? In other words, will the LFC error stutters seem the same as G-Sync stutters, so to speak?

I would safely say that if I were to go with a FreeSync 1 or 2 monitor, I would go 2560x1440 - 3440x1440, 144Hz-200Hz. Any monitor with a refresh rate above 200Hz would be out of he question for me unfortunately

When I play RDR2 locked to 30fps using Nvidia Control Panel, there are times where I do drop below 30fps which can of course cause visual interference of any kind (stutter, etc) but I'm fine with that if that makes sense? Last night I did a test between G-Sync on V-Sync on and G-Sync OFF V-Sync on and could definitely notice a different between the two at 30fps, but not a massive one. G-Sync did just enough to make my experience feel smoother overall.

Is a higher-quality VRR monitor (FreeSync 1 or 2, Nvidia-certified etc) going to render my games locked to 30fps as traditional V-Sync?

I'd be lying, as most would, if I said I 100% understood your responses as they are obviously very expert-level, so I again apologize for asking simple follow-up questions to understand this better

I do appreciate the TL;DRs for me though!

MarshallHurtado wrote: ↑

28 Mar 2020, 15:45

In reference to your LFC response, essentially as long as I stay with a high-rated VRR/Freesync 1 or 2 monitor, one that is preferably Nvidia-certified, I should have an on-par experience with my current G-Sync monitor? In other words, will the LFC error stutters seem the same as G-Sync stutters, so to speak?

Generally, yes. That said, AMD LFC tends to be somewhat better than NVIDIA LFC on AMD FreeSync cards. So there's some potential LFC quality degradation (or loss of LFC) when forcing a G-SYNC Compatible mode on an AMD FreeSync monitor.

MarshallHurtado wrote: ↑

28 Mar 2020, 15:45

I would safely say that if I were to go with a FreeSync 1 or 2 monitor, I would go 2560x1440 - 3440x1440, 144Hz-200Hz. Any monitor with a refresh rate above 200Hz would be out of he question for me unfortunately

See my math above to see if the LFC error margins are acceptable for 144Hz.

MarshallHurtado wrote: ↑

28 Mar 2020, 15:45

When I play RDR2 locked to 30fps using Nvidia Control Panel, there are times where I do drop below 30fps which can of course cause visual interference of any kind (stutter, etc) but I'm fine with that if that makes sense? Last night I did a test between G-Sync on V-Sync on and G-Sync OFF V-Sync on and could definitely notice a different between the two at 30fps, but not a massive one. G-Sync did just enough to make my experience feel smoother overall.

Yes, sometimes the LFC stutter deviations are so minor that it falls in the noisefloor of the low-framerate "regular stutter".

LFC stutter at 48Hz on a 48-144Hz VRR range would, at most, add only 17% erroneous stutterjump. This is vastly superior to VSYNC ON stutters at 60fps vs 30fps since sudden 60fps to 30fps jumps is a far bigger stutter error (full doubling 100% more stutterjump with 60fps suddenly to 30fps, or halving 50% stutterjump 30fps suddenly to 60fps). So LFC stutter is often far by, the lesser of evil, than fixed-Hz stutter errors -- fixed refresh rate stutter is generally always worse than LFC stutter.

The way VRR seamlessly erases framerate transitions outweighs the occasionally injected LFC stutters -- as long as you choose wide VRR ranges. Such as 30-144 range or 48-240 range. LFC stutter is more noticeable on displays with narrower VRR ranges (e.g. 48-120). VRR displays with less than double min Hz, can't do LFC (e.g. 48-75Hz is incapable of LFC because there's not enough time between refresh cycles to insert a repeat-refresh).

The higher the Hz, the easier it is to raise LFC min Hz without stutter artifacts being injected.

Based on what you say, LFC stutter isn't currently bothering you. As long as you've successfully enabled LFC, good LFC shouldn't bother you even at 48Hz. In fact, 48Hz LFC will make "near-perfect 30" more reliable because "almost perfect 30" will rapidly oscillate into/out of LFC territory, increasing LFC stutters at capped framerates equalling the min Hz. Since caps are not microsecond-perfect, and frametimes will vary at least microscopically, this will sometimes drive LFC algorithms crazy, adding permanent LFC stutter (that you may not be seeing) at capped framerates equalling VRR min-Hz.

I would bet smarter LFC algorithms will "watch" the framerate range rather than instantaneous framerate, and simply go into permanent-LFC even at higher Hz, e.g. 31fps or 32fps becomes LFC on a GSYNC "30-144" monitor if it is seeing framerates becoming volatile across the LFC boundary (i.e. 28-35fps with a 31fps or 32fps average). Basically, smart LFC algorithms watch the framerate range, and if the framerate valley falls into LFC range, then the whole range becomes temporarily perma-LFC to become flawless LFC because you cannot have a partially-LFC framerate range without LFC collisions. Then when the framerate range finally stays above the LFC floor, then LFC can deactivate completely seamlessly. That's how smarter LFC algorithms can avoid stutter for minor framerate volatility that "fuzzes across" the LFC boundary.

Nontheless, the point I am making is you might be in perma-LFC without knowing it, and worrying too much about LFC that you actually didn't know you were permanently seeing for your 30fps-capped games.

Think of it this way -- the moment your game runs briefly at 29.99999fps, your 30fps game (Fluctuating in a narrow range between 29.9fps and 30.1fps) goes into permanent LFC until the framerate range goes up far above the LFC min-Hz (example: 34fps-35fps). Good LFC does this to keep LFC-vs-non-LFC transitions seamless. For example.

However, this might not not necessarily be on NVIDIA cards connected to uncertified AdaptiveSync displays -- the drivers may enable VRR but may not bother activating such smart LFC algorithms. (Things may have improved since).

MarshallHurtado wrote: ↑

28 Mar 2020, 15:45

I'd be lying, as most would, if I said I 100% understood your responses as they are obviously very expert-level, so I again apologize for asking simple follow-up questions to understand this better

That's okay. Blur Busters is like that.

MarshallHurtado wrote: ↑

28 Mar 2020, 15:45

I would safely say that if I were to go with a FreeSync 1 or 2 monitor, I would go 2560x1440 - 3440x1440, 144Hz-200Hz. Any monitor with a refresh rate above 200Hz would be out of he question for me unfortunately

Be careful how you pair up the technology, AMD on FreeSync, and NVIDIA on G-SYNC, if you want to get your vendor's good magic-sauce LFC algorithms. We've seen LFC get dicey on NVIDIA cards running on Adaptive Sync displays, so be warned. More study is needed.

LFC can be completely the graphics driver responsibility when it comes to FreeSync / VESA Adaptive-Sync / HDMI VRR. Theoretically, the display doesn't even need to be LFC-aware, it simply needs a refresh rate range wide enough to accomodate intelligent LFC algorithms. Which is when max Hz is slightly more than twice min Hz. Bigger margins are always better for even smarter LFC algorithms, due to the framerate-range fuzz explained above in the previous post.

But it is also the driver vendor's responsibility to intentionally enable their magic-sauce LFC algorithms on FreeSync/AdaptiveSync/HDMI VRR in their graphics drivers, and it is quite possible for a driver to intentionally not do LFC on an uncertified monitor. (There are also technological reasons: Some cheap uncertified Adaptive Sync displays flicker badly when transitioning frequently into and out of LFC, because of the sudden frequent dramatic refreshtime changes that the sudden activation and deactivation of repeat-refreshes can cause).

Until more study is done, if you want guarantee your vendor's best LFC algorithms, keep NVIDIA with G-SYNC. Or AMD with FreeSync Premium (or better).

Fortunately, it is my understanding that G-SYNC Compatible Certified monitors are currently activating some form of LFC with NVIDIA GPUs, so if you get a FreeSync compatible panel, make sure it is also G-SYNC Compatible Certified if you are using an NVIDIA GPU.

After reading all of your explanations I now feel I have a much better understanding of how all of this works, to which I thank you again, you guys know a TON about all of this stuff. As you said, it's best to stick with G-Sync with my Nvidia GPU when possible, and at the very least I should go with an Nvidia-certified monitor if I were to try FreeSync

I did have a follow-up question regarding my RDR2 experience. You said I likely may be already experiencing but not perceiving/noticing LFC stuttering, to which I also find that possible and very interesting. I wanted to explain my frame-limiting method further, just in case you have a different response:

Nvidia Control Panel:

Low-latency mode: Ultra
Max framerate: 30fps
Monitor technology: G-Sync
Triple buffering: On

RDR2 in-game settings:

Refresh rate: 144Hz
Screen type: Fullscreen
VSync: On
Triple buffering: On

MarshallHurtado wrote: ↑

28 Mar 2020, 17:24

Refresh rate: 144Hz
Screen type: Fullscreen
VSync: On
Triple buffering: On

Your settings probably work OK for the specific game you're playing, however:

When enabling G-SYNC + VSYNC ON via drivers (NVIDIA Control Panel), some games function better with VSYNC OFF (Game Settings). That's assuming you don't choose "Application Controlled" in the NVIDIA Control Panel.

Basically, making the game think it's running VSYNC OFF, but ends up becoming G-SYNC + VSYNC ON via the driver settings.

This is not always necessary, just simply that some games use algorithms reserved for fixed refresh rates whenever using a VSYNC ON setting. It's a game-by-game quirk -- no difference in many (most games) -- but some games may work better with GSYNC / FreeSync displays when they're configured to unsynchronized frame rates.

If there's no VRR problems, don't bother changing this setting -- just informing you that there's been situations where GSYNC behaves better on a game that thinks its running in VSYNC OFF mode (Because both VSYNC OFF and GSYNC are asynhronous/dynamic with no fixed intervals).

It depends on how the game developers programmed the particular game, though.

MarshallHurtado wrote: ↑

28 Mar 2020, 17:24

Nvidia Control Panel:

Low-latency mode: Ultra
Max framerate: 30fps
Monitor technology: G-Sync
Triple buffering: On

RDR2 in-game settings:

Refresh rate: 144Hz
Screen type: Fullscreen
VSync: On
Triple buffering: On

Just a quick comment; at best, the triple buffer settings will do nothing for G-SYNC, so you can leave them both off (G-SYNC natively functions on a double buffer).