Upscaling and Frame Generation — When to Use Them and When Not To
Upscaling renders fewer pixels. Frame generation invents whole frames. They have completely different trade-offs and one adds latency.
The short answer
Upscaling renders the game at a lower internal resolution and reconstructs it, which genuinely increases performance and reduces latency. Frame generation inserts synthetic frames between real ones, which raises the displayed frame rate but slightly increases input latency because a generated frame must wait for the next real frame to exist. Upscaling is almost always worth enabling; frame generation is best used when the base frame rate is already above about 60.
Verified and last updated August 12, 2026.
- Time needed
- 12 min
- Difficulty
- Intermediate
- Read time
- 3 min
Upscaling and frame generation get discussed as one feature. They are two very different things with opposite effects on latency, and conflating them is why people have contradictory opinions about both.
The essential difference
Upscaling renders fewer pixels and reconstructs the rest. More performance, less latency. Frame generation invents whole frames between real ones. Higher displayed frame rate, more latency. One is nearly always worth using and the other depends entirely on context.
What each actually does
Upscaling: use it
At quality presets and at 1440p or above, upscaling frequently produces an image indistinguishable from native — and occasionally better, because the reconstruction includes an anti-aliasing pass superior to what most games ship with.
The performance gain is real and latency goes down, because the GPU is doing less work per frame.
Where it struggles: aggressive presets, and low output resolutions. Upscaling to 1080p from a very low internal resolution does not have enough source information, and it shows in motion.
Practical rule: quality preset at 1440p and above, and be more cautious at 1080p.
Frame generation: it depends
The mechanism creates an unavoidable cost. To insert a frame between two real frames, the second real frame must already exist — so it is held back briefly. That is added latency, by construction.
The top row is why the feature has a mixed reputation. Applying it to a low base rate produces something that looks smooth and feels wrong, which is more unpleasant than a low frame rate that at least feels consistent with what it looks like.
Get to 60 first, then generate.
The settings order that works
Enable upscaling at the quality preset 2 min
Nearly free performance with reduced latency. Do this before lowering any graphics settings.
Adjust graphics settings until you hold 60 or more 15 min
Shadows and volumetrics first — they cost the most for the least visible benefit.
Only then consider frame generation 5 min
And only in single-player games. Not in anything competitive.
Enable the vendor latency reduction feature 2 min
It partly offsets generation's cost and helps regardless. There is no reason to leave it off.
Which upscaler
Implementation quality varies more between games than between technologies. A well-implemented open upscaler beats a poorly implemented vendor one.
Vendor-specific options generally look best on their own hardware. Open alternatives work everywhere with slightly softer results.
Try both in the game you actually play and judge in motion. Screenshot comparisons systematically favour whichever reconstructs static detail better, which is not the same as what looks good while moving.
What works
What doesn't
Key takeaways
Frequently asked questions
Does upscaling make games look worse?
At quality presets and higher resolutions, frequently not, and sometimes it looks better than native because the reconstruction includes a superior anti-aliasing pass. The artefacts appear at aggressive presets and at low output resolutions, where there is simply not enough source information to reconstruct from. At 1440p and 4K on a quality preset, most people cannot reliably tell in normal play.
Does frame generation add input lag?
Yes, inherently, because a generated frame is interpolated between two real frames and therefore the second real frame must already exist before the generated one can be shown. That means holding a frame back. The added latency is modest and is partly offset by latency-reduction features, but it cannot be eliminated, which is why it suits single-player games rather than competitive ones.
Should I use frame generation in competitive games?
No. Competitive titles already run at very high frame rates on modest hardware, so there is nothing to gain, and the added latency is a direct cost in exactly the games where latency matters most. Use upscaling if you need performance, and leave frame generation off.
What is the minimum frame rate for frame generation to feel right?
Roughly 60 before generation is applied. Below that, the underlying input latency is already high and generation makes it worse while producing a frame rate number that no longer reflects responsiveness. The result is a game that looks smooth and feels sluggish, which is a genuinely unpleasant combination and the main reason people dislike the feature.
Which upscaler should I use?
Whichever is native to your hardware and best implemented in that specific game, since implementation quality varies more between games than between the technologies. Vendor-specific upscalers generally look best on their own hardware, while the open alternatives work across all hardware with slightly softer results. Try both in the game you play and judge in motion, not in screenshots.
About the author
PC Hardware & Performance Editor
Marcus has been building, breaking and repairing gaming PCs since 2011, first as a weekend favour for friends and then for eight years as a bench technician at an independent repair shop in Portland, where roughly a third of the job was diagnosing machines that "ran fine yesterday".
Found something wrong? Games patch and fixes go stale. If a step here no longer works, tell us at hello@spyld.com and we will retest it. Read how we test and our editorial policy.