Some games exhibit poor DLSS implementation due to inadequate training data, less than ideal integration within the game engine, or simply because developers prioritize other aspects of the game over DLSS visual quality.
It’s frustrating, isn’t it? You enable DLSS expecting a performance boost with minimal visual loss, yet sometimes, the image quality takes a nosedive. The burning question then becomes, why does dlss look so bad in some games? It is something many gamers grapple with after all.
Poor results often stem from how the game uses the technology. A hurried implementation or insufficient training data can produce blurry or artifact-ridden visuals. Let’s explore the reasons for these inconsistencies.
Why Does DLSS Look So Bad in Some Games?
DLSS, or Deep Learning Super Sampling, is like a magic trick for video games. It uses powerful computer smarts to make games look sharper and run smoother. It’s super cool when it works well! But sometimes, instead of looking amazing, DLSS can make games look kind of yucky. Why does this happen? It’s not always a simple answer, and there are actually quite a few reasons why DLSS can look bad in some games. Let’s dive in and explore these reasons, so you can understand what’s going on behind the scenes.
Understanding How DLSS Works (The Simplified Version)
Before we get into the “bad” stuff, let’s quickly talk about how DLSS is supposed to work. Imagine you have a drawing, but it’s a little blurry and low resolution. DLSS is like a super-smart artist who can take that blurry picture and use its knowledge to make a sharper, clearer image. It does this by looking at the image and using special data called “training data” that it has learned from. This training data is from real high-resolution images. It guesses what the missing detail should be, creating a better image than what was originally there. This whole process happens super fast, allowing your game to run at a smoother frame rate with better image quality, or at least that’s the plan.
Why the Picture Gets Fuzzy: Issues With DLSS Implementation
Now, let’s talk about why this picture-perfect process sometimes goes wrong. The truth is, DLSS is a complex technology and a lot of things have to work together perfectly to get the best results. When things are not quite right, it’s easy for visual issues to crop up.
Game Developers and Integration
The game developers play a big part in making DLSS work properly. If they don’t integrate DLSS correctly in their game engine, it won’t matter how good the technology is, the result won’t be pretty. It is like trying to fit a square peg into a round hole – it just won’t work.
- Poor Integration: Sometimes, developers rush and don’t take the time to integrate DLSS well. This could result in blurry, ghosting, or shimmering effects.
- Incorrect Settings: Developers also have to set up DLSS correctly within their game and provide players with different quality settings (like performance, balanced, quality) to select from. Incorrect settings can cause visual artifacts.
- Lack of Testing: If developers don’t test DLSS enough, bugs or visual issues that are only present with DLSS might slip through the cracks.
DLSS and Different Quality Modes
DLSS isn’t just one setting; you usually have different quality modes to choose from. These modes trade off between image quality and performance. Picking the wrong mode for your gaming system or resolution can lead to a poor experience.
- Performance Mode: This setting aims for the highest frame rates but often compromises on image quality, resulting in more blurriness, artifacts, and overall fuzziness. It’s like trying to zoom in too much on a low resolution image.
- Balanced Mode: This tries to find a sweet spot between quality and performance but can still result in slight visual compromises if not implemented well.
- Quality Mode: This setting aims for the best image quality but can still have issues, although generally fewer than performance or balanced modes.
- Ultra Performance mode: Ultra Performance mode aims to maximize performance while drastically dropping the resolution. This means it needs to do a lot of upscaling, and can often result in a very blurry image.
Resolution and Input Image Quality
The resolution of the game and the quality of the input image is important for DLSS to function well. The upscaling algorithm is designed to take in information, and if it has too little to work with, it cannot make a good guess at what is missing.
- Low Base Resolution: If your base resolution (the actual rendering resolution before DLSS) is too low, DLSS has less information to work with, leading to a blurry output. It’s like trying to make a clear picture from a super pixelated photo – there is just not enough data.
- Poor Anti-Aliasing: If the game doesn’t have good anti-aliasing before DLSS, the sharp edges in the original image can cause artifacts and shimmering when DLSS tries to upscale them.
Temporal Instability and Ghosting
One of the most common and frustrating issues with DLSS is “temporal instability” or “ghosting”. This is when objects seem to leave a trail or a faint double image as they move. It’s like a ghost following them. This occurs when the frame data between one frame and the next aren’t aligned perfectly.
- Motion Vector Issues: DLSS relies on motion vectors to understand how objects are moving. If these are incorrect or the game renders them inaccurately, it leads to ghosting.
- Tons of Dynamic Elements: When many things are moving quickly at the same time, such as in a large battle scene, the complex movements can confuse DLSS and lead to ghosting or other artifacts.
Artifacts and Shimmering
Besides ghosting, other visual problems can crop up. Artifacts are strange visual glitches that weren’t part of the original picture. Shimmering is where edges seem to flicker or vibrate.
- Algorithm limitations: Even the best algorithm has its limits. If the data it receives is too noisy or complex, it can produce visual glitches, including shimmering or blocky textures.
- Edge Detection Problems: DLSS can have problems when there are many fine details near edges, which can lead to flickering or weird visual artifacts when trying to reconstruct them.
Training Data and AI Limitations
Remember, DLSS is like a student who learned from a teacher (training data). If the teacher didn’t show the student all the different types of situations it might encounter, the student might make mistakes. The quality of the “training data” used for the AI is critical, and limitations here can affect performance.
- Limited Training Scenarios: DLSS might be well-trained for common scenes but struggle with unique or rare visual scenarios. It may make poor assumptions leading to unexpected visual artifacts.
- Not enough Specific Game Training: DLSS is typically trained on very general scenes, so it can suffer when it encounters very specific assets in a particular game and struggle to process them.
- Changing Algorithms: As NVIDIA updates its DLSS algorithm, there may be growing pains. A newer algorithm might not work well with a specific game which was developed with an earlier version of the algorithm in mind.
Hardware and Driver Compatibility
Sometimes, the problem isn’t with the game or DLSS, but with your computer. Old or incorrect drivers and hardware that is simply not up to the task will hurt the quality of the image and experience.
- Outdated Drivers: If you don’t update your graphics card drivers, you won’t get the latest optimizations for DLSS. This could result in worse performance and poor image quality.
- Insufficient Hardware: Older hardware, while technically compatible with DLSS, might not be able to process it effectively.
- Insufficient VRAM: When you use a higher quality DLSS setting, or play at a higher resolution, this can use up more video memory (VRAM). If you don’t have enough, this can impact performance negatively and lead to visual artifacts.
Specific Examples of Games
We’ve talked about some general reasons why DLSS can look bad. Now, let’s talk about some specific examples. Note, that not every one will have these same experiences, as individual setup and personal preferences can greatly vary the visual experience.
- Poor Implementation: Some older games which received a DLSS patch, did not have very good integration, which has resulted in visual anomalies.
- Incorrect Presets: Often, incorrect presets in the game can lead to a poorer image quality. An example of this, is a game that defaults to ‘performance’ mode when starting out, resulting in a poor first impression.
- Specific Game Engine Issues: Some older game engines don’t mesh well with DLSS, which can lead to issues. A particular game might be using an old version of an engine, which can negatively impact the DLSS experience.
Competing Technologies: DLAA and FSR
It is also important to realize that other similar technologies exist. DLSS isn’t the only option out there. DLAA and FSR are alternatives that also try to improve image quality or performance. They each have their strengths and weaknesses.
- DLAA (Deep Learning Anti-Aliasing): DLAA focuses on improving the image quality but doesn’t have an upscaling function and is therefore not useful for improving performance like DLSS. Some users prefer its sharper, more detailed image.
- FSR (FidelityFX Super Resolution): FSR is an alternative to DLSS developed by AMD, and like DLSS is an upscaling technique that focuses on improving performance while maintaining good visuals. It often has a slightly different visual signature, some users prefer this to DLSS.
It is important to know your options, and see which one works best for you! The visual look of these techniques are often highly subjective.
Finding the Sweet Spot
So, when DLSS looks bad, it’s usually a combination of these factors, and the specific way DLSS is implemented into a specific game. The good news is that developers are constantly learning and improving how they use DLSS. The technology itself also gets better with time. DLSS performance and quality has significantly improved since it was first introduced, and that trend is likely to continue. As hardware gets better and developers get more experience working with AI upscaling, DLSS will hopefully get even more reliable. The main takeaway here is that getting the best results often requires a delicate balance, and unfortunately sometimes the implementation falls short and does not live up to its promise.
DLSS, and similar technologies are a complex field, but it is always good to understand these basics to better appreciate the inner workings of how your games are rendered!
Is DLSS Ruining Games?
Final Thoughts
Ultimately, DLSS’s visual quality varies greatly. The implementation directly affects output. Poor implementation, insufficient training data, or aggressive upscaling settings cause blurry or artifact-filled images.
Thus, why does dlss look so bad in some games? Developers must spend more time optimizing the technology. We need better per-game tuning and fine-tuned algorithms to yield high quality upscaling instead of poor results, this is must.
