Sure, but my point isn’t based on the video at all. With the video or without the video, there’s no technical basis for the claims made by Cryoutilities.

CU but it feels smoother with it installed

That’s how a placebo works. It truly does feel smoother, better, etc.

But when you measure it, you realize it’s just your mind playing tricks on you.

It’s not that it’s “no longer necessary” - it was never useful in the first place.

The UNA framebuffer setting was always a severe misinterpretation by this subreddit (and purposely not fully explained by Cryo) that only matters for Windows users.

The thousand page file tweaks were literally placebos. And the mere idea that Valve, the people working on low level drivers and Proton, wouldn’t know how to properly setup page files for their gaming device is laughable.

The only setting that had the potential of mattering was the increased swap, but thanks to the increased complexity of a shared memory architecture this is extremely circunstancial and really dependent on what the game is doing. Most of the time, the “improvement” was preventing a single frame time spike every once in a while.

Now you might think “jeez, but why are people so angry when talking about this?” because whenever somebody questioned the claims of this tool, or more importantly, the inner workings of the took, this subreddit would downvote them to oblivion. It’s bizarre that a literal placebo became almost worshipped in this place. And boy the clickbaity “make HOGWARTS LEGACY amazing” Cryo videos that took 30 minutes to show, after all the possible tweaks, margin of error differences that were explained as “oh no, I can’t measure any real improvement but trust me it feels better!”

If somebody is showing you supposed results of their tool, but also claiming they can’t really be measured by any real metric and must be “felt”… they’re gaslighting you.

If you go back 6 years, you’d see something similar on Android communities. Somebody would make a post about how you could change some build.prop parameters and delete the battery calibration file (which Google engineers explained was not even used) and that would make your performance significantly better! Of course, the person posting this amazing tip was also promoting their YouTube channel, and their PayPal tips jar. Rings a bell?

Will using the tool somehow harm your device? No, it won’t. Will it improve performance? No, it won’t. Can we finally put this whole ordeal to rest?

At 60Hz, each frame is shown for ~16ms, while at 70Hz each frame gets displayed for ~14ms.

2ms is not nothing… But it’s below the treshold for human perception. While LCD screens have a little bit of a blur as the time for each subpixel to transition is somewhat long, the change from 60Hz to 70Hz amounts to nothing but a placebo.

This post, and the comments on the previous post, are worded in a bizarre way that somehow tries to imply Valve is attempting to hide the issue, or purposely not fixing it.

At least 3 Valve developers already commented about it - several people already provided clean captures of the issue. They already claimed to be working on it.

If you look at the thread on GitHub, there’s also a lot of noise, unnecessary comments and redundant videos. So yes, closing the issue and fixing it internally is a good move.

What else do you expect them to do? Broadcast an apology on live television? Replace your Steam Deck with a gold bar? Personally send a DM to every Redditor?

The extremely noisy greys are also pretty annoying.

This, compounded with the subpixel matrix, at this resolution and screen size, creates the illusion the display is covered by a thin layer of dust, or that the outlines of objects (or thin text) are missing dots.

I haven’t got either unit to test, but in the past, Samsung has shipped some pretty bad panels too. It’s less about the manufacturer and more about how much Valve is paying and how much variance they tolerate.

“BRICKED DEVICES” on all caps and bold font.

You can’t brick a Deck by an update stalling at “x%”, you can simply boot an external .iso and reimage the OS.

Cool beans

you don’t have a clue what you’re talking about.

Feel free to pretend what you just wrote is true and overclock your Samsung RAM.

In fact, I encourage you to do so without a backup.

If you have an LCD model with non-Samsung memory, or know somebody with one, you can actually enable a memory overclock and test!

Though it’s the most unstable and easily brickable of all Steam Deck overclocks, so I wouldn’t, personally. If you do have Samsung RAM… don’t even think about it.

The Deck is heavily memory bandwidth limited. The CPU improvements from the latest patch will probably give an improvement, but it will not be the full ~20% observed by Digital Foundry.

It looks amazing, and I hope they keep this new design.

Sometimes Reddit freaks out about gaussian blur on user interfaces, but I think it looks beautiful and it’s running smoothly.