Denuvo reportedly cutting DMC5 framerates by up to 25% - Sensationalist Headlines

https://www.pcgamesn.com/devil-may-cry-5/dmc-5-fps In a post on the game’s Steam community page, one player discovered a branch likely used by the QA team to test the game, but had accidentally been left public. Capcom has since deleted the branch, but players can still access the Denuvo-free file in the Steam console. From there, it can be downloaded, and then used to overwrite the pre-existing version of the file. The author of the Steam post claims that the downloading the file has increased the game’s frame rate from 74 FPS to 93, an increase of 19, or around 25%. They also state that the new file has kept their FPS more stable than before. Both Denuvo and several major publishers have previously claimed that the anti-piracy software does not impact game performance. While Denuvo workarounds are often attached to videogame piracy, this time around that’s not particularly likely to be the case. While Devil May Cry 5 does appear to have already been cracked, a pirated version would already be free of Denuvo’s protection. In order to take advantage of this particular trick, you would need to have actually purchased the game through Steam.

So even if I have to buy it, I have to pirate it to play it at a good framerate? This is some galaxy brain DRM right here.

Depends whether crackers bothered to removes DENUVO instead of just making workaround for running game with it

I was watching yesterday someone play few missions early and I was restraining myself hard not to impulse buy it on the spot. I've been still thinking about it this whole day and then this shit is revealed. Thanks I guess for saving me money.

While I am strongly opposed to invasive anti-piracy measures like Denuvo, I always have to view these claims with high skepticism because there are a lot of false claims that get circulated about it's impact on performance or other system elements like the hard drive.

https://www.youtube.com/watch?v=1VpWKwIjwLk

While Denuvo workarounds are often attached to videogame piracy, this time around that’s not particularly likely to be the case. While Devil May Cry 5 does appear to have already been cracked, a pirated version would already be free of Denuvo’s protection. In order to take advantage of this particular trick, you would need to have actually purchased the game through Steam. That article states this but I seriously doubt this wasn't in some way used to crack the game, it's incredibly rare to have a day 1 crack for a Denuvo game and I don't think this is a coincidence.

can only assume that its a shoddy implementation of the drm, since other games using it had their fair share of investigations on the matter, but all for the most part came out empty handed yikes though, pirates get to play the game for free and get substantially better performance, even worse, the game got cracked essentially on day 1 all the negatives of Denuvo seems to have manifested in this release

AFAIK the day 1 "crack" is just the DRM free exe, same thing happened with FFXV at launch too, no need to make a crack when the company just hands you the keys.

Even if we give you the benefit of the doubt, it's weird that an older revision of the executable lighter and much better performance wise, so is very posible that this difference is thanks to the Denuvo implementation. Now the big question is: How much the performance impact depends on the implementation and how much on Denuvo itself? This is probably the first time that we have such an easy "Before-After Denuvo", so I hope this gives us more information.

Yeah this is what it is apparently, the DRMless .exe file was on the Steam branch. Man what a fuck up, all that money wasted on buying and implementing Denuvo, someone's gonna get their ears lit.

Same thing happened to DOOM too. Afaik the demo executable worked for the final game content, and it wasn't protected. Correct. Denuvo really doesn't cause a significant frame drop unless its implemented incorrectly and without care. You're not supposed to let Denuvo protect the most performance-critical code, so the devs didn't do their job here. Of course, I guess this news is just going to get used to ridicule Denuvo, but it's not really their fault in this case.

Oh that's a good fact! I have read in the past that the game developers give Denuvo the game project and some private keys, they decompile-recompile the game and give the devs the final result. I always wonder how true that was.

Sounds badly paraphrased. As far as I know, this is the process: Developers buy a license. Denuvo asks for your executable and debug database (.exe and .pdb). They send back a custom protector executable that can be used during the build process. Developers compile their game as usual, as a final step they put it through the protector executable. The protector sends the executable to the Denuvo servers (encrypted), here it is automatically disassembled and random portions of instructions are converted into equivalent encrypted bytecode that is run by a Denuvo VM (which is embedded into the game executable). The patched executable and debug database return. In the case of bad performance (typically reported by QA - who are supposed to compare unprotected and protected builds): Bad performance is reported to Denuvo. They ask you to provide a list of functions that should be exempt from the virtualization process. I believe they also offer to profile and find the best candidates for exemption (we just provided a list of functions). The protection server is changed to take exemptions into account, subsequent builds should run better. There's probably a little more to it, but that's the gist of it.

Most denuvo cracks simply emulate the checks so just like with IJ2/H2 the issues will never be gone until and if devs remove it

That clarifies a lot of my doubts. Thanks for the explanation!

All Denuvo cracks* The only time Denuvo has been actually removed is when they used a workaround (think demo exe being used to run retail content like Final Fantasy) or the devs actually removing it.

I had to refund DMC5 because of unplayable levels of stuttering every second and stuttering in cutscenes desyncing audio. Same problem was in RE2 and was never fixed. Have to wait until Wednesday to get the console version. Thanks Capcom. Denuvo had nothing to do with it either, people claim the denuvo-less exe doesn't fix it.

I never had this issue in RE2 or RE7 but games like Metro Last Light and Metro Exodus I had a stuttering problem all the time as the game tries to load assets. Usually it qas most rampant when I just started up the game. Things like firing a gun/animations would stutter the game for a second while it tries to find the audio to play.

friend that bought tried the no denuvo exe and didn't have an FPS counter to tell, but noticed a vast difference in load time

Whats your build , I'm running a pretty mid range pc and can get 100+ fps most of the time without stuttering

I have a good framerate aside from the stuttering. It's not a graphics card issue. Message board topics about this say it happens to people with all different ranges of specs.

Denuvo pretty much always increases load times by a lot because thats when the DRM is active basically.

Digital Foundry's Richard Leadbetter put this to the test. tl;dr, there does seem to be a performance hit, but it's very small (about 7% on average in this case) and something those with decent CPUs may not even notice due to the game being more GPU bound. We don't know if this is because of a lack of Denuvo, or if there are other changes in the executable that improve performance (the EXE is from a QA branch, after all, so it could come with some tweaks and fixes). More testing is required essentially. Still, if it's something that increases performance a bit and can be obtained perfectly legally, why not give it a go?

Some games aren't affected, performance wise, by Denuvo. Most are. And I believe all of them have seriously increased load times with it, as well

So are you going to provide us some empirical evidence of this? Or is it a "gut feeling"? The way that we understand how Denuvo works (industry secrets and all that) should not impact performance in a noticeable way if implemented competently. Load times shouldn't increase drastically either, as it isn't encrypting content. It's mangling core game functions. Decryption when you have the correct key isn't a slow process. A minor performance impact is to be expected, a few frames or an extra second of loading at worst. If it's anything worse than that, that's on the developer for not doing their job right. The games that have had Denuvo removed largely didn't see a massive change in their performance. DOOM (2016) being one of the most noteworthy cases of removal.

They are. https://youtu.be/ByfLg9wGB4o Some cases have double the loading times, most have about 25% increase and some show no significant increase.

So in those cases, I'm guessing Denuvo is being applied to filesystem code. Developers should exempt filesystem code from being protected by Denuvo and then it'd lead to very little impact.

Thats the thing though, devs should do that but a lot of Denuvo implementations are flawed and a good amount of those games have increased load times, bigger performance impacts or other issues caused by that. Denuvo has a hand in all those applications, so they also take part of the blame in my opinion.

Well I disagree. It's a tool, they're not responsible for it being used optimally. Developers should be held fully responsible. Like I said, they should compare unprotected and protected builds and figure out what's impacted.