FRAPS stole my frames!
HTML-код
- Опубликовано: 11 сен 2024
- Well that didn't go as planned. I discovered a performance-draining issue with the program FRAPS, and some new tests later, we see what the effects were!
Platform specs:
Athlon XP-M 2400+ @ 2.4GHz
512MB PC3200 @ 266MHz
Asus A7S333 (SiS 745)
Aureal Vortex 2
Western Digital WD800 80GB HDD
Windows 98SE Unofficial SP3
Drivers:
Rage Fury MAXX: A6.40CD15
Voodoo5 5500: v1.04.01
GeForce 256 32MB: FW 45.23
GeForce DDR 32MB: FW 45.23
Radeon DDR 32MB: Cat 4.1
Intro Animation By Ken Gruca Jr - Inquire at kjgruca@gmail.com!
○Join the Pixel Talk Discord server! / discord
○Instagram! / pixelpipes
○Follow me on Twitter! @pixelpipes
○Donate via Paypal: @pixelpipes
![MISSING: SLIM SHADY [Expanded Mourner’s Edition Trailer]](http://i.ytimg.com/vi/xh8qgHrOO1g/mqdefault.jpg)
Hello! Nathan here with PixelPipes, and just a heads up, this video is a follow-up to my previous Card Battles video featuring the Voodoo5 5500 versus the Rage Fury MAXX. Please watch that first or you won't understand what's happening here:
=> ruclips.net/video/4be8HYOGrl4/видео.html
For those that just want the numbers and don't want to hear any explanation, go to the timestamps:
11:15 for redone tests
21:13 for new tests with additional bonus cards
Thank you all for being patient with me! We'll be back on track with the next video!
Hey, Nathan very interesting video. Luckily l'm old enough to remember and actually use fraps back in the early 2000's. My 'awsome' 2003 was an AMD Duron 1Ghz, 256 MB of DDR (upgraded from 128 megs of SD RAM) and the 'legendary' GeForce Fx 5200 8x AGP edition. 😅 I was told that by using Fraps the fps difference could be anything from 5 to 15 fps. Since the 5200 is not a good GPU for gaming to begin with, it was obvious when games were playable and above 30 fps, or producing a 15 fps slideshow while using Fraps. In my opinion the impact is caused by windows 98 itself, which wasn't built for multi tasking to begin with. I remember game instructions from this era reminding to close all tabs and folder before launching the game. Today we're used to leave the browser open, or having steam, uplay, origin, MSI Afterburner, Shadowplay , Discord or twitch running in the background. I usually play some single player game while downloading something else on steam with a minimal performance hit. Back in the day my performance tanked when l was trying to download a 3mb mp3 song and try to play some games at the same time. So l think it's the operating system.
You do you, don't worry what people on the internet think, let the numbers show for themselves. You are doing great, keep up the good work. Nice to see someone spend money, or the time to get donations on hard to find items. Voodoo 5's are very expensive to just throw money at, as are most the cards you have, even if they are donated, its not something everyone has.
Fraps has been known for its overhead since the 2000s, i find it more remarkable it is used at all these days.
people just use msi afterburner
Tools like Fraps influence the FPS has been known problem for a long time
MrJ0mmy TBH Afterburner kinda sucks for benchmarking. It doesn’t have a time cutoff and it doesn’t provide logs for frame time data. OCAT & FrameView are both much better for benchmarking newer hardware.
It was known that using it for recording would tank fps far more than even other recording software. It absolutely was not commonly known for lowering fps just having it running effectively doing nothing at all.
It might be known by some people, but I imagine the common knowledge of the recording feature being so bad has muddied the information and become mixed up. Especially as he even says it is not common in all circumstances!
@@wyterabitt2149 100% correct bud. It is well known that FRAPS causes FPS loss when recording. It is not known for causing large FPS loss like this when just running.
I think Nathan is correct, this appears to be a very specific bug that is only effecting the Voodoo card and only in dual GPU mode, as he pointed out. Viewers here are getting confused with FRAPS behavior like you stated, and are also forgetting what Nathan pointed out.
This is just a random bug that has crept up, the type that can and does happen to anybody. These things happen to the best of us. This issue was unpredictable because it is not common behavior for FRAPS.
Hey Nathan,
I'm sure i speak for other people but often times these "mistakes" can end up being even more interesting than the original video content. I think this was awesome and I appreciate you going back and trying to understand exactly what happened. This is the kind of content that I think there just isn't enough of.
I'm just glad to see you making vids again tbh
What this really tells me is that we need a new tool for old Windows that's more reliable than FRAPS. Everyone knows that FRAPS has issues with overhead and such, but if you're testing old software/hardware on Windows 98 or something your options are somewhat limited.
For someone like me who was pretty late to the retro party (and would be lost without RUclips help), it is seeing an entire journey, mistakes and all, where I learn the most. Like you, I tended to just let fraps run on my 98se machine (Voodoo3 w 600mhz coppermine). I actually believed the fraps FAQ where it claimed the performance penalty was minimal. Gonna try my own benchmarks now. Big thanks for the update bro!!
Probably minimal for mid-late 2000s standards...which translates to _big_ on 90s hardware.
Perfect reaction to your discovery. Of course people in the comments go all Dothraki like "It is known...!". Wise-assing about some supposed overhead that FRAPS has AAAAALWAYS had (I for one only remember it being very heavy when capturing video).
But what you found is actually much more specific. If it was an inherent overhead of Fraps, then it would affect all your results the same way, but it did affect the results differently! So that in itself is already an interesting discovery. I certainly didn't think Fraps would have a preference in regards to GPU architecture or drivers (and quite honestly, it would be very interesting to figure out why it impacts the Voodoo so much).
The new Bench-results make it all the more intriguing... Somewhere there is some weird bottleneck.
i agree ,so tired of these pansies saying "ooohh it's over head was known X number of years ago" ... BULLSHIT i call. I never heard about it's overhead , and i used it up to dx 11 become common (because it has issues with dx 11) . and actually if you look it up in the late 90's early 2000's it DID NOT have an overhead issue. that issue started happening in the version that updated for DX9-c. so i know these guys are full of shit. as dx9-c didnt' exist till like 2003/
Fantastic video. I think you're doing a great job. Everyone goofs up; it's part of being human. It's obvious that you're very sincere about rectifying any such mistakes with more accurate info, and that matters most in the end.
fraps has been known for a long time to have a performance decrease just having it run not doing anything.
@Lassi Kinnunen kind of hard isint in rivatuners vocabulary
Okay, but how do you get rivatuner to do it in Windows 98? Because I'm spoiled by msi afterburner, and just got my 98 project build playing nice finally.
It's been known to decrease performance slightly as any monitoring tool does (Rivatuner or whatever modern equivalent you use included) simply because the act of monitoring a bit of hardware uses cycles to get that information, *this* is something different and shows that some cards/APIs don't play nicely with it.
I can't believe I missed this whole ordeal. Lots of arm chair quarterbacks out there. Don't sweat it Nathan.
As far as im aware, you are the best retro channel on RUclips, i was thinking yesterday that i dont understand how you dont have over 100k subs, its all good Nathan, we are all human.
D3D being faster in UT on the V5 5500 is almost guaranteed to be because of your very fast CPU. If you run the same benchmark on a period correct CPU, The glide score would very likely be much faster. The biggest strength of Glide was the low CPU overhead.As for asshats on the internet, best to ignore them. Your videos are some of the best available on the topic and a huge influence on my collecting and testing methodology.
Keep on keeping on, youtube needs more quality retro content!
Would love to see a Voodoo 3 3500 and TnT2 Ultra together with those scores towards the end of the video, to anchor them compared to the previous generation.
Very interesting, going to take this into account when benching
Dang dude, what an awesome video! Not only do I get to see these charts again, but now it makes way more sense based on your empirical data :).
You’ve done a good job with this mea culpa. As someone who didn’t have any follow up interaction with the previous video, I really appreciate your corrections and discussion about them and what went wrong. Owning the mistake and making good says much more about you and the channel than the original mistake ever could.
Achieving perfect performance on these old video cards is just pain. Even if you track down the best driver for your video card that works best with your single favorite game, you can still have issues because of motherboard/chipset specific hardware/driver problems. Or even because of some other random PCI card that causes slowdowns. Also I'd like to suggest to ignore h0stile people in retro community. And I'm very grateful to you (and others) for doing all those retro benchmarking/exploring stuff, both with and without mistakes :-). It is very under-appreciated routine work, at which I would've lasted a week at best.
I knew some data were off with Voodoo5 5500, I still remember back in the days, many benchmark reviews showed that Voodoo5 had the capability to compete with GeForce 2 GTS, though it might still lost to GF2GTS, but not too much.
it traded blows with the geforce 2 gts , but the geforce 2 ultra trounced the shit out of it. which is why i went geforce 2 ultra on my amd thunderbird build. back then the "ultra" really meant a shit ton in performance, as teh geforce 2 ultra was about 30-40% faster than the gts. not like today where the Ti moniker might see 15% speed over the non ti versions.
god damn i miss that computer. at the time it was so unbeleivably fast compared to every other computer i had built or owned at the time. 850 mhz thunderbird core slot a processor , 312 megs of system ram (ol sdr ram 333), and a geforce 2 ultra. it was a beast of a machine for it's time.
As many people have stated here, don't sweat it. There are so many variables that can come into play between systems, even with the same hardware, even down to what updates have been installed on the operating system. We have to keep in mind that no one is saving the world by having more accurate benchmarking results. (Not everyone can get as picky as Steve from G.N...nor do we have to be.)
I've caught only a couple of your videos in the past that showed up in my recommended feed. For whatever reason, those weren't compelling enough for me to subscribe. However, the transparency and and candor you display here was what finally got me to click that subscribe button. We need more people like you here on RUclips. There's a big problem on this platform of people covering up their mistakes and denying their audience the opportunity to learn something from it. Often it's far more helpful to know what obstacles a person encountered than just trying to replicate their successes. Thank you.
Hey Nathan, you’re right: Episode 1 Racer is an undeniable classic. Eeee kuuntaa!!
I think people don’t realize retro hardware benchmarking is harder than the latest hardware. You can’t just use the latest build of windows 10, the most recent bios revisions and the latest drivers available to download. Who else has to tell themselves ok I’m going to benchmark this card that’s more than a decade older than the software I need to use to benchmark it and i need to check the my ARCHIVES for the best drivers!
Indeed, well put.
Fraps stole my pixels and.. my pipes..)
Nathan as long you do your best there is no problem, I enjoy your channel.
I love your channel and your vids! Keep it up! Cheers
i discovered that fraps issue a long time ago. But more than the number, i came to your channel for the hardware itself... Numbers are justs numbers...
I do recall fraps having some overhead, but it has been so many years that it never dawned on me. And your initial results seemed ok to me. But I would not have ever thought the V5 was under performing as I use it on a lower end CPU where its performance is probably more in line with your original numbers. (P3-733EB/P3B-F1 CL2 ram). And my experience with fraps overhead was when I was using lower cpu's as well so I subconsciously wouldn't have suspected it on your setup. Higher end cpu's in my collection generally run XP with newer versions of Fraps.
In fact, way back in the day I was always used to lower numbers because I had ran benchmarks on my daily driver which was always running software which could affect performance. /shrugs
But this is an excellent video, I really do appreciate all your effort put in to this. And this has been one of my favorite retro channels so far.
16:20
Am I reading that right, the _PCI_ version of the Voodoo 5 is 10fps faster than the AGP variant? How in the hell does that work?
Huh, very interesting. I knew Fraps could impact performance although I'd put it down to a placebo effect as I couldn't find any difference in benchmarks and had no idea it was that card/game specific. (ie. I thought that you either had a tiny performance loss across the board with it running at worst or that I was just "feeling" one where there wasn't one, not a difference that was sometimes big on one card but small on another or big on one game but small in another as your results show) Guess it may not have been a placebo effect...Mind you, this was on GeForce 6800 Series, Radeon x1600 series and GeForce 9600/9800 series with XP and Vista so who knows? I might do some testing with my XP Retro rig if I remember to next time I'm using it; it has a 9800GT in it.
Also, don't sweat the incorrect results in the previous video. It happens to everyone and this is absolutely a textbook demonstration of how to handle such a situation: With respect for the audience, an explanation of what the problem is and how it was discovered, new fixed results *and* something extra for us all to talk about to boot!
Love the intro!
Congratulations! This is why we love your videos!
My guess (regarding the odd Rage Fury Maxx performance bumps) is driver eccentricities that don't handle the bloat and bugs inherent in an old W98se install all that well. The AFR on the Rage Fury Maxx is done in software only, so running those drivers on a fresh install might have allowed smoother operation, leading to higher FPS. The single chip bumps were fairly modest, what one would expect from a fresh install.
I've been hearing of this issue with Fraps for years, decades really. You'd think they would have worked out that issue by now
Benchmarking is an attempt to assess the performance of a system... for a specific task! :) The last part has its importance. Great video though! Keep it up!
Hey, Nathan. I've really been enjoying your content. I think you handled this situation really well. Keep up the good work, I'm gonna keep watching!
Keep up the good work, your videos are awesome! You can't please all the people, but don't stop doing what you are doing, retro hardware IS cool!
Don't worry, it's all okay :)
Thanks! I remember reading back then that fraps could affect your frame rates, but didn't knew that it could affect two cards with such a difference in the same system, even more being it so powerful! I wonder how much it would affect to people using Tualatins, P4 and Athlon XP and slower systems!
If all of the benchmarks being compared were done with FRAPS, then it's still valid. Without a DigitalFoundary style capture setup, you don't have a huge number of options for frame counting on old PCs
Good job Nathan as always! Your commitment to the retro pc community is unquestionable and there is no need to apologise for simple mistakes that happen all the time, especially with 20 year old software.
I might have forgiven you but boy oh boy, voodoo gods are angry at you :O BTW. It would be nice to somehow test how bad MAXX input lag is in games. With high speed camera or something. Same for when you will test SLI/CF configurations. Have a nice voodooritic day :)
Admitting to and learning from your mistakes makes you *more* of an expert, not less.
I remember Fraps and just how much improvement it really would need. Stuff we have nowadays, benchmarks, framerate counters, hardware video encoding stuff like that. All had to be done with fraps in software.
So yeah, running fraps would mean taking away fresources.
Of course, the video files it could create were huge. Because, AVI with almost no compression. And at that time, HDD's where not big so... trouble.
Nowadays we have it much better.
It'd be nice if you could show which driver version you used in the charts under the card name. That would help a ton for people wanting to replicate your results!
Since you're already capturing the computers why don't you try using the program TRDROP? You might be limited by the max fps of your capture card but it might be okay for a 60fps 120fps 240fps baseline. I've done an old old video on fraps stealing frames in No Man's Sky. I also did one with afterburner doing the same. Both had to do with just drawing the overlay.
On an aside, this kind of stuff is why I've been avoiding doing Gaming benchmarks for my website, The Chip Collective. Multiple passes of synthetic CPU benchmarks with a sound method is just way more comfortable lol. But I've been benchmarking a 2070 super the past 2 weeks and it's hell lol. 5 passes of every benchmark in my suite. A big Excel sheet.
I'm glad you recognized your error though. Keep on keeping on my dude!
I knew something was wrong when I saw the results for UT99, it's impossible for V5 5500 to lose vs Fury Maxx, V5 5500 will beat (in glide) vs opengl/d3d other cards even a GF2 Ultra or a Radeon 7500.
another thing that happens is that by using overpowered cpus you're actually alleviating some of the voodoo5's shortcomings, such as lack of T&L, which did shift some of the load from the cpu to the gpu, helping frame rates on period-appropriate cpus. but since you use a much more powerful cpu, it handles the T&L via software just fine so the voodoo5 isn't hurt by the lack of hardware acceleration.
@PixelPipes - Nathan nice Video 👍🏻 would be nice if you could do a video about the S3 and S4 cards and their special highres texture compression compared to other cards. In some way these cards where way ahead of its time 😁
It's known for ages, why do you think fraps was mainly put aside?
When I saw the video I just assumed you were testing them both, isolated... to see which was better of the 2 (discount fraps on both and you'll have the answer)
I've always heard "Add a few FPS due to FRAPS", assuming it was robbing a few resources, thus skewing the projected results just a tad from what the user would typically see without FRAPS running. But I also assumed this was common knowledge too. I haven't had a 3dFX card since my Voodoo4 4500 so I had no horse in the race. I came for the entertainment and was not disappointed. Keep em coming and forget the haters.
To be fair here, I would never in my life have expected that Fraps could have a significant impact on performance. Even as a former Voodoo5 5500 owner with a even more Overkill CPU your results never felt wrong to me, but then again I never Benchmarked my card, but the results simply didn't look off to me.
when i was using fraps in first half of 2000's on my xp and 98se machines the missing 2-5 frames was standard thing so i stoped to use fraps all together so yeah i confimr there was a problem with it but i never tryed to find out what it was. thanks for the video
what you need is a 98-compatible re-implementation of FCAT. Use a tool to draw color bars over the image and do external capture and use a program to count the color bars.
Great channel and thanks for all the hard work! If anyone feels the urge to criticize they should at least have some data to back it up. Too many biased and undocumented opinions in this world...
I find it weird how he's just now (in May) finding out Fraps hurts performance just having it open... I knew about this when I used it as a kid on a Pentium 4 with the junk integrated graphics...
I've been watching through some of your videos and was curious, what GPUs and CPUs are you looking for your collection? I honestly would have to look again to be sure (that I still have them) but I have some older hardware sitting around from the late 90s and early 00s I wouldn't mind parting with.
Guess it's time to go full Digital Foundry method for checking the fps and frame times.
I was *FRAMED*
Yeah I could of told you FRAPS has a larger performance hit than they lead one to believe always has. They try to keep it hidden from people considered it's popularity but it's still there and can be very measurable in some titles and benchmarks. Also the performance impact while there shouldn't have been that large on the Voodoo 5 unless it's preventing scaling because it doesn't know how to handle the dual chip setup and forces it to use one.
What I think would be fun is a shootout between a card of this era and a modern GPU :p. Problem is they won't run on the same system motherboard or processor :(.
Me this was the Time frame I started to use the Intergraph / 3DLabs Cards Workstation Cards for my Main System..
The new benchmarking confirms that the Fury Maxx has been historicly undervalued, despite whoever likes it, despite whoever disagrees. Thank you Nathan for your honesty and excellence
Man you need antivirus, system booster, ram cleaner, ccleaner monitoring, another antivirus and windows defender in background if you want speedy PC. Dont forget win update.
What about testing the Bitchin' Fast 3D 2000 with Bungholio Marks? :)
Fraps is sh!t mate. I try to use msi afterburner or the built in stat's on newer AMD cards. Notice I call them AMD and not ATi. Fraps was giving me issues because it drains resources.
Oh and BTW nice to see you again. I hope you find a better solution soon.
Wow, I was still using a Pentium 4 when I stopped using fraps because of the fps issue. It became a big controversy back then. Really weird to see someone find this out in 2020. I guess we all make mistakes. No biggie.
How is it weird Pogo?
Personally when i used TnT2 and fraps back in days i discovered when benchmarking Quake 2 with fraps running in background the FPS was lower.
Latter i always think that i was limited by P-2 333mhz.
Use bandicam mate if you must go the software route. It's very light weight.
Try using a old version of riva tuner. It's monitoring system is Wat became afterburner an if for sure has less overhead and more features too
that silver card on the bottom shelve, with the two small fans(?), I remember seeing it in a magazine for like 2k$ when I was a kid, thinking how powerful it must be. What card is it?
That's a Leadtek FX 5900 Ultra
Hey man, mistakes were made but more importantly, mistakes were discovered.
What really matters is the reaction and how you fixes your mistakes.
Retro hardware testing these days must be quite harder in some ways than testing the latest tech. You are running old hardware that might not perform as it was new, you are using old software and drivers that had issues fixed later (but dropped support for the old stuff).
If someone is watching retro hardware reviews without keeping that in it's mind, he's widely mistaken.
Why are you making excuses like that? Yes, you made a mistake, but your reviews are one of the best in retro scene. Do NOT make like that anymore, you are cool man and do not hear any amgry stuped “sofa experts”
Real enthusiasts know when to admit they've made a mistake
Sorry, this is not enough Nathan. You have LIED to us. You gave me hope for my MAXX.
_remembers your lie about the 3000+_
You made a mistake.
But who does not?
IMO the real mistake is not admitting and not learning from a mistake.
I am looking at you The Verge.
About the Unreal tournament score, how about "detail texturing" ? I know that by default, its enabled in glide but disabled in D3D while using the original render libraries and ini settings. If you ran with it enabled in glide and disabled in D3D it can easily explain why D3D was faster, but come on, having that off is such eye-sore on any close-up textures that you would have surely noticed. Glide generally helps a lot in CPU limited scenarios, once you become purely fill-rate limited it has little to offer.
There could very well be more effects going on in Glide mode. I only looked at the settings through the game menu but there might be more unexposed settings in the cfg files.
@@PixelPipes Yeah the game menu does not have much, you have to look in the unrealtournament.ini file. Detail texturing is easy to spot, just look close to any surface texture and if its a low-ress mess its off, if it has another layer of finer texture on it that appears only when you look from closer distance, then its ON. By default its ON only in glide, other APIs and drivers back in the day where too buggy for them to risk it ON globally.
what about rivatuner?
Yeah, it's known that FRAPS affects performance.
However I didn't know that the difference is that big! :D Great follow-up video! ^_^
I heard software can misrep numbers so I would blame software issues too bad their aren't phy gauges that could be hooked up to measure fps but it not a real thing lol
Eh, the people who mind don't matter, and the people who matter don't mind. Everyone has goofups and owning up to them makes you more mature anyways. Don't sweat it, it happens. You could even reason that it caused you to make this informative follow up video. I didn't know about this beforehand and now I do.
Send Fraps to jail officer...
ah FRAPS the software that hogs your resources, you would think by now there would be a open source tool to use for recording FPS on all of the operating systems.
Dude. I have been a retro gamer before it was called retro. i was a gamer when most you tubers on the subject were either yet to be born or begging their 'rents for a voodoo card or a 2400 baud modem to get on the bbs's with. Don't sweat the dumb stuff, people don't know everything and wiki is not always accurate. Thing is, don't worry about screwing up and when ya do don't worry about trolls, and expect to get jibes. However, never ever diss a component unless you been there and know the pain of why a component sucks.
Wait, how do you record Glide FPS then?
Just have to hope the game has a feature that will do it for you! Although I hear there is a new tool out for that purpose...
@@PixelPipes Ooh?
Started playing rage 2 on my channel not a bad game
i really don't think of these (cards) as retro. to me retro is still 8 bit and 16 bit gaming like what i had when i was a kid , these cards came out when i was an adult (i was 19-20 at that time). to me they are things i still remember reading back when building my pc's then, I remember it like it was yesterday.
The sad reality of aging! 😆
@@PixelPipes don't laughy face me dude .. you said you were 31 , you'll be where i'm at in 10 more years ... it goes REAL fast. (im 41 now ).LOL
Oh god I know... It's already started for me! People playing games these days don't even know what pixel pipes are. It's all CUDA cores, etc to them! Guess that's why I make these videos, so I can sit on my porch and yell at them about "the good ol days". 😜
@@PixelPipes well you do a good job with presenting your videos. your "retro" tech videos are really the only ones i can sit through all the way. just finished your video on those ol' 3d tech demos it was awesome. even learned a few things i missed back in the day.
@@DenverStarkey Thank you!
fraps is poop
The Steam overlay (to display FPS) does the same... That POS made me lose ~30FPS in Doom Eternal
Yip. Doom Eternal is known to have overlay issues. The devs did confirm as well. Which is weird, because DOOM 2016 didn't have the same issues, at least not for me.
I am sure there is some id tech engine reason for Doom Eternals overlay issues, but I have not investigated it online for an possible answer yet.
Bethesda ruined doom eternal. That's all that needs to be said. 2016 was boring, but mostly bug free at launch. Eternal is a hot glitchy mess this side of fallout 76.
@@ZeroHourProductions407 not to forget Eternal running like ass on pre pascal nvidia cards while its AMD equivalence runs times better
@@3dcomrade True. But that is down to Pascals inferior DX12/Vulcan architecture, which is not as good or complete as AMDs or NVIDIAs latest Turing architecture. What we are seeing is AMD and Turing gaining performance and not Pascal losing performance.
Pascal is well known to have inferior or incomplete DX12/Vulcan hardware. ASYNC Compute is one of Pascals main issues.
As new DX12/Vulcan games start to use all of those APIs features, Pascal will fall behind the FPS gained by AMD and Turing. That is why NVIDIA corrected the architecture with Turing.
But it really is more a case of AMD cards and Turing cards gaining performance, then it is a case of Pascal losing performance. If Pascal is losing performance, it is minimal. Pascal can even gain a tiny amount of FPS in some incredibly light ASYNC scenarios, minimal.
So it is newer NVIDIA cards and AMD cards gaining a lot of performance from ASYNC compute and maybe Pascal losing a few FPS, if that. So it is possibly both. Pascal could be losing a few FPS and newer cards are gaining lots of FPS. And if NVIDIA implemented proper ASYNC support like they should have, Pascal would be gaining as well.
But you should absolutely expect the FPS performance gap between Pascal cards and new cards to grow with new games that use these features fully. Just look at Red Dead Redemption 2, same issue for Pascal cards. Pascal gains nothing in RD2 but newer cards gain a lot of FPS.
The next gen consoles and their RDNA2 GPUs and low level APIs (custom version of DX12.2 for Series X) will cement this. I had to upgrade my 1070ti to a 2080s because I know of Pascal issues.
I blame NVIDIA. And I would have gone an all AMD build but they still have to many people complaining about driver issues, so I stuck with NVIDIA.
So I know where you are coming from bud, but the situation will only get worse. DX12 and Vulkan are the future, NVIDIA messed up and now Pascal owners are paying for it, I was one of them and I blame NVIDIA.
@@joeyvdm1 doesnt mean idtech have excuses as there are other technologies that can still run decently on not so async ready cards
wow, i didn't think you were in your 30s, you look more like someone in the 20s
Hey stop apologies. There is no other way to show fps in old games. One core cpus. Windows xp and 98. Crap drivers. Off course there is a hit. I did my own benches in xp with a super powered pc for retro (ati x800. P4 3.0. 1gb) fraps barely did nothing to fps
Who still uses Fraps in 2020?
Seems like you're being too hard on yourself Nathan. Haters are going to hate so to speak.
Buddy FRAPS has been killing frames since ever , this is old news we tought you knew this
Dont worry no harm done :)
Try benchmarking like this:
-First pick few games that you can easily bench using in game counters
For example use games that use different game engines (you dont need to benchmark doom 3 and quake 4 just pick one of them because thy use the same game engine)
-You dont have to use external capture card , just record it with a phone or something
-Turn off any background process that you dont need (old os-es were great for this)
-Use time accurate versions of the games when possible (example hl2 is terrible for old time accurate hardware with all the new engine updates compared to the vanilla)
-Speak with other retro channels and ask them what fps counter are they using for benchmarking retro games
-compare your benchmark scores to the scores back in the day (this could be easy Video Material , tech then and now)
-Dont use virtual machines
Call yourself a expert is always way to fail.. Fraps and other tool have bugs.. they are not ideal, but they are best tool which we have, better tools has only Digital Foundry,.. and some special capture cards software.. which are actually counting actual frames from video card output, its much more work..
Most of people are playing with Fraps on, some Fraps result are actually what they would get, if there is some better Fraps alternative for Win98 let me know.. Afaik there is not eve good Fraps like tool for Glide games..
That is why I never use capture tools at all
@@toby2581 ☝️☝️☝️
So basically the conclusion we can come to here is that ATi's drivers for a one off card were a lot more robust than 3DFx's drivers for their flagship product that was years in the making.
I think I'll have to stick with the Rage Fury MAXX on this one, if the Voodoo 5 5500's drivers are so buggy that running popular benchmarking software in the background can drop it's performance by 50% then I think it's probably a pretty terrible card.
Good on you for correcting your mistake, but I think you shamed the Voodoo 5 even more by revealing this...
:)
..fraps hasnt been updated in 7 years and was first released in 1999...why youd be still using it now is beyond me, its well known to be off in FPS count... (msi afterburner! ha! you think fraps is bad for performance! afterburner sucks ass!)..steam has it built in for all dx games ive seen no hit..
Great, link me to the Windows 98 version of Steam.
@@PixelPipes 😂
You're just jumping on trends you hipster!