Gadgetory

welcome back to hard Rome box today's video is a simple one and a bit of a follow-up to my review of the asou rog swift PG 27 you queued that we published a couple of days ago that video is pretty long as it is but there was one thing I decided to cut out for time reasons and that's the performance impact of playing games in the HDR mode compared to the standard SDR mode now you might be thinking surely there is no performance impact option to play an HDR compared to SDA after all you're basically playing at the same quality settings with the same textures the same character and object models same shadows and so forth but the reality is that switching on HDR and specifically HDR which g-sync does impact your framerate in certain games and that's something we'll be exploring in this video a couple of notes before we show the benchmarks all testing was done on my personal rig which includes a risin 7 1700 X overclocked 3.9 gigahertz 32 gig of ddr4 3000 memory and an NVIDIA GeForce Titan X Pascal which if you might recall is around the level of a GTX 10 atti the monitor I'm using is of course the asou speech in UQ which is a 4k 144 Hertz 10 bit HDR panel with g-sync HDR support however for most of this analysis I used the monitor at 98 Hertz to avoid chroma subsampling though I'll show you the impact of subsampling later all games were played at 4k with ultra quality settings the only two differences being whether the HDR toggle was enabled or disabled in the game and when the g-sync was enabled or disabled the other thing to note is this performance analysis only applies to gsync HDR at this point our plan revisiting the topic when I get my hands on a freezing to monitor to see whether AMD GPUs impacted particularly when more games come out with free sync to a special HDR processing pipeline but for now we're just testing with an NVIDIA GPU and the current best HDR display on the market let's get into the benchmarks starting with Star Wars Battlefront 2 which has an OK HDR implementation but it's definitely not the best I've seen you'll see here there are three configurations we've tested it the standard SDR mode with chasing enabled HDR with juicing disabled and then finally both HDR and gsync enabled as I mentioned earlier we're testing at 4k Ultra settings with the monitors set to 98 Hertz and connecting over DisplayPort the first thing you'll notice is all three modes delivered different performance for the best mode in terms of performance is the SDR mode and just in case you are wondering there is no difference in performance between g-sync on or off in the SDR mode in this game or any other game I tested with then switching to the HDR mode with juicing off reduces performance by 4% in average frame rates but the big hit comes when you enable both HDR and gsync it's here that frame rates are 13% lower than the SDR mode on average and 11% lower in 1% lows the results of embell front you are painting a rather interesting picture about HDR and specifically g-sync 8g up let's do some further testing hitman has the oldest HDR implementation of any game i'm testing with and in this game the impact from HDR is a lot lower while HDR does reduce performance by around 3 to 4 percent there is no significant difference between g-sync on or off in the HDR mode assassin's creed origins performs more simply to battlefront 2 in terms of the impact HDR has on performance merely enabling HDR cuts down performance by six percent but enabling HDR and gsync sees frame rates hit by 11 percent tree sinker is having a surprisingly large impact considering it has no impact when running the game in the SDR mode fire cry 5 is the only game I tested where there is no appreciable performance difference between the HDR and SDR modes far cry 5 has in my opinion the best HD our implementation available in a game today and it also supports aim DS racing - technology perhaps this game is using a different newer HDR implementation that does not result in a performance it it's really hard to say at this point Middle Earth shadow of war does not have a built-in HDR switch instead it relies on your windows HDR and WCG setting to determine whether to use HDR or not this game is impacted by HDR processing in that you get a 5 percent performance heat from simply enabling HDR while g-sync HDR takes a huge 14 percent hit the final game i tested was battlefield one in its a similar story here enabling age results in a 4% hit to frame rates and that jumps up to 11% when g-sync is added into the mix it's no surprise that these results are nearly identical to Star Wars Battlefront 2 because both battlefield 1 and battlefront 2 use the same frostbite engine and for those wondering whether 4 to 2 chroma subsampling impacts performance on the PG 27 YouTube when you bump up their refresh rate to 144 Hertz you'll be glad to hear it does not hurt performance whatsoever here the Star Wars Battlefront 2 results and the results with juicy HDR and chroma subsampling are essentially identical to the results without subsampling enabled this set of benchmarks is an interesting one we have four games that are noticeably impacted by chasing HDR and HDR in general battlefield 1 battlefront 2 Assassin's Creed origins and shadow of war all four impacted in roughly the same way with HDR reducing frame rates by approximately 4% and gsync HDR reducing performance by 12% both compared to SDR but then there's far cry 5 which shows no performance impact and hitman which is slightly impacted by hjr but g-sync appears to make a little difference with these sorts of results you can't simply say that enabling HDR or g-sync HDR will impact performance in every situation appears to be a case where some engines need to do a significant amount of extra processing to support HDR such as the frostbite engine which results in a performance hit while others use a different HDR technique but doesn't impact performance as significantly gets doing there are many differences already between game engines this varied approach to HDR with the very performance heat sounds fairly typical the more interesting situation here is when you add g-sync into the mix alongside HDR the performance hit in some games is substantial a game like shadow of war in its HDR mode sees a 10% performance penalty from simply an abling g-sync and that's definitely not an ideal situation you can sort of stomach that performance hit from just enabling HDR's there is a visual quality improvement as we know most other settings and improved visual quality impact performance but it's disappointing that enabling g-sync as well which has zero performance in SDR games also hits frame rates in some titles do you think I guess it does improve smoothness and the overall experience so for a lot of it's gonna be worth that performance in particular at 4k but I don't think enabling adaptive sync should come at such a high cost it seems that for some games at least the driver needs to do a fair bit of heavy lifting to package up both adaptive sync and HDR metadata into the display signal so that the g-sync HDR module on the display side can process it with minimal latency with g-sync disabled the module is effectively bypassed so less PC side packaging and processing is required for the HDR signal and then for some games the HDR pipeline is designed such that far less driver side work needs to be done at least that's what this data seems to be indicating while it's not good to see hgr or juicing compact performance in some games the results in far cry 5 are promising in this title there was no impact when switching on HDR or g-sync and the ACA implementation is visually excellent so there's a chance that other upcoming games will be able to implement HD in a similar way after all we are only in the early stages of HDR support on PC so it could take a few iterations to work these sorts of things out that's it for this short look at HDR game performance if you like our display related coverage consider supporting us on patreon to get access to our exclusive discord chat Steve and I are both pretty active in there chatting with you guys every day so it's definitely worth signing up I've got to go test some more monitors now so I'll definitely see you in the next one