Sleeping Dogs Gameplay Performance and IQ Review @ [H]

FrgMstr

Just Plain Mean
Staff member
Joined
May 18, 1997
Messages
55,719
Sleeping Dogs Gameplay Performance and IQ Review - Sleeping Dogs is a new game that has impressed us with its ability to graphically challenge this generation of video cards. This game offers DX11 graphics, with a unique in-game SSAA quality setting. We will look at gameplay performance and image quality and show you that even GTX 680 SLI and 7970 CFX aren't enough for the highest in-game settings.
 
I have a HD 6950 2GB at 850 Mhz and I have been playing the game with high res textures and all settings maxed out except AA at High at 1080p. The game is perfectly playable with frame rates in high 30s and 40s in the outdoors while driving and in the 50s while indoors.

A overclocked HD 7970 Ghz at 1150 - 1200 Mhz will be able to play 2560 x 1600 with high AA at avg 40+ fps with min 30+ fps. I don't see why thats not playable. A racing game or a fast paced multiplayer action game would require 60 fps to be enjoyable but not this game.

http://www.hardocp.com/article/2012/10/02/sleeping_dogs_gameplay_performance_iq_review/5
 
Last edited:
Well, please do keep in mind that we lean towards "playable" as being something we can all agree upon. There are a lot of [H] folks that find even somewhat mushy play simply not to their liking, so our bar may be a bit high, in fact, I am almost sure it is. If we find lagging framerates impacting any of the games fluidity or immersion, we call it "unplayable." You on the other hand may find that same experience totally OK. This process is subjective and that is why we give you guys the specifics and the data to compare with.
 
Thanks for the review! I luv'd the game, almost made me feel like i was in hong kong again :)
 
I have a HD 6950 2GB at 850 Mhz and I have been playing the game with high res textures and all settings maxed out except AA at High at 1080p. The game is perfectly playable with frame rates in high 30s and 40s in the outdoors while driving and in the 50s while indoors.

A overclocked HD 7970 Ghz at 1150 - 1200 Mhz will be able to play 2560 x 1600 with high AA at avg 40+ fps with min 30+ fps. I don't see why thats not playable. A racing game or a fast paced multiplayer action game would require 60 fps to be enjoyable but not this game.

http://www.hardocp.com/article/2012/10/02/sleeping_dogs_gameplay_performance_iq_review/5

We weren't using an overclocked 7970 at 1200MHz, ours was stock GHz Edition speeds.

I felt 40 fps avg was the sweet spot for this game.

Keep in mind we test the entire game, and find spots that have the lowest framerates, and use those areas for our run-through. Therefore, you can be sure the settings we specify that are playable are playable in the ENTIRE game, and not just in parts of it. The goal is to not have to lower settings in particular parts of the game, you can take our settings and rest assured that the entire game, from start to finish is playable at the settings shown on each card.

Of course there are areas that have high framerates, but there are also areas that have low framerates in the game, we have to take that into consideration in our testing, else we aren't painting the whole picture of the game. You can take the settings we've shown, apply it to the cards shown, and you should get a good gameplay experience. It's a painstaking and thorough process we go through to ensure the entire game is playable.
 
I agree with the part about 7970 CF having very choppy framerates, especially when driving. From your review the 7970 single card performance seems to be fine so I'm hoping it's just a driver issue which can be fixed, because this is an awesome game :)
 
Don't understand why you say that extreme quality 1080p was not playable on a 7970 GHz edition... 30fps minimum 35fps average is better than most console games can do, and is perfectly playable. Did you just not want to admit that the GHz edition made something playable that was not playable on the 680? :confused:

It's not smooth but I guarantee you that you're in a tiny minority of gamers if you find 30fps minimum framerate to be unplayable in a single player game.

That 7970 CF performance is weird... I'm running 7970 CF (@ 1200/1800) and getting rock solid 60fps with the game maxed out 1080p and both GPUs typically around 70% usage, so I'm unsure why they performed so poorly there. Sounds like something AMD will have to fix in a driver or CAP update (how often have we seen that before... *sigh*).

Also don't understand why you compare a 7950 to a 670 and a 7870 to a 660ti. Based on US pricing, the 7950 should be compared to the 660ti and the 670 should be compared to a vanilla 7970. Of course a $400 GPU beats a $300 one and a $300 one beats a $240 one. Didn't need a review to find that out.
 
Last edited:
As a 660ti owner I also agree that the 660ti should be compared to the 7950 based on price (even in UK).
 
Other than the bridge snapshot, I think I like High Quality better than extreme.

It would be great if you could tweak FXAA and SSAA independently.
 
We will look at gameplay performance and image quality and show you that even GTX 680 SLI and 7970 CFX aren't enough for the highest in-game settings.

And the point of such craptacular coding is precisely what? There is some sort of virtue in being 'so good' that no one can run the SW maxed out?
 
Now first of all thank you very much for a great article on a great game. These game analysis articles are what makes [H] stand out.

And then the question that made me finally register on these forums: why, oh why, no CF/SLI testing at FullHD with extreme AA? This game is one of those rare cases that could really benefit a FullHD gamer for having a second high end graphics card.
 
SSAA kill the performance, that's why it's never used anywhere but this game and Witcher 2 (UBER SAMPLING)...
 
Playable frame rates is a matter of opinion, yet the makers of this game give the option to limit frame rate all the way down to 15FPS :eek:

The developers of Sleeping Dogs believe 15FPS is playable?
 
I thoroughly enjoyed this game. It was a bit short, but it was a fun ride non the less. I have two GTX 580 cards. I am running on three 24" screens for surround gaming at 5760x1200. I have everything maxed except for the AA (set to low), shadows (can not remember but I know it is not set to high), and world density (set to high instead of extreme). I have the high res pack installed even though I did not see an extreme difference while playing the game. I feel the game runs very smooth on three monitors with these two cards. Just my two cents in case anyone was wondering about last gen Nvidia cards.
 
So this article is about image quality and performance in this game with these cards. My question is what image quality/performance do you get at an eyefinity resolution.

There has to be a ton of guys out there that would like to know this information. So what you cant max out settings at 2560x1600 that is only a 4mp picture. 5760x1080 is a 6.2mp picture. Where are the eyecandy settings going to be up there? can I still use the high res pack at this resolution with these cards?

You might come back at me with..well we cant max out at 2560 so whats the point. The point is I have three screens and I like to use them. I would like to know where these caards stand at 5040x1050 or 5760x1080 in a game image quality test as well.

You guys were the biggest promoters of eyefinity for good reason. But now you never put any test results in your reviews for these res.
 
Playable frame rates is a matter of opinion, yet the makers of this game give the option to limit frame rate all the way down to 15FPS :eek:

The developers of Sleeping Dogs believe 15FPS is playable?

Maybe if someone hasn't played any games since the N64 days 15fps would seem playable... lol

But let me put it this way: the Halo games have sold a bajillion copies and are the most popular FPS next to Call of Duty. Halo games run at 30-35fps. So, how in the world can someone say 30fps is not playable?

I think it's fair to say that you'd prefer to turn down some settings to get higher framerates, but I don't think it's realistic to say that 30-35fps is not playable. The only time I can see that being "unplayable" is if your fps is jumping down to 30 from like 60... that type of stuttering can be really annoying and unpleasant, but 30-40 fps? That's fine for 99% of people.
 
I'm into this game a couple hours, I have everything maxed at 1080p except for AA which I set to HIGH and I have fps capped at 60 and it never moves on a stock clocked 7970.... Not worth the fps dipping into the 30's for max SSAA setting to me, the game looks sweet like this, I have 16X AF enabled in the drivers too.

My cpu is a 2600k at 4.2ghz stock volts and 8gb of 1866 at 9-9-9-18 1T for comparison, using same drivers as review.
 
SSAA kill the performance, that's why it's never used anywhere but this game and Witcher 2 (UBER SAMPLING)...

That just happens to be another one of those rare "FullHD with CF/SLI for the win" settings. As in really not an option with any single GPU but maybe two would just be enough...
 
This game is very well optimized imo, but SSAA is a sure fire way to fucking demolish your otherwise butter smooth performance for a pretty minor change in gameplay given how busy the visuals generally are.
 
Don't understand why you say that extreme quality 1080p was not playable on a 7970 GHz edition... 30fps minimum 35fps average is better than most console games can do, and is perfectly playable. Did you just not want to admit that the GHz edition made something playable that was not playable on the 680? :confused:

It's not smooth but I guarantee you that you're in a tiny minority of gamers if you find 30fps minimum framerate to be unplayable in a single player game.

That 7970 CF performance is weird... I'm running 7970 CF (@ 1200/1800) and getting rock solid 60fps with the game maxed out 1080p and both GPUs typically around 70% usage, so I'm unsure why they performed so poorly there. Sounds like something AMD will have to fix in a driver or CAP update (how often have we seen that before... *sigh*).

Also don't understand why you compare a 7950 to a 670 and a 7870 to a 660ti. Based on US pricing, the 7950 should be compared to the 660ti and the 670 should be compared to a vanilla 7970. Of course a $400 GPU beats a $300 one and a $300 one beats a $240 one. Didn't need a review to find that out.

You say it yourself, it's not smooth. Raw framerates can be misleading to performance. I wish I could sit you in front of the computer and let you try it, you'd also conclude it isn't playable. 30-35 fps was laggy, choppy, all of the above. It wasn't a good gameplay experience that I can recommend in the game.
 
But let me put it this way: the Halo games have sold a bajillion copies and are the most popular FPS next to Call of Duty. Halo games run at 30-35fps. So, how in the world can someone say 30fps is not playable?

I think it's fair to say that you'd prefer to turn down some settings to get higher framerates, but I don't think it's realistic to say that 30-35fps is not playable. The only time I can see that being "unplayable" is if your fps is jumping down to 30 from like 60... that type of stuttering can be really annoying and unpleasant, but 30-40 fps? That's fine for 99% of people.

It depends on the game, a lot. Some games feel fine at 30fps, other games demand higher fps to feel smooth. Some games fluctuate so much in fps between scenes you neeed to maintain a higher amount of fps. I want people to be able to take the settings we specify, apply it to the card in question, and have a good gameplay experience from start to finish in each game. I don't want anyone to take our suggestions, apply them, and then scream out loud that they aren't playable. We are a bit conservative on settings, but we have to be, we have to consider the highs and lows in the game, and no one likes the lows. We aim for a consistent gameplay experience, of non-fluctuating fps, with smooth performance that everyone can agree is playable. It is not an easy task I assure you.
 
I might be misinterpreting the tone of the article (sorry if I am), but the author seems very excited that (a) the game is unplayable at highest settings/resolution, and (b) there's DLC that improves graphics. Somehow this means that the game is pushing hardware to it's limits and is therefore a good thing.

Shouldn't the quality of the graphics determine whether the game is actually using PC technology to it's fullest? I'm not saying the game looks bad, but are the lower framerates justified when compared to other games?
 
It wouldn't be so bloody demanding if wasn't so f'ing buggy.

What bugs, I played the game from beginning to end and had maybe 1 crash and one issue with a scripted sequence... it's probably the most solidly put together open world game I've played in years (most entries in the genre are usually broken as all fuck).
 
Out of curiosity, what were the results with 2xHD7970s at 1920x1200 (or 1920x1080) with all settings maxed out? Based on the conclusion, I'd think that this would be the perfect compromise and should deliver very good frame rates. Unfortunately I don't own the game yet, but I think this would be a great fit. The other obvious solution is to crank down some settings to achieve smooth results at 2560x1600. (I realize this isn't the aim of the test at that resolution)
 
Out of curiosity, what were the results with 2xHD7970s at 1920x1200 (or 1920x1080) with all settings maxed out? Based on the conclusion, I'd think that this would be the perfect compromise and should deliver very good frame rates. Unfortunately I don't own the game yet, but I think this would be a great fit. The other obvious solution is to crank down some settings to achieve smooth results at 2560x1600. (I realize this isn't the aim of the test at that resolution)

My point exactly. It feels like [H] decided to go with the "current gen can't max it out" theme and leave out the results showing well and truly playable frame rates with CF/SLI at FullHD maxed out.
 
Out of curiosity, what were the results with 2xHD7970s at 1920x1200 (or 1920x1080) with all settings maxed out? Based on the conclusion, I'd think that this would be the perfect compromise and should deliver very good frame rates. Unfortunately I don't own the game yet, but I think this would be a great fit. The other obvious solution is to crank down some settings to achieve smooth results at 2560x1600. (I realize this isn't the aim of the test at that resolution)

Well I can tell you what my setup gets, from the in-game benchmarking tool, maxed out 1920x1200...

Min: 45
Max: 155
Average: 84

I still don't buy in any way shape or form that 30fps minimum fps is unplayable, but everyone's entitled to their opinion. I do understand that you guys (and any review sites) have to be extra cautious about what you recommend. I have played TONS of single player games in the past with 30fps *average* framerate and found it just fine, though. Heck, I remember enjoying Doom 3 with like 20-25fps average framerates on my 9600 pro (though I doubt I would enjoy that today). I (and most people I know/have talked to) only need more than that for multiplayer gaming where 40fps+ minimum framerates are pretty much required.
 
Last edited:
sleepingdogs3.jpg



Rig in sig
 
Can sombody post some perf scores in eyefinity.

It seems eyefinity is too much trouble for the [H] to bother with anymore.

Would be really interested in 7870 7950 and 7970 scores if sombody has a few minutes.

TIA
 
[H] should be posting the max image quality you can achive at different resolutions up to and including eyefinity. Not what is the max resolution you can run at the max settings. I dont think most people with an eyefinity rig would lower res to one screen if they dont have to just to max out all the eye candy. I think they would go for the best overall compramise. For example running 5040x 1050 rather than 5760x1200 then running as much eye candy as possible. They would not say well at 2560 they cant max out the game so I am just going to play my game at 1900. This is like part of a image quality review. We know what the image quality is like at 1900 and 2560 but not 5040 or 5760 or even if it is even playable at all with generally good settings and the high res pack.
 
Can sombody post some perf scores in eyefinity.

It seems eyefinity is too much trouble for the [H] to bother with anymore.

Would be really interested in 7870 7950 and 7970 scores if sombody has a few minutes.

TIA

Here you go.
Running at max settings except high AA and motion blur off (quite a performance hit and I don't like it anyway). Rig in sig, 7970's are running at 1125Mhz core and 1575Mhz memory.

Crossfire:
83449314.jpg


Single card:
noncf.jpg



Crossfire produces some stuttering and is not smooth which is the same conclusion as the [H] review. Hoping for a driver update...
 
Thanks for that info slh28! :) I am currently running 3x 5770s and generally they do fairly well and it is hard to get a feel for what single card I would have to buy to replace those 3 to be worth my time and money. It seems my 3 cards are about the same as a 6950 or so. The 12.8 drivers smoothed things out alot for me. But my shaders really hold me back.

I am not just trying to bitch at Kyle and Brent for no reason. But when you are checking out a review from guys you respect and have followed since q2 was big and they dont give you the whole story one gets frustrated. They were pimpin eyefinity so much the last two years you would think they could at least include one chart and a few words showing readers what perf and IQ to expect if they choose to go that route on a demanding game like this.
 
Wow this game made my dated 5870 cry. It made my GPU overclock unstable, I bumped my voltage to 1.212 and all is fine.

Remarkable performance though, I get a solid 60 FPS @ 1080P max settings SSAA/off

And also tested with SSAA on, but using vsynce and 30 FPS limit. The game ran flawlessly!

So one more test, I tried a benchmark run with SSAA vsynce and 30 FPS limit at 1080p. Score was something like,
avg30 max40 min20, funny results because I kept my eyes on the FPS OSD timer and it was pinned to 30 FPS the entire time.

Personally I cant see a hair of difference with SSAA on, but certainly can differentiate 30 and 60 FPS.

This is currently my ongoing experience with Sleeping Dogs, running on dated hardware.
 
hmm could be the bench im using but my numbers just arent remotely similar to what H is seeing im getting a minimum of 30 average of 44 max 60 on Ultra Quality with Extreme AA settings but then again im using the Adrenaline bench
 
Just wanted to say thanks for the tips on the high res DLC and forcing 16x AF in the driver settings!
 
hmm could be the bench im using but my numbers just arent remotely similar to what H is seeing im getting a minimum of 30 average of 44 max 60 on Ultra Quality with Extreme AA settings but then again im using the Adrenaline bench

The benchmark is way different than the manual in-game real-time gameplay we performed. We look for areas that are the hardest on graphics, and we play the game using all the aspects of gameplay we can incorporate. I've run the benchmark, it's not comparable or real-world with what you actually experience as you play the game.
 
The benchmark is way different than the manual in-game real-time gameplay we performed. We look for areas that are the hardest on graphics, and we play the game using all the aspects of gameplay we can incorporate. I've run the benchmark, it's not comparable or real-world with what you actually experience as you play the game.

Agreed. That happens alot, unfortunately. People should really judge performance by actually playing rather than running misleading benchmarks.
 
Back
Top