Techreport 7950 vs. GTX 660 Ti "Smoothness" videos

blastingcap

Diamond Member
Sep 16, 2010
6,654
5
76
Thanks for the link, but this is not the kind of hi-speed test I had in mind. The kind I had in mind was basically the finalize-output version of frame times that TR has been using... in other words, TR's frame time graphs but with actual photos representing ticks rather than FRAPS's counting of frames. TR's frame time reporting is NOT at the end of the graphics chain which is problematic; whereas photographing the monitor IS at the end of the graphics chain.

That said, the video should indeed result in lively discussion.
 

blastingcap

Diamond Member
Sep 16, 2010
6,654
5
76
So is the moral of the story that we should stick to Catalyst 12.8 or earlier? I am still on 12.8 because I didn't really need any more speed and didn't want the instability that the early beta of 12.11 gave me.
 

Rvenger

Elite Member <br> Super Moderator <br> Video Cards
Apr 6, 2004
6,283
5
81
Anything else other than Skyrim? I can confirm that AMD cards do hitch with the 12.11 drivers on Skyrim. I have been playing BF3, LA Noire, Borderlands 2, Dead Space 2, L4D2, Rage, Dirt Showdown, Crysis 2 with no signs of choppiness or hitching... or at least nothing that is different to a GTX 670. I will also stress that the HD7950 has too low of a clock speed at stock and some games at high settings can suffer. That 200mhz OC over stock makes a huge difference.
 

SlowSpyder

Lifer
Jan 12, 2005
17,305
1,001
126
So is the moral of the story that we should stick to Catalyst 12.8 or earlier? I am still on 12.8 because I didn't really need any more speed and didn't want the instability that the early beta of 12.11 gave me.


I'm still on 12.8 and do not notice any stuttering in the games I play. Haven't tried newer drivers yet, not sure if I should.
 

cmdrdredd

Lifer
Dec 12, 2001
27,052
357
126
I'm still on 12.8 and do not notice any stuttering in the games I play. Haven't tried newer drivers yet, not sure if I should.

It might be worth testing for yourself to see if there is a difference. I know I'd be interested to hear results of user testing. Can you feel or see any differences outside of increased FPS etc.
 

Enigmoid

Platinum Member
Sep 27, 2012
2,907
31
91
I have a question. Are these video's relevant when youtube only plays videos at 30 fps. What is the point of filming at 240 fps and playing it at 30 fps?
 

Keysplayr

Elite Member
Jan 16, 2003
21,209
50
91
I have a question. Are these video's relevant when youtube only plays videos at 30 fps. What is the point of filming at 240 fps and playing it at 30 fps?

What do you see when watching these videos? The whole point is to record at very high fps and playback slowly.
 

BrightCandle

Diamond Member
Mar 15, 2007
4,762
0
76
Certainly its been controversial so far. I did my own test of 12.11 to see if fixed the stutter I saw before the driver but it didn't. The data I got suggested it was worse by 1-2 ms variation which is pretty dramatic. I also saw single card fps stutter noticeably as well where before I would say that was smooth but not quite right.

But then so far I have established my tolerance for stutter is much much less than average. Its not as if the nvidia cards don't do it as well, they do. The charts at tech report show a smaller but periodic variation which might affect some people.
 

3DVagabond

Lifer
Aug 10, 2009
11,951
204
106
They should make the vids available for DL and realtime speeds. As has been pointed out, You Tube might not be a good medium for the test and we need to see the smoothness, or lack of it, in real time, not slow-mo. They are both jerky in slow-mo. At least on my monitor with my internet connection.

I'd also like to see them with VSync on.
 

Rvenger

Elite Member <br> Super Moderator <br> Video Cards
Apr 6, 2004
6,283
5
81
I want to see more examples... Skyrim is just one example and you can't make a conclusion on it.
 

notty22

Diamond Member
Jan 1, 2010
3,375
0
0
You can use the website 'Keepvid.com' and d/l the 720p version and play it with (for example) MPC player . If your connection causes a viewing issue. It's still a recording, in a movie format VS PC output. I'm just stating that I know the obvious variables.
 

BrightCandle

Diamond Member
Mar 15, 2007
4,762
0
76
I agree I don't think either is brilliantly smooth, the vsync jerks at the bottom are quite horrible in both. But the hitches that show up on the 7950 are really bad. I don't in all honesty thing either looks acceptable honestly, but then I almost always run vsync on with settings quite a bit lower than most games suggest to ensure that never happens to me. I find that unplayable myself.
 

Haserath

Senior member
Sep 12, 2010
793
1
81
They should make the vids available for DL and realtime speeds. As has been pointed out, You Tube might not be a good medium for the test and we need to see the smoothness, or lack of it, in real time, not slow-mo. They are both jerky in slow-mo. At least on my monitor with my internet connection.

I'd also like to see them with VSync on.

They're both meant to be jerky to see individual frames. The 7950 shows much less consistent frame times though.
 

BrightCandle

Diamond Member
Mar 15, 2007
4,762
0
76
If anything both these videos show just how bad 60 fps actually is. You only have to slow it down by half to really notice its not nice at all. IQ is dreadful with vsync off but without it the hitching would likely be really bad.
 

KingFatty

Diamond Member
Dec 29, 2010
3,034
1
81
HEY, anyone know how to combine both videos into a single "split" scene? So it's like one video, with a line in the middle? I'd like to see that split-screen view, instead of the two entire videos shown side-by-side like this.
 

3DVagabond

Lifer
Aug 10, 2009
11,951
204
106
They're both meant to be jerky to see individual frames. The 7950 shows much less consistent frame times though.

I see the hitching, but in slo-mo @30fps. In real time I've never seen a game look like that when set up and adjusted properly. Actually I've never seen anything look as bad as either of them. :)
 

VulgarDisplay

Diamond Member
Apr 3, 2009
6,193
2
76
Instead of hitching I saw horrible tearing on both sides? Who knows, may need to watch it again. Real time videos would be nice also, or a much higher framerate slow motion video.

EDIT: Watching again I see one major hitch early on for the 7950. One thing I notice is that when I watch the horizon I see no hitching on either card. When you watch the bottom of the screen it looks awful which screams tearing to me. I wish they would have the FPS on screen during this. That would help me grasp what I'm seeing a lot more. It's strange to me that they didn't have Afterburner or something give us an FPS readout.
 
Last edited:

Leyawiin

Diamond Member
Nov 11, 2008
3,204
52
91
I bought a Sapphire HD 7850 OC in July and was playing Skryim almost exclusively (had been since release). On the Ultra preset it was not particularly smooth on my 60 Mhz monitor - exhibiting camera panning stutter pretty noticeable until I had been moving around in an area for a bit (guess textures were loading). My old GTX 460 1GB @ 865 Mhz was smoother on the Ultra preset...until the smaller VRAM filled up. Returned it for a GTX 670 and its smooth as butter at all times with that game.

As far as other games on my PC at the time I had a mix of RPGs - Divinity II: The Dragon Knight Saga was smooth on the HD 7850. Two Worlds II was horrible - very choppy. Sacred II was smooth, but I couldn't use PhysX (or course - all you get is nice grass and leaf movements though). Fallout 3 and New Vegas were fine with it. The GTX 670 is perfect with all of them (as was the GTX 460, but I couldn't max all quality settings on most of them).

I've mentioned this here once before and it could just be my particular hardware mix or AMD drivers at the time. Its not a judgement on anyone's experience with Skyrim and Radeon cards. It may have just been that specific card, but I didn't feel like sending it back to Newegg, trying another and discovering it did the same thing.
 
Last edited:

Ferzerp

Diamond Member
Oct 12, 1999
6,438
107
106
If anything both these videos show just how bad 60 fps actually is. You only have to slow it down by half to really notice its not nice at all. IQ is dreadful with vsync off but without it the hitching would likely be really bad.


There is nothing that suggests that they are only slowed to half speed.

They were filmed at 120Hz and 240Hz, but it doesn't say what they are being played back at.
 

cmdrdredd

Lifer
Dec 12, 2001
27,052
357
126
If anything both these videos show just how bad 60 fps actually is. You only have to slow it down by half to really notice its not nice at all. IQ is dreadful with vsync off but without it the hitching would likely be really bad.

With vsync on you have input lag. Yes I notice it in skyrim too lol.

You either have some tearing or input lag. No way yet to eliminate both 100% of the time.
 

Final8ty

Golden Member
Jun 13, 2007
1,172
13
81
Instead of hitching I saw horrible tearing on both sides? Who knows, may need to watch it again. Real time videos would be nice also, or a much higher framerate slow motion video.

EDIT: Watching again I see one major hitch early on for the 7950. One thing I notice is that when I watch the horizon I see no hitching on either card. When you watch the bottom of the screen it looks awful which screams tearing to me. I wish they would have the FPS on screen during this. That would help me grasp what I'm seeing a lot more. It's strange to me that they didn't have Afterburner or something give us an FPS readout.

They both looked like bad to me and i would not play that game like that on either setup.

And i see examples like this on Youtube all the time and i think its horrid.
http://www.youtube.com/watch?v=vwC_r2R_0To
 

VulgarDisplay

Diamond Member
Apr 3, 2009
6,193
2
76
I watched both sides seperately and paused when I saw a hitch in the video. I focused about 1cm above the cursor. I only noticed a hitch at 30 seconds on the ATi side. If my math is correct that's about frame 1700 in the graph. There is a spike there, but it's not the biggest one. Perhaps these fraps graphs are showing something because my math put a hitch at that point, but strange also that I did not perceive the major spike in the graph at 1450. Looking at it again though I maybe should not use 5000 frames as the actual number of what is in the video since the AMD side drops to 0 at around 4200. That could put the spike I see at 1450 in the right spot.
 
Last edited: