I hardly see any difference.

same here...
looks crap on both - hehe (take a look at the scene with the old woman - the skirt is totally edgy - thats state of the art)

The difference is insignificant.

make sure you also apply http://skyrim.nexusmods.com/download...d=8236#content

I see no difference.

Are you sure you didn't muck up something? Both vids are the same lol

On two, maybe three images i spotted any noticeable difference. HD textures pack gives nothing.

you'll always notice the difference on weapons, many are 4x the original resolution.

There is a slight difference. Some textures look a bit sharper and more detailed but not drastic.

LOD bias -10

just an FYI to viewers: even full screen, this particular youtube embed won't go above 360 resolution, so you have to go to youtube to see it in full resolution and see the differences.

The improvements are barely noticeable. I was expecting a lot more

Nope mate

yes ncaine, the site embed player is limited to 360p

Yeah, just went to youtube to watch in 1080p. No difference. Seems like a lot of work went into this and not a whole lot came back out of it. There is a slight difference but if you were actually playing and not just standing there staring at the grass you would never notice.

It was fixed this morning. You should be able to switch to full screen and adjust video resolution.

Alot of the common textures like wood, rock, and many ground textures were replaced. It really was difficult to make them out at a quick glance, but once you know the textures and compare them you will notice quite a difference.

Things to keep in mind, the youtube videos are still very compress and the native resolution is actually a bunch of pixels below the selected resolution.. ie unless you choose the 1080p option, it will be hard to notice anything. Personally I thought skyrim did very well with its textures in the original release... and the update pack is simply a nice bonus for us pc users.

I think people were probably expecting the huge differences you usualy see with user made texture packs, without realizing that a) the new texture is a completely different source from the original and b) the performance drop on these megatextures is often huge and needs a poerful pc while a nominal pack like this one probably had no performance impact.

btw image verification sucks balls if it doesn't even work. I type whats i the image and it says it is wrong.

Most of the textures from this pack are only 1024x1024. It's only useful right now because modders are still working on retexturing everything.

No.

Looks to me like all they've done is run a sharpen filter over the standard textures and some think those are HI-Res sorry but I've got 4k textures in some of the mods from Nexus if they can't compete with that then why bother

yep.

Nice video

HD Video on 27", 2560x1440 monitor.

I looked very carefully and could only see the difference in about 25% of the comparisons.

Whoever made this video screwed up. Hardly a big endorsement. HD pack not recommended if it drops your frame rate below 60FPS as you want VSYNC to prevent screen tearing.

Not too many systems could handle:
- HD PACK, and
- 60FPS, and
- Maximum quality settings.

you have no idea what you're talking about.

you want below 60fps for vsync anyway as it prevents lag.

You get screen tearing when you go past 60FPS as more images are fed to the screen then it can actually display. vsync prevents that, but if you drop below 60 there is really no worries as far as tearing goes.

Don't understand how it prevents lag (lag relating to what, exactly?) though, maybe squall there can shed some light on the matter.

synchronisng a framerate that is higher than 60 introduces a delay period of approximately 20ms (where a framecapped no vsync 60fps is 16ms) This can be felt in most first person shooters with vsync enabled (on 60hz displays atleast).

capping the framerate at just below 60 when using vsync+TB allows the frame to display Just-In-Time, and thus the latency is always 16ms max.

Im sure theres a far more clear way to explain it, but put simply. try running Doom 3 with vsync (mouse smoothing completely off) and you will find it feels like you're dragging the aiming reticle around the screen, where as inputting 57 or 58fps as a frame cap (Nvidia inspector or Dxtory) the input will feel far more responsive.

Homeworld 2 is another title where the frame cap really helps.

But all of this is valid only if your hardware can do >60FPS.

This "buffer" is in hardware or software?

swap buffer, bit of both.

you also have the vsync lag when fps suddenly drops from 60, or is unsteady with triplebuffering. AC2 had a lot of instances of that.

It does looks kinda the same, the new texture set is slightly better, but i would rather use a Fan HD texture pack.

Did you really remember to activate the high resolution textures in the launcher? Cant see any difference.

re: VSYNC:
It seems a lot of people here are confused on this issue.

VSYNC OFF:
You get screen tearing no matter what, though it varies from not very noticeable to very annoying depending on the refresh rates your getting and the way the game was made.

VSYNC ON:
The game will ONLY generate the correct number of frames (i.e. 60FPS). It will not generate "100 frames and buffer the rest." Try an older game and you can actually monitor the CPU usage. A game running at 180FPS vs 60FPS will actually use 3x more CPU power. I've tested several games.

VSYNC can NOT function if the game can't produce a high enough frame rate. If you drop below 60FPS it's the same as VSYNC being off.

VSYNC vs quality:
If the game doesn't show screen tearing then great, maybe running at 45FPS isn't so bad, but most games do and it shows up worse on a large screen.

What I'm saying is, if possible try to tweak your quality settings so you can achieve VSYNC most of the time. Obviously it's not always possible.

Sluggish with VSYNC?
This issue is actually quite complicated and beyond the scope here.

Tearing occurs any time an incomplete frame is swapped into the front buffer, this can happen at any frame rate.

Wrong, Framerates heavily exceeding the vsync will create an input buffer which is aproximately 4-16ms of delay. 1000fps vsynced to 60 has more then 30ms of input latency, where 80fps synched to 60 is closer to 20.

Yeah?... THATS JUST PLAIN STUPID. framerates below the sync rate will divide by 1.5 - 2x or more as you get further from the refresh rate because of doublebuffering. framerates under the refresh rate will still tear with vsync off (WHICH THEY DON'T WHEN VSYNC IS ON), Hence your nonsense that vsync disables when under 60 is pure stupidity.

All games tear. it doesn't matter what effects they use, they all end up rendering to the display the same way and the swap chain introduces tearing as a result in the case where vsync is off.

Is this again based on your ignorant/incorrect assumption that vsync is disabled when under the refresh rate?

You just didn't want to look more like an idiot then you already did.


You have absolutely no idea what you are on about.

Vsync introduces input lag because the input frame does not match the frame currently in the front buffer, its actually 1 frame behind it in an ideal situation. the higher the frame rate prior to vsync the more frames it can be behind. Eduke32 at 1000fps synched to 60 is 4 frames behind the frame in the front buffer.

How did you arrive to the figure of 4 frames in this last example?

squall_leonhart,
I'm sorry but you are incorrect on everything you've said here.

For example, find an older game and run it with FRAPS to show the frame rate and VSYNC disabled. Leave the Task Manager on in the background to monitor the CPU usage, then close the game and see what it was.

Let's say your game ran at about 180FPS with VSYNC disabled and your CPU usage was 75% .

Now if you run the game with VSYNC enabled getting 60FPS you'll discover that your CPU usage will be about 25%. It's remarkably proportional to the frame rate.

The reason for this is simple. With VSYNC DISABLED the computer generates as many frames as possible, limited either by the CPU or the Graphics Card.

With VSYNC ENABLED, the computer generates ONLY one frame for each refresh on the screen. THAT IS WHAT VSYNC MEANS (synched to the vertical refresh).

*As for the rest, I don't know what to say. I'm a computer technician. I haven't the slightest doubt about what I'm saying. Don't get mad, I'm not being a jerk. Obviously you think you are correct.

SKYRIM and HD Textures:
Again, if possible the OPTIMAL way you want to run Skyrim (or most games) is at 60FPS with VSYNC enabled. If your HD Texture Pack brings you BELOW 60FPS you not only can NOT VSYNC (screen tearing) but you also have a lower frame rate, and possibly severe stuttering at times.

It's a tradeoff. Quality vs framerate. Plus the VSYNC (screen tearing) issue.

(so I ONLY recommend the HD Texture Pack to those people with powerful gaming PC's who can achieve 60FPS VSYNC'd with the Pack applied. That's very few people.)

CRT monitors vs LCD monitors:
Everything I've said about VSYNC applies to CRT monitors. The only difference is that CRT monitors usually run at 70FPS to 85FPS on average. (due to flickering caused by the phosphor glow dissipation which is worse at low refresh.)

I know a lot of computer technicians who know jack shit on how PC subsystems operate, so that fact on its own means literally nothing.

Is he? (I can't believe I'm going to defend squall here, one of most monumental jerks of all time; and in fact, I'm not defending him, I'm defending common sense, reason and the undeniability of facts)

This is 100% correct and if you think any differently you're just being silly. And a bad computer technician. The fact of the matter is, the screen knows jack shit about what's going on in the frame buffer. It could be ready or being copied over or the copying process has just started or it may be near its end.

EDIT: this is a possible source for misunderstanding; with doublebuffering you get two buffers (duh) which are being swapped when they're done being drawn on. Without syncing the system doesn't pay attention to which one is being currently displayed on the screen so the gpu may write to the "wrong" buffer which leads to tearing.

I'm not sure about this part and some of these numbers, but if some frames are dropped, lag has to occur. Since 1/60 = 16ms, some of those numbers sound right.

Again, 100% correct. The fact of the matter is, if you have double buffering and the card (system) can't keep up with screen refresh rate, then the screen will just have to display double frames as there isn't anything new to use. Lets say, the card fills the frame buffer once it's done with rendering on screen frame 1. On screen frame 2 the rendering is not complete so the screen has to display the previous frame which is still in the frame buffer. On screen frame 3, the card manages to do it's business, fills the frame buffer with a new frame and it gets displayed on the screen. Guess what happens with screen frame number 4. If you have a screen capped at 60Hz, you'll get about 30FPS effectively. So, vsync off is not equal to vsync on with framerates below the screen refreshrate. He's not wrong, you are. Not because I say so, but because there is a certain way things work. A lot smarter people then me designed those systems and I wouldn't dare saying anything against them.

This one is a rerun and still correct.

Again a rerun, again correct, again just facts about how the whole rendering and displaying process works.

This was a question and by very definition it can't be wrong.

I'm afraid both of these are very correct.

This is obviously correct as by definition vsync makes frames drop so it's plainly evident you have to lose frames. You lose them anyway as the screen just can't use them all, it's just so that vsync prevents tearing (well, if done correctly, some broken games tear no matter what, i.e. GTAIV). The good thing about it is: more frames there are shorter bits of time they represent so the lag is fairly limited.

EDIT: actually, you get a mix and match of various frames between the tears. So you get displayed part of new and old information.

This is one of the bits I don't get and I'll refrain my judgement on pending further explanation.

Its an opengl specific oddity. i don't quite get it myself but games like Homeworld 1 and Doom 3 get more then 1 frames worth of delay. Its probably a quirk of triplebuffering since the additional backbuffer would create an extra frame of delay and both of these titles use it by default. Frame pre-rendering does not help either in some cases causing more latency on top of it.... BF3's default internal setting of 3 stacks to 6 frames prerendered by the cpu both by the engine and the driver.

I previously used Release 290 with the frame limiter to cap doom 3 and the homeworld games, unfortuntely though the setting is not working for opengl in release 295.

Infact one game that reverses the situation above is NWN. it uses more GPU and CPU processing with triplebuffering off, incurring lag.

The program itself is responsible for producing this sync state and the buffer swap? I just can't understand why would some games act like they do with vsync on. Already mentioned GTAIV with vsync enabled will give infinite loading screens and will not stop tearing anyway. Can the game ignore the vsync enforced through the driver?

GTA4.. wasn't that a bug specific to D3DOverride?

Well, isn't that nvidia specific? I had problems on ATI card, don't know if the later game versions fixed it, but on patch 4 I had tearing with (ingame) vsync both on and off. Haven't tried forcing it from the driver, though.

its bundled with rivatuner, but its not restricted to nvidia only.

Well, I did use rivatuner, but only for setting the fan speed profiles. I left all the other settings alone.

This is my last reply:
There was some truth in the lag issue (which I said was actually a complicated issue, and it is).

As for the rest, I'm not concerned if you think I'm incorrect. Your both throwing around terms with some truth but your fundamental concept of how VSYNC works is flawed. The entire point is for the computer to generate one frame to be buffered and synched for each frame the monitor draws. Yes, it's SLIGHTLY more complicated then that if you wish to discuss doublebuffering but the above is the basic principle.

The two most important things to keep in mind are this:
#1. Computers will generate as many frames as possible with VSYNC disabled (I had over 300FPS in an old game) and processing on the CPU is almost perfectly proportional to the framerate, thus 60FPS uses about 20% the processing on the CPU and GPU as 300FPS) so the computer runs much cooler in this case with VSYNC enabled.

And it's easy to test with the Task Manager monitoring CPU usage. Go ahead and try a game with VSYNC enabled and disabled (turn down all quality settings to lowest to make the FPS difference higher).

#2. Tearing is a main reason to use VSYNC if you can achieve the required frame rate, such as 60FPS to match the monitor's refresh. It's a simple matter to find games with incredibly annoying screen tearing that are fixed by enabling VSYNC, though tearing varies from almost unnoticeable to unplayable with VSYNC disabled.

None of those points address any of the crap you spouted a few posts above. But yeah, I do understand why would you want to change the subject, it's ok, just the next time check your facts beforehand.