Kwakkelde Kwak

JoyofRLC Acker wrote: Draw distance: 900 I'm not sure if that's the cause of all your problems, but this is certainly "miles" too far for your computer. If you have your draw distance that far, SL will have to download everything within a 900 meter range, which is almost 4 sims away in every direction. This means excessive amount of data that needs to be downloaded, an excessive amount of geometry and textures to process by your computer and chances are your cache will be used in an excessive way. This last thing could cause the HD light you see. With the specifications you post I wouldn't set draw distance over 128 meters. Maybe 64 or even 32 in a club or other busy sim and maybe 256 when you want to take a picture.

Very simple answer: yes that could be it. There's one way to find out if it actually is it, just assign the material and try to upload again. That can't take more than 2 minutes.

Ciaran Laval wrote: Flip has gone, which viewer were you using it or did you just flip it by removing the minus sign? I'd say by adding 1.000, so -0.1 becomes +0.9.

Trinity Yazimoto wrote: i tried with the same file but with 20 % and then 30 % of opacity in the middle layer in phsop and it was too much transparent. Too much transparent as in completely tranparent? If so, masking, blending..give it a go On the higher graphical settings SL likes to make a guess what it will do for default. With the new material system you can set it yourself.

Looks like you have a couple of issues, the biggest being the lighting/renderer. What kind of lights do you have in the scene and how did you set up the renderer? Here's an example of something very similair: I added a daylight system (create -> systems -> daylight) I used the arch&design materials and raised the FG quality (under indirect illumination) for some better lighting effects I also added a floor to the scene for those lighting effects I used the mental ray renderer and raised the settings as shown in red for better quality

It's a long shot, but is the texture in SL by any chance set to alpha masking instead of alpha blending?

Mara Inkpen wrote: SL knows nothing about GTX 780, it forced my settings to the lowest, and when i returned them to Ultra, i found it grayed out all the advanced options: shades, shaders, advanced lightning model, so SL looks like crap for me atm (comparing with what i had with 680) o, cruel world... Take a look at this thread: Issue with GTX780

Excellent, I'll have to remember that!

I've recently read about someone else with the very same issue with a GTX780, the greyed out options that is. As I remember, the lower end 600 cards had the same issue when they were new. The drivers you have are the latest (they always show up as windows drivers in SL), your computer is well over the requirements needed for running SL on ultra settings. It's probably not what you want to hear, but I'm afraid you'll have to give it some time. The delay in loading could be a network issue.

If Maeve's suggestion doesn't fix it, my best guess is you move the hip joint (pelvis) in your animation. Unlike a standard avatar in a modeling program, the SL avatar is only pinned at the pelvis. What does this mean? Well, if you move the hip forward in an animation, the entire avatar will move a bit in SL, sliding feet and everything. The only solution I found, if you really need to move the pelvis, is moving the feet in the opposite direction when animating (hands too). Take in mind this isn't perfect either, since not all avatar legs are equally long. So if the legs of a very short avatar stay in place, legs of long avatars will slide and vice versa. I hope that makes sense

The page Rolig linked should tell you enough. If you buy L$, SL will take the US$ from your account. You already have your US$ there, so all you have to do is "buy" the L$ in your viewer or on your dashboard. On the dashboard it's under shopping -> funds, in the viewer you can click your balance. All in all the buying and selling of lindens is a two step process. (Paypal/credit card - US$ balance - L$ balance) You just need to take that second step.

I saw them higher up the thread, it has to be the new Alienware 14 As I said in the other posts, I am pretty sure it will get you 20fps on med-high settings in a club with 20 people. To be absolutely sure, someone with a similair Alienware computer (or other similair laptop) will have to share their experience. I'm sorry I can't give you the definitive answer, but that's just the way it is. I'm also not entirely sure about how the video memory is managed, maybe someone who does can shed some light on that. I showed the results with my old card with only 512 MB and it was close to what you want, but on my 2GB 670 I sometimes see video memory use slightly over 1GB. (in the three clubs I visited I saw around 900MB btw). If you can spare the 200 bucks you could get the Alienware 17 instead, which has a 765M with 2GB of memory. I just don't think the difference (if there is any) in SL performance will justify the difference in price. If you want to play other games which make more use of your graphics card it might be worth it though. To make a comparison: benchmark GT525M (743) benchmark GT750M (2543) benchmark GTX765M (4116) To make it even more confusing, the 750M in the bench has more but slower memory than the one you want.

Orca Flotta wrote: Let's get over with the infights and let's just agree on this: BETTER IS BETTER! Sorry, I'll try and stay civil Of course better is better and everything you posted about the versions and subversions is true. However, better is also more expensive and more importantly, a better graphics card doesn't neccesarily mean better SL performance, so that extra money could very well be money down the drain. I just fired up my old computer, running linux on an amd 6000+ cpu from 2007, with an underclocked 9600GT 512MB from 2008 @1920x1080 (this card is very close to the 525M actually) , wireless (and pretty poor I have to say). I went to three clubs with 15, 20 and 40 people. I got 20, 15 and 10 fps on med-high. Not exactly scientific, but a good indicator nonetheless. Even that old card wasn't pushed to its max, far from actually. Unfortunately I haven't found any good monitoring tools for Linux (yet), but the temperature was a good 10 C less than when the card is pushed, exactly between full load and idle temperature. The one year old computer with a 3770K (no OC), GTX670 @ 1680x1050, win 7 Pro and wired connection gives me 55, 45 and 40 fps on med-high settings. (45, 40 and 35 on ultra). The load on the gpu is around 40%. Wouldn't the 40% gpu load mean the bottleneck is not the graphics card? Doesn't the framerate on my old computer (almost reaching the OP's desired results) mean you don't need a very powerful graphics card? The 9600GT (that's not even my underclocked one) has a benchmark of 755, against the more than double 1685 for the 750M. Is it strange to assume a card that much better, in combination with a faster CPU and memory, on a lower resolution (1366 x 768 for the Alienware laptop in question) would be plenty to grab those extra 5 fps?

I'll bite.... Suspiria Finucane wrote: Saying the 750M will cut it is only your opinion (advice) and not the best opinion (advice) as far as my opinion(advice). I provided evidence to support my opinion(advice) while you provided rhetoric. Whether a 750M will cut it is not a matter of opinion, it's a matter of fact, whether it's true or not. If I say a yard is as long as a foot, that's not an opinion, that's a false fact, unless we are talking about a clown. Opinion would be if the 750M would give you "good" performance, which is very subjective. The 750M giving 20 fps in a typical club on med-high settings is not subjective, apart from what is considered a "typical club". The only evidence you offered is that the desktop GTX670 performs better than the laptop 750M. That's (pretty obvious and) pretty useless information, afterall, the OP is looking for a new laptop. If anyone can provide an accredited video of the 750M performing to the above specifications, I'll reverse my position. You continually state a falsehood as to "the" question. There was more than one question. Why can't you be specific? I've been specific in each of my replies which should be of more help than someone generalizing. So you take a position of the 750M not "cutting it" on the basis that a 670 performs a lot better? I at least said I suspect it will meet the OP's demands, you bluntly say it doesn't. Then you say the 750M WILL meet the OP's demands, then again you don't? And you get offended because someone points that out. What did you initially post again?: For what you want to do, the 750 doesn't cut it. I really don't see where you read the second question. The comment at the end of the post is a part of the question, which obviously was and still is: "Will I be able to get 20 fps on med-high settings in a club with a 750M?" If for some reason you think those framerates aren't acceptable, which can very well be the case, that's your opinion. Maybe you need to be more specific about what "cutting it" actually means. Claiming to help and actually being of help are two different things entirely. Differing opinions are good. Unfortunately, you seem to think only your opinion is the "best" one. How sad for the community if only advisors can post their opinions(advice) . I don't see you attempting to discredit Jean in the same way who stated pretty much the same as I have. Of course different opinions are good, but I haven't given my opinion on anything, I have made an educated guess on whether the 750M will get 20fps under the circumstances described by the OP. Who ever said only advisors can give advice? I haven't applied for it, there are just some forum statistics that will determine your "rank", apparantly I ticked the right boxes for advisor. My advice isn't worth any more or less than that of a member, moderator, helper or whatever just because I have a purple puppet next to my name. I am not trying to discredit anyone, I have no reason to. Jean said the 750M won't give very good results on ultra settings. She doesn't like laptops, that doesn't change the fact she said on med-high settings the 750M might perform good enough for the OP. I can see the guise quite clearly. I'm not sure why you state "why would you make these forums a challenge". I said nothing about making the forums a challenge. Once again that is your incorrect interpretation (opinion/advice). The challenge I referred to was something entirely different and if I wanted to make the forums a challenge, I would have stated it specifically. Then you might want to be specific about what challenge you mean, because I don't get it. An interpretation isn't the same as an opinion or advice btw. If that's not what you ment, you might want to specify that too. I think the biggest challenge for you is to admit you made a mistake. I do it all the time, it's really not that big a deal. That is making mistakes and accepting them. In ending, if you didn't want to make a big deal out of nothing you wouldn't have replied in defense of your fellow advisor. Instead you would have stuck to the issue being debated. Theresa made a valid point and a correct one at that. She doesn't need any defence. If you were an advisor and she was a member, I wouldn't have posted anything else than what I did. What's your obsession with those ranks anyway? If you desperately want to become an advisor, just answer a couple of questions in the "answers" section, give some advice here and there in different forums that get you kudo's and before you know it you'll be part of our superduper advisor club. It's not a big deal, it's trivial at the most.

Why would you say there's a motive behind Theresa's post and why would you make these forums a challenge? As far as I'm concerned, these forums are here so people can help eachother out with advice. Saying the 750M doesn't cut it just isn't the best advice, it should allow the OP to visit those places with a framerate of 20, which is what was asked. If there's a sim where the 750M doesn't give those framerates at med-high levels, neither will the 670, or a titan or 780, simply because the gpu won't be the bottleneck. No need to confuse eachother or make a big deal out of anything. Let's leave it at that.

I'm afraid that's too quick and simple of a question. If you have a fully textured and UV mapped high LoD, you can make a copy and remove loops etc to simplify the model. In that case the answer is "yes". If you however build a new model for another LoD, the vertices of model and UV map don't match. If the shape is similiar to the high LoD, you can still use the texture, but you will have to make a new UV map. So, it really depends.

Nice example Mikki, No wet surface so many people complain about and a very low "oh look at me!" level. Very subtle, very nice, makes a lot of difference.

Coby Foden wrote: How about placing invisible (100 % transparent) light source prim in the flame? The prim would move about in the flame randomly by a script. The intensity of the light prim would also be controlled by script. Perhaps this arrangement could give nice realistic flame lighting effect? I've done that in the past, well it doesn't move around, but the colour randomly changes slightly to more reddish or yellowish and the intensity and range vary slightly. It's not perfect, since you can't change the settings all that often (per second). It does look pretty nice though.

The thing is, as I said a couple of times, the tracert doesn't show any "bogs", I get a 110 ms ping to the login server. SL doesn't show any either, with a ping sim of 150-180. I'm pretty sure there aren't any bogs, yet those speedtests insist on showing a 200+ ms ping, no matter where to I test. I appreciate the feedback, but am afraid this one will go in the great book of computer mysteries...

Ah ok, I misunderstood your comparison then. Still you really do not need a video card that fast, especially if you consider the costs of the 780M. I don't get very high framerates on a full sim with my GTX670. The laggier the place is, the less my video card does, according to my gpu monitor. Temperature, gpu load, memory use, it all drops. I get around 25 fps in a laggy sim, on ultra settings, lowering the graphics to med-high will increase the fps to around 30-35 if I remember correctly. That tells me the graphics card isn't really the bottleneck in a full sim. I figure it's the rate at which data reaches your computer, which depends on the connection to and the internals of the LL servers. The question was if the 750M would get around 20 fps on med-high settings. My guess is it probably will, but it's just a guess.

Aren't you comparing apples with, uhm grapes? If I'm not mistaken, the 750 in your comparison is a mobile card, the 670 a desktop card. According to this site, the 750M and 670M (the mobile version) are pretty close if you compare them to the desktop GTX670. (benchmarks of 1667, 1909 and 5366) The only mobile gpu that comes close to the desktop 670, is the brand new 780M and you need really deep pockets for that card. The cheapest laptop I can find after a really quick search is almost 2000 dollars and isn't on sale (just) yet.

Given the complexity of the SL internals and externals (a userbase with hardware varying much much more than any other 3d application), I'd say it's a very wise choice to implement the material system the way they did it now. If there are any real issues with the new material system, it would be far easier to tackle them now it's not woven into every piece of the software yet. LL can only test so much, even through a beta program that reaches a fraction of the user base. Let them get it 100% right for objects, then see if they can implement it for avatars and terrain. Imagine the responses from the user base if LL wouldn't implement the material system right now, simply because it doesn't apply to the avatar. People have been using it for months now and it works, good enough for a release. Browse the forums for the deformer project and you'll get a good impression on how the reactions would be on a postponed material release. I don't think anyone disagrees with Penny that it would be a nice feature, but I'm happy LL decided it's time to pull the system out of beta.

I understand all that, but especially the ping times still don't make any sense at all. Like I said, both the windows tracert to the US and SL itself (through the ctrl-shift-1 menu) give me lower ping times than the speedtests give me to my own ISP which is 4800 miles closer to my home. Now that defies the laws of mathematics. The speed of light doesn't explain a ping of over 200 to my own ISP either, let alone 200 ms to the US. The time it takes light to travel from my home to Texas, 4800 miles away, is around 2.5 ms. All the delay will be in the transfer stations and metal wiring. It's as if the test is taking a detour or something. I'm sure the number the test gives me is the ping it senses, it's just that it simply doesn't reflect the actual times I get when using the internet.

Now I have a picture in my head for the rest of the day...of a half empty glass....or was it half full?

Since Drongle is offline.... You must have missed what he said about transparancy: "Note that if you set the overall transparency of the texture to anything other than 0, then it will revert to alpha blending." So it doesn't look almost the same, it is the same. I haven't tested it, but it would make sense that masking is easier on your computer than blending, since it takes less calculation. It's all local, so if your computer is up to the task, both will work. I'd use the alpha mode that best suits what you are after. If you don't see any or a lot of difference, I'd always pick masking. For your object though, you can't easily use masking. Masking means either full transparancy or full opaqueness. So parts of your glass would either be invisible or they wouldn't be transparant at all, like you see in the top middle picture you posted. Alpha masks are for example good for trees if you don't mind the jagged edges or for holes in very low poly walls. Normal, specular and alpha are three different characteristics, so you can mix and match any way you like.