Monkey Frog Studio
-
Posts
165 -
Joined
-
Last visited
Content Type
Blogs
Forums
Store
Gallery
Videos
Posts posted by Monkey Frog Studio
-
-
Ok. According to those that test video cards (just read several benchmarks), there is a pretty decent difference between a mobile 1070 and the desktop version. So, that is most likely the answer. So, like I had originally stated, the 1070, at least desktop, is fine for VR. Your slow-down issues are most likely due to the differences here. Since the OP was originally getting a card for a desktop, these differences don't matter. And since he had already purchased a 1080, it doubly doesn't matter.
EDIT: apparently the main difference in speed comes from the power the GPU can draw upon. On the desktop, with a proper power supply, it's "unlimited". Not so on the average laptop. This can slow down the same GPU (1070 on a desktop vs. mobile) by 15% or more. On some game tests, the frame rate was HALVED on the mobile version.
-
Sigh. You're not answering the question I had posted before. Is the 1070 mobile the same as a 1070 for desktop? Also, what are the system specs of that laptop that NASA is using with the 1070 installed. It can all make a difference. In any case, it doesn't affect me. I am not (yet) developing for VR. However, as stated previously, from the articles and reviews I've read thus far, there is actually little difference between the performance of the 1070 and the 1080. Yes, there is some. But for most, it is not noticeable. So, if there is a big difference between Leadwerks on a laptop with a 1070 and Turbo running on a completely different PC with a 1080, I would guess there are more differences than just the video cards.
-
1 minute ago, Wafflesoft said:
wow half a mill in polys seems like alot. My game doesn't have any models that use more then 30k polys.
Half a million? That was just the Rover model. He stated some of the models they were using were over 2 million.
Even so, I know this isn't for a real-time game engine, but I can normally manipulate a 3D model with well over 9 million polys in real-time in my modeling program. Heck, in Blender 2.8 I've had many millions more than that and was still able to move about in the view port in real time. -
Did you read through that thread I linked you to? Several of us have been having issues with Leadwerks not going true full screen. We played with just about every Nvidia setting we could think of and Josh eventually said it was a bug and he would make a fix for it ... eventually.
-
12 minutes ago, Josh said:
That says that we are running an application with a load heavy enough that the 1070 mobile GPU cannot process it within 11 milliseconds. Some of the models we use are two million polygons, with thousands of sub-parts, so that is not surprising. I think the Rover model we were using was 500,000 polys alone.
Right. So, that's not really a problem with the 1070 and VR, but with optimizing the scene. But, once again, what version of the 1070 is this? Is it really a mobile version? Or a full 1070? I don't know how Nvidia is making their mobile cards these days, but there used to be big differences in performance between mobile cards and desktop cards, even in with the Nvidia 9 series.
-
For Leadwerks not actually running full screen, check out this thread:
-
1
-
-
22 minutes ago, Wafflesoft said:
The 1080 seems to be working amazingly. I just have an issue with it and leadwerks. Its something to do with higher resolutions and dpi scaling. I posted another thread in general about it. Leadwerks as well as my published game aren't going into full screen properly and giving off weird resolution numbers in the game's menu.
Check out this thread:
There seems to be an issue with Leadwerks on some systems when attempting to run full screen. The thing I notice is that all of us have Nvidia 10 series cards (1050, 1070, 1080).
-
Yes. I get that you have a slowdown with VR on a 1070. Others don't have a noticeable slowdown with VR on a 1070. So, what does that say?
However, which mobile card is it on the laptop? What are the specs of the laptop compared to the system with the 1080? I'm just curious. -
Yeah, I obviously can't speak to that as I don't have access to Turbo, etc. But with the games and VR applications made for and released to the public, the 1070 works just fine. In most cases, according to reviews and end-user experiences, the difference between the two cards is either negligible or not at all noticeable. So, if I were purchasing a video card (and not doing what you do) and was interested in VR, I would not hesitate to get the 1070.
QuoteTurbo is faster
Than what?
If Turbo is experiencing a noticeable slow down on the 1070 and not the 1080 in VR and these other games and applications are not (or not as noticeably) then there might be something else going on, I would think. Yes, the 1070 is slower overall than the 1080, but they use the same chip set (basically). -
According to just about everything I've read, the 1070 does just fine with VR (not the 1060, as it's a different chip, if I remember). The 1070 is a little slower and uses slower VRAM, but is not all that different than the 1080. One article even called the 1070 the "VR Sweet spot".
https://newatlas.com/nvidia-geforce-gtx-1070-review-vr/44509/
On one of the REDDIT channels, they are saying VR on the 1070 is "smooth as butter" in all the games they played on it in VR.Now, that's not to say that the 1080 doesn't have more power. It does. However, a 1070 is just fine (and, in many cases, more than fine) for VR.
-
Who brought up the 1060?
-
You would do well with either the GTX 1070 or 1080. You really have to look into the differences between the two to see if the price difference is worth it. I am running a 1070 at the moment and there isn't a thing I can't run on it. All games run at their highest (ultra) settings, I create complex 3D models for work on it, etc. So, for me, I didn't have a need to dish out the extra for the 1080. But that's me.
-
If I run MyGame.exe, it's the same as I reported in my previous post (nothing happens). If I run test.bat, I get the dialog stating to press any key. Once I do, it simply disappears. So, even though test.bat is set to run MyGame.exe, it doesn't run.
Just for FYI - I created a folder on my desktop and dumped all the files from the ZIP into it and ran it from there.
-
Screentest3 did not even run on my system. I double-clicked the EXE, the cursor spun for a moment, then nothing.
-
4 minutes ago, Josh said:
@Monkey Frog Studio If you set your monitor to the native resolution does this work properly?
My main monitor has been set to it's native resolution, which is 3840x2160, throughout all this testing.
-
16 minutes ago, Josh said:
Try this. All I did was make a call to SetWindowPos() after creating the window, setting the position back to 0,0.
It will just be a spinning blue box in the middle of the screen, with no menu.
Nope. Same issue. It's not full-screen (desktop can be seen to the right and bottom). I cannot tell if any of it is off-screen because hitting escape simply exits the program (instead of showing game controls). But it looked like it wasn't off screen to me (though I could be wrong).
-
Since this seems to be a fairly isolated incident (only two of us reporting it), then I wouldn't do something that could potentially add issues.
-
Yes, this behavior continues if I unplug the HDMI cable of the second monitor.
The scaling factor of my primary display is 125%, but this also happens if I have it set to 100% (thus, both monitors 1 and 2 are set to the same scaling).
I have the pro edition of Leadwerks, despite not being a programmer.
Can you tell me where to add the SetProcessDPIAware() and if there are any parameters I need to set?
Running your testscreen app via Troubleshoot Compatibility did not turn up anything and I tried several different parameters.
-
Here's a screen shot:
It looks like the screen has been resized, but app is shoved off to the left and top. Notice how the game controls to the left are off screen.
-
I have a GeForce GTX 1070 (Founders Edition). I've running the latest Nvidia drivers, too.
-
I'm not sure what to tell you. The screen test file you'd sent me previously and any of my own will not run full screen. This is not the case with any other game (modern like Prey or old like System Shock 2) or game creation system (like Unreal, Unity, or Godot).
-
Just wanted to update this thread. I did a complete uninstall of all Nvidia drivers using DDU (Display Driver Uninstaller), which basically boots you into Safe Mode and totally cleans out any remnants of your graphics driver. Then I did a reinstall of the latest Nvidia drivers. This, unfortunately, did not solve the full screen window issue for me. When I run Josh's screen test app or a Leadwerk's game at full screen, the game is not full screen and is partially off screen to the top and left. So, this does not appear to be a driver issue.
-
1
-
-
Jorn Theunissen has a few Leadwerks tutorials on terrain creation on YouTube. Of particular interest to you might be:
-
1
-
1
-
-
32 minutes ago, Josh said:
It is possible they might have modeled it in a way to make it appear that way, like maybe it's actually really really short.
I don't think they model it that way. In most games the rifle moves smoothly into position from it being held to right in front of your eyes. I think what they do (and I could well be wrong) is use another camera view for inside the scope that is zoomed in.
GPU card
in General Discussion
Posted
Well, yeah. I agree. However, that's part of the problem right there ... CAD models. CAD is great when you need to physically build something from it. But it's pretty rotten when you need it for other things, like rendering or real-time 3D stuff. I've had a few jobs in the past taking CAD models for clients and reconstructing them accurately as SDS/poly models so that we could have reasonable render times. You'll definitely get the performance you need if you can not only convert these CAD files, but get someone to reduce them down. They look the same once reduced, but have a heck of a lot fewer polys!