HELP: How fast is your Video-Card | BluffTitler Community

Next Previous HomeHELP: How fast is your Video-Card

komies | 9 years ago | 11 comments | 1 like | 4,064 views

vincent likes this!

Can you beat my video export times: if you can't beat me, please send data also. Send anonymously to: testground@live.nl if you don't want to post on the forum.

Because I am also curious which video-card is better with Blufftitler.

We use all the same show:
c:\outerspace software\blufftitler\media\shows\cubemap\ReflectiveSphere1.bt

Try 3 different resolutions with 3 different settings of shadow use on layer 2.
And export as uncompressed double NTSC (59.94)
Make sure that your render screen is as small as possible before you export to video.

Windows info: goto the right of the screen / settings / pc-info or W7 This PC right click properties
GPU info: dxdiag.exe click on start / windows logo type dxdiag
stopwatch.onlineclock.net

After more results maybe we can figure out if AMD or Nvidea is faster for Blufftitler.
And compare the different video-cards on GPUBOSS.com to see what makes them faster for Blufftitler.

Windows: 10
CPU: I5
Memory: 8Gb DDR3
GPU: ATI R7 260x
Memory: 2Gb GDDR5

Show resolution:

1920 x 1080
No shadow ...........0:12 seconds
Shadow 0.2............0:23
Softness 0.5...........2:05

UHD 4K
No shadow...........0:47 seconds
Shadow 0.2..........1:01
Softness 0.5.........4:57

UHD 8K
No shadow..........3:08 minutes
Shadow 0.2..........3:08
Softness 0.5.........5:21

komies, 9 years ago

Very interesting.

vincent, 9 years ago

Windows: 10
CPU: Q6600
Memory: 8Gb DDR3
GPU: ATI HD7950
Memory: 3Gb GDDR5

Show resolution:

1920 x 1080
No shadow ...........0:15 seconds
Shadow 0.2............0:16
Softness 0.5...........0:16

UHD 4K
No shadow..........0:55 seconds
Shadow 0.2..........0:55
Softness 0.5.........1:00

UHD 8K
No shadow..........3:47 minutes
Shadow 0.2..........3:47
Softness 0.5.........3:48

gpuboss.com...

komies, 9 years ago

Without shadow is the R7 260X faster.

With just Shadow on is the R7 260X still faster. Except on the 4K?

With shadow + softness it gets a beating from the HD7950 it doesn't seem to slow down with shadow and softness on.

So that is a result, but to get a better understanding before any conclusion can be made I need more results to extrapolate a conclusion.

komies, 9 years ago

First test on my NVIDIA GeForce GT 620 took 1:15 minutes, which is 6 times slower than your 12 seconds!!! Don't dare to do the other tests... 8-[

All the demo shows and BixPacks must run like wildfire on your systems!

michiel, 9 years ago

That is what I am trying to find out, which card is best for Blufftitler.

What makes a videocard better for Blufftitler

Is it a higher number of shader cores?
Is it a higher number of compute units?
Is it a higher number of render output processors?
is it a higher F.P.U.?
is it a higher texture rate?
is it a higher memory clock speed?
or combination of these factors?

a R7 260x costs new around € 100,-
the 7950 costs new around € 350,-

So please sent the data, because it can benefit you too.

gpuboss.com...

komies, 9 years ago

i7 -2,8 Ghz
12 GB Ram
Nvidia Geforce GTX 550 Ti

komies you have a very top to other pc and this allows you to do more heavy spectacles.
This can be a problem when we sell a job and the customer does not have a pc of these characteristics.
It is best to try to do jobs that do not comsuman many PC resources.

B.Projects, 8 years ago

Hi Alex,

Call me thick, but I'm not sure I'm understanding the parameters for this test. I understand how to set the show resolution, and did so for the first run, but am confused on the shadow settings. I also realize there is no longer a Shadow Softness setting, but I think I don't have the other two settings, (No Shadow and Shadow=0.2) understood. I went to Layer 02 (Point Light), and the closest parameter I can see is Shadow Intensity. I tried setting that for 0.200 and ran the first test, but it rendered in a mere 5 seconds. That has to be wrong, as your two systems take between 12-15 seconds. Maybe I don't need a new computer after all? Ha! No way my 9+ year old beast is THAT fast.

So my question is, can you please try to explain the settings to this old guy so I can understand them, and hopefully get my baseline results posted in a meaningful way?

Thanks!
Amiga

amiga, 5 years ago

Meanwhile I believe that Michiel has optimized the program especially on the shadows, so it makes the comparisons not accurate. Maybe BT from 3 years ago and BT from today are a little different.

vincent, 5 years ago

@amiga

Yes, BluffTitler is evolving in version 13.7.01 (08/10/18) the shadow system was changed.

To make it work with these new paramaters is as following:

Open shows\cubemap\ReflectiveSphere1.bt
Set point light on Layer 2 Shadow intensity to 0.2 and for no Shadow to 0.
Make sure that in Styles it stands on Shadow Volumes
And under settings the High Quality is marked.

The resolutions are the same, the export in Double NTSC 59,94 fps is the same and I would suggest using avi as export while the rendering takes the same time in MP4 then you have to wait for the encoding which we don't count in total render time.

Good Luck

komies, 5 years ago

Well, found the time to get it done, and thanks Alex for the more clear instructions.

Windows: 10 Pro
CPU: Intel Core i7-920 @ 2.67 Ghz
Memory: 16Gb DDR3
GPU: AMD Radeon RX 580
Memory: 8Gb GDDR5

Show resolution:

1920 x 1080
No shadow ...........0:09 seconds
Shadow 0.2............0:15

UHD 4K
No shadow..........0:33 seconds
Shadow 0.2..........0:47

UHD 8K
No shadow..........4:50 minutes
Shadow 0.2..........4:47

I found a few weird things, and hopefully you can shed some (shadowless) light on them ;-)

Your older results were the same within each of the resolutions for "No Shadow" and "Shadow 0.2", but mine were quite a few seconds different. Even more odd, on the 8K test, "No Shadow" actually took LONGER than "Shadow 0.2", but only by a tiny bit. Odd...

Hopefully more folks will take the (less than) 15 minutes to run their own tests using the optimized shadow version of BT and your goal of comparing systems will be achieved! Thanks for all the help and advice you have provided me and others on this forum!

After I finally get around to building the new system, I'll update this.

Thanks all!
Amiga