Brno BVE Website
+35
BillEWS
ap1991
phileakins
MattD6R
Dexter
cris_smith14
phontanka
Quork
LabRatAndy
JimH
MrReset
jsiren
vpodhola
leezer3
buckysam
fcancalon
Stevegr
alex_farlie
rick1984
JLegrand
Stephen Cross
Drag0nflamez
Neil
skydo
theflyingoreo
Northern Line
mrknowitall
Egg
johnsinden
viccylinefan
Tstageman
pedned
graymac
busheyheath
kwijiboenator
39 posters
Page 13 of 16
Page 13 of 16 • 1 ... 8 ... 12, 13, 14, 15, 16
Re: Brno BVE Website
What did the passengers say about that?
I think I've learned my first lesson of how to swear in Czech!
Re: Brno BVE Website
Gray, have you been given the lighting file in the download? The latest 163 has a nightcab and I have implemented brightness command in the route... it is in a separate file, though. I am not sure if I messed it up or not... Thanks.
Re: Brno BVE Website
I did notice the cab darkening in the tunnels, which you hadn't done before, so you haven't messed up!
Re: Brno BVE Website
Due to absolute lack of feedback / reactions the development has been suspended until further notice.
Re: Brno BVE Website
Hello guys,
there is a new video that should explain you the signs in the route. If you are interested, take a look. I suggest watching it in full screen mode. well, once Youtube processes it, that is.
there is a new video that should explain you the signs in the route. If you are interested, take a look. I suggest watching it in full screen mode. well, once Youtube processes it, that is.
Re: Brno BVE Website
You're clearly better at building than at driving Get better soon!
Great idea of doing such a route knowledge/signal explanation video!
Great idea of doing such a route knowledge/signal explanation video!
Quork- Posts : 1438
Join date : 2012-05-05
Age : 32
Location : Hofheim a.T., Hessen (Hesse), European Union
Re: Brno BVE Website
A lot of the signs remind me of the german railways
Northern Line- Posts : 323
Join date : 2011-07-12
Age : 30
Location : London, UK
Re: Brno BVE Website
Northern Line wrote:A lot of the signs remind me of the german railways
Well, we are neighboring countries after all, so the differences might not be big.
Re: Brno BVE Website
Right,
on the 13th of October I have placed an announcement on my website asking for help with testing the optimized route - the purpose was to enable more people to be actually able to play it. There has been no response whatsoever (despite approx. 50 site views per day).
Have a good day everyone.
on the 13th of October I have placed an announcement on my website asking for help with testing the optimized route - the purpose was to enable more people to be actually able to play it. There has been no response whatsoever (despite approx. 50 site views per day).
Have a good day everyone.
Last edited by Dexter on Thu Oct 29, 2015 8:00 pm; edited 1 time in total
Re: Brno BVE Website
Sorry, I'm not watching websites regularly. The boards are a better place for such requests IMHO. 50 views can easily be achieved by crawler bots (Google etc.), especially if the view counter is rather primitive. From my nearly decade of webmastering experience pageviews are a) waaay higher than and b) not particularily correlating to unique visitors. A unique visitor can easily generate dozens of pageviews at a time.
Quork- Posts : 1438
Join date : 2012-05-05
Age : 32
Location : Hofheim a.T., Hessen (Hesse), European Union
Re: Brno BVE Website
I wouldn't mind helping with the testing but sure not what this involves, though I only have limited time I want to spend on this. I not sure whether my computer specs are what you are after for testing but mine is:
CPU:Intel® Core™ i5 processor 480M,2.66GHz, 1066MHz FSB, 3MB L2 Cache (Dual Core)
Operating System: Windows® 7 Home Premium 64bit
Memory: 4GB DDR3 (1066MHz)
Graphics: ATI Mobility Radeon™ HD 5650 (2295MB Total: 1GB discrete memory + 1271MB shared memory
CPU:Intel® Core™ i5 processor 480M,2.66GHz, 1066MHz FSB, 3MB L2 Cache (Dual Core)
Operating System: Windows® 7 Home Premium 64bit
Memory: 4GB DDR3 (1066MHz)
Graphics: ATI Mobility Radeon™ HD 5650 (2295MB Total: 1GB discrete memory + 1271MB shared memory
MattD6R- Posts : 264
Join date : 2013-06-16
Location : Brisbane, Australia
Re: Brno BVE Website
MattD6R wrote:I wouldn't mind helping with the testing but sure not what this involves, though I only have limited time I want to spend on this. I not sure whether my computer specs are what you are after for testing but mine is:
CPU:Intel® Core™ i5 processor 480M,2.66GHz, 1066MHz FSB, 3MB L2 Cache (Dual Core)
Operating System: Windows® 7 Home Premium 64bit
Memory: 4GB DDR3 (1066MHz)
Graphics: ATI Mobility Radeon™ HD 5650 (2295MB Total: 1GB discrete memory + 1271MB shared memory
Thank you, Matt.
Sorry for the delay, I was extremely busy with family stuff this week. I would like you to do the following - please download the route from my website (the old, unoptimized version) and run through it. Please take notes about any significant fps drops and where they have occured. I would also like you to note down the lowest fps + where.
The above will be the first phase of the test. Once completed, I will send you the links to the reworked version of the route and you will do the same with that one. I am trying to troubleshoot stutterings and fps drops. I have finally managed to find what caused them - by a couple of modifications I was able to lower the no. of polygons in the route from 190K to about 73K and lowering.
Sorry for the delay, I was extremely busy with family stuff this week. I would like you to do the following - please download the route from my website (the old, unoptimized version) and run through it. Please take notes about any significant fps drops and where they have occured. I would also like you to note down the lowest fps + where.
The above will be the first phase of the test. Once completed, I will send you the links to the reworked version of the route and you will do the same with that one. I am trying to troubleshoot stutterings and fps drops. I have finally managed to find what caused them - by a couple of modifications I was able to lower the no. of polygons in the route from 190K to about 73K and lowering.
Re: Brno BVE Website
I ran the route and with fps values in brackets with the CPU at maximum. It took 65 seconds to load.
5 fps (11) upon loading. At the start I get 5 to 6 fps (13-14) This slowly improves to 7 to 9 (20) at 3000m. Through the tunnel I get 10 (22). This improves to about 10 to 14 (25-29). It then droped to 9 to 11 (20 to 23) at Senice. Then dropped to about 7 to 9 (13-17) in the tunnel and to 4 to 8 from 7500m. It improved to 7-8 at Vroanov -jih. 5 to 7 (14-15) after this station and 8 in aprroach to Vroanov and at the station. 9 to 10 (18-22) after the station. 10 to 11 from 11km but dropping to 5 to 8 (12-14) afterwards from 11.5km (after the tunnel). Dropped to 4-5 (8-9) at Lipnice nad Sivou and after the station. 3 to 5 (7 to 8 ) on approach and through Lipnice nad Sivou Haje. After the station it was 3 (4-6) and to 2 (5-6) at 14km and dropping to 1 (2-4) at 17km and right through to the last station.
I did download and run this version after you released it in June with the same results compared to the above results. But just to make sure I did delete and re downloaded the files.
5 fps (11) upon loading. At the start I get 5 to 6 fps (13-14) This slowly improves to 7 to 9 (20) at 3000m. Through the tunnel I get 10 (22). This improves to about 10 to 14 (25-29). It then droped to 9 to 11 (20 to 23) at Senice. Then dropped to about 7 to 9 (13-17) in the tunnel and to 4 to 8 from 7500m. It improved to 7-8 at Vroanov -jih. 5 to 7 (14-15) after this station and 8 in aprroach to Vroanov and at the station. 9 to 10 (18-22) after the station. 10 to 11 from 11km but dropping to 5 to 8 (12-14) afterwards from 11.5km (after the tunnel). Dropped to 4-5 (8-9) at Lipnice nad Sivou and after the station. 3 to 5 (7 to 8 ) on approach and through Lipnice nad Sivou Haje. After the station it was 3 (4-6) and to 2 (5-6) at 14km and dropping to 1 (2-4) at 17km and right through to the last station.
I did download and run this version after you released it in June with the same results compared to the above results. But just to make sure I did delete and re downloaded the files.
MattD6R- Posts : 264
Join date : 2013-06-16
Location : Brisbane, Australia
Re: Brno BVE Website
Thanks, Matt, exactly what I needed. From those values I assume the route is more or less unplayable for you, correct?
Later today, I will send you a pm with links to the optimized version. I will be more than happy if you do the same with it.
Later today, I will send you a pm with links to the optimized version. I will be more than happy if you do the same with it.
Re: Brno BVE Website
Well it is bearable but at the end when the frame rates get very low it is not playable as it very hard to control the train. But it certainty would be more enjoyable with higher frame rates with the hardware I have. I look forward to testing your optimized version.
MattD6R- Posts : 264
Join date : 2013-06-16
Location : Brisbane, Australia
Re: Brno BVE Website
Guys, I do apologize about the delay. Family stuff. Regarding the polygons - I think I have done my best in the optimized version of the route. I have managed to take the total amount from 195K down to 33K, which is quite massive. Now I need to take a look at the current route ending - it is stuttering because of the immense no. of dynamic alpha faces present there (vegetation stuff).
I will ship the polygon-optimised version very soon, I would like to see the difference in fps and loading time. Be ready.
Re: Brno BVE Website
Ok - waiting!
Phil
Phil
phileakins- Posts : 95
Join date : 2011-08-08
Age : 77
Location : Swanage, Dorset
Re: Brno BVE Website
I ran your optimized version with fps values in brackets with the CPU at maximum. This version is a better but there are sections particularly the end where the fps is rather low.
It took 60 seconds to load and at the start of the route I got 7 to 9 fps (14-20). This increases to 8 to 11 (15-23) after the station and further to 15 to 20 (33-50) from 1500m. From 2500m I get 15 to 23 (45-50). Then from the tunnel and through the tunnel I get 33-41(69-80). This continues after the tunnel dropping slightly to 25-35 from 4500m to through to Senice. 19-24 through the tunnel and 15-21 (30-35) after the tunnel and dropping to 10-19 (23-26)from 6500m. This dropped to 7-11 (14-17)soon after and continued thorugh to Vranov jih and 5 to 8 (14-18)through the station. 7 to 9 after the station and improving slightly to 7 to 10 (18-21) on approach to Vranov. 10-13 (29-32) through Vranov. 13-20 (38-43) after station and 17-21 (38-42) through the tunnel. 13-19 (30-34) after tunnel. 11-14 (20-22) from 11700m and 8-14(18-21 on approach to Lipnice nad Divou and 6-7 (7-11) though the station. 4-6 (10-12) after station and through to Lipnice nad Divou Haje. 3-4 (6- after station and 3 (6-7) from 14000m dropping to 2-3 from 16500m and to 2 (4-5) from 17000m and through to Deblin.
It took 60 seconds to load and at the start of the route I got 7 to 9 fps (14-20). This increases to 8 to 11 (15-23) after the station and further to 15 to 20 (33-50) from 1500m. From 2500m I get 15 to 23 (45-50). Then from the tunnel and through the tunnel I get 33-41(69-80). This continues after the tunnel dropping slightly to 25-35 from 4500m to through to Senice. 19-24 through the tunnel and 15-21 (30-35) after the tunnel and dropping to 10-19 (23-26)from 6500m. This dropped to 7-11 (14-17)soon after and continued thorugh to Vranov jih and 5 to 8 (14-18)through the station. 7 to 9 after the station and improving slightly to 7 to 10 (18-21) on approach to Vranov. 10-13 (29-32) through Vranov. 13-20 (38-43) after station and 17-21 (38-42) through the tunnel. 13-19 (30-34) after tunnel. 11-14 (20-22) from 11700m and 8-14(18-21 on approach to Lipnice nad Divou and 6-7 (7-11) though the station. 4-6 (10-12) after station and through to Lipnice nad Divou Haje. 3-4 (6- after station and 3 (6-7) from 14000m dropping to 2-3 from 16500m and to 2 (4-5) from 17000m and through to Deblin.
MattD6R- Posts : 264
Join date : 2013-06-16
Location : Brisbane, Australia
Re: Brno BVE Website
Bit of a mixed bag Dexter.
I used OBVE 1.4.3.1-dev dated 24/08, full screen with an interpolation of mipmapping (if that makes any difference).
Loading time was 25 seconds.
Stationary at Brno - 45 fps average.
between Brno and the 50kph limit sign 25 to 40 fps
1.0k 77fps
1.2k 104fps
then to 2.0k average over 100fps although at 2.0kfps were 60. In the following tunnel fps were between 117 and 162!
3.6k 139fps dropping to 125fps by 4.2k.
Fps stayed over 100 until 5.2k then dropped to 91
5.6k 75fps
6.0k 50fps
6.4k 37fps
7.0k 28fps
7.4k 22fps
After Vranov (stop 50fps) things pick up -
Between 9.2k (56 fps) and 9.8k (97fps) then drop off
10.4k 70fps
11.4k 40fps
12.0 25fps.
Between Lipnice nad Sivou and Lipnice nad Haje fps vary between 15 and 23.
13.2k to 13.8k 23fps down to 11fps
14.0k to 14.4k 9 to 10fps then up to 20fps at 15.8k
After 16.0k fps went from 13 down to 8 at 16.2k and then down to a steady 6fps at 16.6k until 5fps at Deblin
I did find an unexpected extra station Dolni Skalky had appeared at about 13k. That was a surprise, particularly to be told off for not stopping.
Much better at the beginning that the first version I tried and more fps at the end, still 'stuttering' at several quite random places though.
Looking forward to trying the next version - anything more you want to know about this run or do you want me to try it again with an earlier OBVE version?
Phil
I used OBVE 1.4.3.1-dev dated 24/08, full screen with an interpolation of mipmapping (if that makes any difference).
Loading time was 25 seconds.
Stationary at Brno - 45 fps average.
between Brno and the 50kph limit sign 25 to 40 fps
1.0k 77fps
1.2k 104fps
then to 2.0k average over 100fps although at 2.0kfps were 60. In the following tunnel fps were between 117 and 162!
3.6k 139fps dropping to 125fps by 4.2k.
Fps stayed over 100 until 5.2k then dropped to 91
5.6k 75fps
6.0k 50fps
6.4k 37fps
7.0k 28fps
7.4k 22fps
After Vranov (stop 50fps) things pick up -
Between 9.2k (56 fps) and 9.8k (97fps) then drop off
10.4k 70fps
11.4k 40fps
12.0 25fps.
Between Lipnice nad Sivou and Lipnice nad Haje fps vary between 15 and 23.
13.2k to 13.8k 23fps down to 11fps
14.0k to 14.4k 9 to 10fps then up to 20fps at 15.8k
After 16.0k fps went from 13 down to 8 at 16.2k and then down to a steady 6fps at 16.6k until 5fps at Deblin
I did find an unexpected extra station Dolni Skalky had appeared at about 13k. That was a surprise, particularly to be told off for not stopping.
Much better at the beginning that the first version I tried and more fps at the end, still 'stuttering' at several quite random places though.
Looking forward to trying the next version - anything more you want to know about this run or do you want me to try it again with an earlier OBVE version?
Phil
phileakins- Posts : 95
Join date : 2011-08-08
Age : 77
Location : Swanage, Dorset
Re: Brno BVE Website
Cheers guys,
I have managed to start the optimization of the route - part 2. I have identified what causes the heavy lagging at the end of the route. Unsurprisingly, it is caused by the vegetation (that's what I though earlier), but it is not the trees, it is the dense bushes alongside the tracks. The question now is how to deal with this, since I do not want to use plain grass only...
Sorry for possible typos, my daughter is assisting me with me with typing... and also with scrolling the mouse, clicking random stuff etc. Just in case you thought I am mad...
Re: Brno BVE Website
Dexter wrote:Cheers guys,I have managed to start the optimization of the route - part 2. I have identified what causes the heavy lagging at the end of the route. Unsurprisingly, it is caused by the vegetation (that's what I though earlier), but it is not the trees, it is the dense bushes alongside the tracks. The question now is how to deal with this, since I do not want to use plain grass only...Sorry for possible typos, my daughter is assisting me with me with typing... and also with scrolling the mouse, clicking random stuff etc. Just in case you thought I am mad...
Whilst I haven't got the optimized version, a few thoughts based on the current release version:
Your bushes are always going to be *nasty*
A brief explanation as to why:
Faces using transparency must be rendered in order from the back of the scene to the front of the scene.
Each time the texture changes, GL.BindTexture must be called to change the texture being drawn. This is the slowest call the engine makes.
Look at the attached screenshot (Screenshot.jpg)
You're probably binding the texture ~100 times a frame, as you, which is the cause of your slowdown, as your transparent bushes change texture every ~5m along the track.
The easiest way to speed these up would be to change all the bushes to the same texture, and drop in a periodic different bush to break up the monotony.
An alternative, and better approach would be to make a basic texture atlas.
This is very simple:
Take both of your bush textures, and combine into a single texture file. (Texture.jpg is a very quick example, although this isn't intended for use)
Depending on whether you need your bushes to tile vertically or horizontally, just use the other axis.
Now, use a little basic texture mapping, so that all the bushes use this texture
Suddenly, your 100 calls to GL.BindTexture drop to one!
Notes:
You want one line of transparent pixels at the top, bottom and in between the two textures.
This prevents 'bleeding' between the two textures.
Cheers
Chris Lees
http://www.bvecornwall.co.uk
- Attachments
Re: Brno BVE Website
Cheers Chris,
thank you for the example. You are using slightly different textures than those I am working with right now, but the principle is explained very well. Let me show you a bit of code of one of my "animated" array of bushes:
- Code:
;;;;;;;;;;;;;;;;;;;;;;;;;;
[object]
Position = 0, -0.5, 0
RotateYFunction = -1.1
states = 3DBush.csv
[object]
Position = 1.4, 0, 1
RotateYFunction = 1
states = 3DBush2.csv
[object]
Position = 5,-0.75, 0
RotateYFunction = -0.4
states = 3DBush.csv
;;;;;;;;;;;;;;;;;;;;;;;;;;
[object]
Position = 0, -0.8, 2
RotateYFunction = -0.8
states = 3DBush2.csv
[object]
Position = 1.4, -0.5, 2.5
RotateYFunction = 1
states = 3DBush2.csv
[object]
Position = 5,-1, 2
RotateYFunction = -0.4
states = 3DBush2.csv
;;;;;;;;;;;;;;;;;;;;;;;;;;
[object]
Position = 0, -0.3, 4
RotateYFunction = -0.95
states = 3DBush.csv
[object]
Position = 1.4, -0.5, 4.5
RotateYFunction = 0.7
states = 3DBush2.csv
[object]
Position = 5,-1, 4
RotateYFunction = -0.6
states = 3DBush2.csv
;;;;;;;;;;;;;;;;;;;;;;;;;;
[object]
Position = 0, -0.65, 6
RotateYFunction = -1.1
states = 3DBush.csv
[object]
Position = 1.4, -0.8, 6.5
RotateYFunction = 0.5
states = 3DBush.csv
[object]
Position = 5,-0.1, 6
RotateYFunction = -0.8
states = 3DBush.csv
;;;;;;;;;;;;;;;;;;;;;;;;;;
[object]
Position = 0, -0.8, 8
RotateYFunction = -1.1
states = 3DBush.csv
[object]
Position = 1.4, -0.33, 8.5
RotateYFunction = 0.5
states = 3DBush2.csv
[object]
Position = 5,-0.77, 8
RotateYFunction = -0.8
states = 3DBush2.csv
;;;;;;;;;;;;;;;;;;;;;;;;;;
[object]
Position = 0, -0.45, 10
RotateYFunction = -1.1
states = 3DBush.csv
[object]
Position = 1.4, -0.7, 10.5
RotateYFunction = 0.5
states = 3DBush.csv
[object]
Position = 5,-0.44, 10
RotateYFunction = -0.8
states = 3DBush2.csv
;;;;;;;;;;;;;;;;;;;;;;;;;;
[object]
Position = 0, -0.8, 12
RotateYFunction = -1.1
states = 3DBush2.csv
[object]
Position = 1.4, -0.25, 12.5
RotateYFunction = 0.5
states = 3DBush2.csv
[object]
Position = 5,-0.7, 12
RotateYFunction = -0.8
states = 3DBush.csv
;;;;;;;;;;;;;;;;;;;;;;;;;;
[object]
Position = 0, -0.23, 14
RotateYFunction = -1.1
states = 3DBush2.csv
[object]
Position = 1.4, -0.61, 14.5
RotateYFunction = 0.5
states = 3DBush2.csv
[object]
Position = 5,-0.5, 14
RotateYFunction = -0.8
states = 3DBush.csv
;;;;;;;;;;;;;;;;;;;;;;;;;;
[object]
Position = 0, -0.5, 16
RotateYFunction = -1.1
states = 3DBush.csv
[object]
Position = 1.4, -0.13, 16.5
RotateYFunction = 0.5
states = 3DBush.csv
[object]
Position = 5,-0.33, 16
RotateYFunction = -0.8
states = 3DBush.csv
;;;;;;;;;;;;;;;;;;;;;;;;;;
[object]
Position = 0, -0.9, 18
RotateYFunction = -1.1
states = 3DBush.csv
[object]
Position = 1.4, -0.67, 18.5
RotateYFunction = 0.5
states = 3DBush2.csv
[object]
Position = 5,-0.22, 18
RotateYFunction = -0.8
states = 3DBush2.csv
;;;;;;;;;;;;;;;;;;;;;;;;;;
[object]
Position = 0, -0.75, 20
RotateYFunction = -1.1
states = 3DBush.csv
[object]
Position = 1.4, -0.12, 20.5
RotateYFunction = 0.5
states = 3DBush.csv
[object]
Position = 5,-0.55, 20
RotateYFunction = -0.8
states = 3DBush.csv
;;;;;;;;;;;;;;;;;;;;;;;;;;
[object]
Position = 0, -0.35, 22
RotateYFunction = -1.1
states = 3DBush2.csv
[object]
Position = 0.7, -0.55, 22.5
RotateYFunction = 0.5
states = 3DBush2.csv
[object]
Position = 5,-0.1, 22
RotateYFunction = -0.8
states = 3DBush.csv
;;;;;;;;;;;;;;;;;;;;;;;;;;
Now, this array of bushes only uses two textures. It is precisely this bush array along with another very similar one (different bush type), which causes the fps meltdowns. What I am thinking is that the texture is being called many times and also, the render has to verify the animation every frame, which is probably ovewhelming. Would it help to have a large csv with pre-rotated bushes instead of this "animated" file? It is then rendered only once and not further verified by the renderer, if I understand it correctly...
Re: Brno BVE Website
Hello everyone, Phase II of the optimization is currently in progress. Some major issues have been found and addressed. Who's in for testing during the upcoming weekend?
Page 13 of 16 • 1 ... 8 ... 12, 13, 14, 15, 16
Similar topics
» [wip]First Brno Track development
» Brno BVE Website now has its domain
» Brno BVE Website - updates
» Brno BVE Website donation system finished
» First brno track class 163 help
» Brno BVE Website now has its domain
» Brno BVE Website - updates
» Brno BVE Website donation system finished
» First brno track class 163 help
Page 13 of 16
Permissions in this forum:
You cannot reply to topics in this forum
|
|