Beta Release: GPT (cockpit texture extraction, remote cockpit control, shm mirror)
-
hello GiGurra
Thanks for your works
I had build a specific post in French C6 forum
http://www.checksix-forums.com/viewtopic.php?f=286&t=185523&start=75#p1512317I had make a specific build v0.31 with in “components” 3 parts :'GPT_Transmitter", “GPT_Receiver_F-16” & “GPT_Receiver_M2K”.
The “GPT_Receiver_M2K” had the *.json" with specific x and y for Mirage2000. It is for dummies…I hope you are ok with that…
Thanks
Sharknoir -
GiGurra,
Thanks so much for these tools! I used the previous version on 4.32 and it was perfect.
In the new version I’m getting a bit of lag between my actions in the sim, and what I see on the receiver displays. Could changing the JPEG quality help minimize this? Does it affect the delay? Going from 50hz to 30hz definitely helped, but I’m thinking if I go lower than 30hz, there might be less for the receiver to deal with, but it won’t be smooth.
-
Hello,
The new version 0.3.1 only work if i have bms with Cockpit Displays to External Windows. This is correct?
I have a gpt old version that work with bms 4.33 x86 in full screen that have better fps…
-
Hello,
The new version 0.3.1 only work if i have bms with Cockpit Displays to External Windows. This is correct?
Not correct. See the readme. You need to enable two settings in a bms config file.
do NOT enable “Cockpit Displays to External Windows”GPT now runs entirely outside bms, and if you have at least a 4 core CPU should not affect BMS fps at all.
What affects bms fps slightly (exactly same as old gpt) is download of the cockpit textures to ram. -
GiGurra, can you please comment why is this version of displays extraction stops extracting after alt+tab. Previous version was working fine after switching back and forth.
-
GiGurra, can you please comment why is this version of displays extraction stops extracting after alt+tab. Previous version was working fine after switching back and forth.
Because of a bug in BMS. Confirmed by devs in this thread:
https://www.benchmarksims.org/forum/showthread.php?23390-Bug-Report-FalconTexturesSharedMemoryArea-stops-updating-after-ctrl-alt-delThey said it will be fixed in their next patch.
In short: The old GPT used a hack to get display data from BMS - which doesn’t work with 64bit. The new GPT uses proper display data extraction from BMS - but that proper method has a bug which can be seen if you minimize bms (like alt-tab or ctrl-alt-del).
As soon as the BMS devs patch BMS -> GPT will work with alttab etc.
-
Hello,
All work fine until i plug 2 usb samsumg u70 on my pc, the Displays Transmitter still waiting for bms, no found that was started.
with the 2 Samsung u70 pluged only work if i have enable “Cockpit Displays to External Windows”.Tx
-
Hello,
All work fine until i plug 2 usb samsumg u70 on my pc, the Displays Transmitter still waiting for bms, no found that was started.
with the 2 Samsung u70 pluged only work if i have enable “Cockpit Displays to External Windows”.Tx
Sorry no clue what could be the problem. Unless you accidentally removed the export configuration from bms.cfg.
Do you have some buddies to help you with configuring GPT?GPT should NOT in any way require “Cockpit Displays to External Windows”.
-
All works fine without the u70.
With the u70 “Displays Transmitter” not found that falcon bms is running.
With old gpt version i don’t have problems with the u70.
-
Just to confirm, I use GPT with 2 small side monitors for MFD powered by 2 vid cards on the same PC and do not enable the BMS “Cockpit Displays….” Must be something with your individual setup. Win 7-64, NVidia GTX 760, BMS 4.33. Very difficult to assist remotely, though, beyond the obvious. I start the Displays Transmitter, then Displays Receiver, then BMS.
-
Samsung u70 are usb.
I test gpt with a extra monitor conected to video card and all works fine.I think that is something with resolution that make that gpt not found the bms when i plug the u70.
In old gpt version there are this;
<shm_tex_w>1200</shm_tex_w>
<shm_tex_h>1200</shm_tex_h>
In new version is someway to force this and not search for the game -
Sorry if i ask something that has been already wrote, but if so i cannot find it
Anyway i am looking for a way to extract mfd and instruments to a windows tablet and using its touchscreen for issuing key commands.
With MFDE the instruments and MFD extraction works good, but there is no way to send keystrokes to the main rig running Falcon.
Is there a way to do that with this program?I would like to try it, but only if i can do this -
Samsung u70 are usb.
I test gpt with a extra monitor conected to video card and all works fine.I think that is something with resolution that make that gpt not found the bms when i plug the u70.
In old gpt version there are this;
<shm_tex_w>1200</shm_tex_w>
<shm_tex_h>1200</shm_tex_h>
In new version is someway to force this and not search for the gameCheck the displaysreceiver config. You’ll want to edit the target positions and sizes of displays. The source doesn’t matter.
-
Sorry if i ask something that has been already wrote, but if so i cannot find it
Anyway i am looking for a way to extract mfd and instruments to a windows tablet and using its touchscreen for issuing key commands.
With MFDE the instruments and MFD extraction works good, but there is no way to send keystrokes to the main rig running Falcon.
Is there a way to do that with this program?I would like to try it, but only if i can do thisGPT only works on PCs unfortunately.
-
GPT only works on PCs unfortunately.
yes i know…my tablet is windows so it is a pc and i already managed to showbl instruments on it with your program. what i am asking is how to use mfd and cockpit keys sending the callbacks to my main pc running falcon…thanks
-
yes i know…my tablet is windows so it is a pc and i already managed to showbl instruments on it with your program. what i am asking is how to use mfd and cockpit keys sending the callbacks to my main pc running falcon…thanks
Well GPT only handles forwarding of keystrokes and display of mfd textures. GPT does NOT display or create any mfd buttons or kesytrokes - That you need other software for. I personally use Helios for switches and Thrustmaster f16 mfd buttons.
You could write your own keystrokes generator and send signals to GPT keyboard receiver. If you got the programming skills…
Check https://github.com/gigurra/gpt for the protocol. Might switch that to ZMQ in the future instead of my own hack. Essentially just json data right now over a custom (mnet) message layer. -
I’m still getting lag using these new versions This is on a gigabit LAN. Oftentimes if I hit a button, such as SWAP, it’ll take a few seconds before it shows up on the Client.
On previous versions (those for 4.32), I could get good results even when the Client was on WiFi. Any tips on reducing lag? Any settings that would affect this besides max_hz and jpeg quality? (for example, the old version has a max_kbps variable).
thanks.
-
I’m still getting lag using these new versions This is on a gigabit LAN. Oftentimes if I hit a button, such as SWAP, it’ll take a few seconds before it shows up on the Client.
On previous versions (those for 4.32), I could get good results even when the Client was on WiFi. Any tips on reducing lag? Any settings that would affect this besides max_hz and jpeg quality? (for example, the old version has a max_kbps variable).
thanks.
The delay is caused by the slave not being as fast as the game PC. That means the game pc is producing more mfd frames/images than the slave can render. Download 0.3.3 and set the max fps setting for the displaystransmitter in the cfg to something lower. There should be absolutely zero delay between the game and the slave displays. Well, in the order one frame time (< 30 ms worst case).
The slave code hasn’t changed at all in 0.3 so the same issue must have existed previously as well… Probably something else you have that has changed - or maybe you had the maxfps parameter set back then.
Note that in 0.3.0 the fps limiter was broken and that created a lot of lag for a lot of people.
0.3.1 and newer has this fixed. -
Well GPT only handles forwarding of keystrokes and display of mfd textures. GPT does NOT display or create any mfd buttons or kesytrokes - That you need other software for. I personally use Helios for switches and Thrustmaster f16 mfd buttons.
You could write your own keystrokes generator and send signals to GPT keyboard receiver. If you got the programming skills…
Check https://github.com/gigurra/gpt for the protocol. Might switch that to ZMQ in the future instead of my own hack. Essentially just json data right now over a custom (mnet) message layer.But you use Helios over network?Or with the same PC?Maybe i could configure helios send the actual keystrokes(changing the falcon keyfile in order to have and your software intercept it and then forward it to the receiver that is listening on the main Pc?
-
G’day Gigurra,
Is there any chance you would create additional exports for GPT?
Now that the Backup UHF panel is useable it would be great to have the UHF Preset channel and UHF Frequency exported to use with Helios panels.
Thanks for your time and efforts,
- Bear