Livestream development - rpy dir & pribluda

[Olaf Lehmann]

Livestream from Yesterday works today on my tv.
I have seen that the new current speed values are frozen too short. I have changed the "refrigerator values". Now we will have lesser high and low speed measure points, but the displays are not so nervous, I think it's better.

That will be the only livestream development from my side before the next race.
I hope until the race on the 8th of November the director will be not more ignore the leader.


It seems yesterday the livestreamer client wasn't on track for a moment. That should be the reason that the trackpositionboard was dissorted.
Sorry for the circumstances Andre' , but I can't make it easier.

BR Olaf

P.S.: I saw the sectortimes display in quali. Pribluda has only memory adresses for one driver.
In case of using replaydirector the same adresses will be used for different drivers. That makes invitably wrong results.

[Dag_J]

It seems yesterday the livestreamer client wasn't on track for a moment. That should be the reason that the trackpositionboard was dissorted

Indeed Olaf , we had some issues with the machine that runs the stream so we had to start from scratch yesterday afternoon and in our haste it seems a couple of details went right out the window. Putting the tv car on track was one of them it seems. Lets call it weapons of mass confusion..

[Olaf Lehmann]

I have good and bad news:
Bad: I have lumbago or slppped disc. I have to wait what the doctor says.
Good: Because I can't do anything in the garden and had time for replaydirector development. So what I wrote in post 21 is ready before my "self-said" deadline, see srmz replaydirector thread.

BR Olaf

[Dag_J]

Very sorry to hear about your back issues Olaf, brings me back a few weeks , its no fun when your back is out of comission. Hope you have a quick and speedy recovery. Great news about the rpy director , we are gonna do some testing this week so thank you very much for that. Get well soon Olaf.

EDIT : Andrè just tested this and the leader focus works.

[André]

Very sorry to hear about your back issues Olaf. Hope you recovery soon.

I have now tested the new version and it works great :-) Picks up the leader as a charm :-D

Some ides to some small live stream improvements:

Now that you have been able to recognize the leader. It would be create if we could follow the leader at the start. At least in to the first corner. That would be a create feature.

And if we also could hard code or configured that TV1 was always used in the start, heading in to the first corner. That would also be a nice feature.

This would allow us to configure the ini file to drive inboard for the rest of the race. Since that gives the audience the best viewing experience.

Get well soon!

BR André

[André]

Hallo Andre,
if have copied the link from the website (above) and from Kyalami thread (below)



The Kyalami link works for me.
Could it be the "/" at the end is the delinguent?

BR Olaf

Hi, Olaf!

Have finally managed to solve the link issue to YouTube from our website.

I needed to add the following at the end of the link address: /" target="_blank

Should work perfect now, both from any mobile device, and any browser.

BR André

[Olaf Lehmann]

Have finally managed to solve the link issue to YouTube from our website.

I needed to add the following at the end of the link address: /" target="_blank

Should work perfect now, both from any mobile device, and any browser.

Thanks Andre', it works.

Now that you have been able to recognize the leader. It would be create if we could follow the leader at the start. At least in to the first corner. That would be a create feature.

And if we also could hard code or configured that TV1 was always used in the start, heading in to the first corner. That would also be a nice feature.

Two good ideas, but only the second works.
I have upload the new version. In the complete first lap the camera type for near distances will be used (it's normally the tv1 view).

The other idea to set the focus until the first corner couldn't work correct.
The reason is: All the time I (it means the patch do it - I'm not so hardworking) compare all the time the number of laps and the driving distance from start/finish line. So the patch knows what is the leader.
But unfortunately it doesn't work correct at the beginning, because the number of laps is zero in the first lap and before the first crossing of the start finish line.
So the patch thinks the drivers directly before the first crossing at the start/finish are the leaders. The result would be more changing of the focus than before.

By the way. Personally I find it a little bit nicer if about the race sometimes the camera type changes from onboard to tv, but it'sa matter of taste...

BR Olaf

P.S.: I have rearranged my computer desc. So I can stand when I write/work. To sit is bad for my back at the moment.
My doc is chiropractor too. He is otiptistic he can make my frozen back flexible again. Tomorrow he will set his hands on...

[André]

Hi Olaf,

Good to hear that you have a movable computer desk. Working with some fun stuff makes the sick leave days better :-D

I will test the new version, and give you a feedback :-)

I did not fully understand the reason for way the leader focus on the first lap/in to the first corner wasn't possible.

You wrote:
But unfortunately it doesn't work correct at the beginning, because the number of laps is zero in the first lap and before the first crossing of the start finish line.
So the patch thinks the drivers directly before the first crossing at the start/finish are the leaders. The result would be more changing of the focus than before.

Did you mean the second crossing of the start finish line? Not at the start itself, but at the end of the first lap. If yes then I understand.

Because if the leader is recolonized directly at the start when crossing the start finish line the first time, then it should be possible.

Good luck at the chiropractor, hopefully he has some magic hands! :-D

BR André

[André]

Hi Olaf,

I have now tested the new version.

The first lap feature is working great, but I am lacking the control function for when to move from the near distance camera to the middle distance one. It seams to be hard coded to 32 meters.

Is it possible to get that back? Or is that breaking the other feature?

During this test I also noticed another thing. In Pribluda you can set the color of the leader, and that's already correct at the starting grid before the race.

So if you figure out how Pribluda can know that already then, then maybe we will be able to set the focus to the leader already at the starting grid.

BR, André

[Olaf Lehmann]

You wrote:
But unfortunately it doesn't work correct at the beginning, because the number of laps is zero in the first lap and before the first crossing of the start finish line.
So the patch thinks the drivers directly before the first crossing at the start/finish are the leaders. The result would be more changing of the focus than before.
Did you mean the second crossing of the start finish line? Not at the start itself, but at the end of the first lap. If yes then I understand.
Because if the leader is recolonized directly at the start when crossing the start finish line the first time, then it should be possible.

I mean the first crossing of start/finish!
Let's say the track is 5000metres long.
0L/4999m 0L/4998m 0L/4997m 0L/4996m
could be the values of the first four drivers directly after the start (before very first crossing of start/finish)
Next values could be:
0L/0.2m 0L/4999m 0L/4998m 0L/4997m
Now for the patch the second driver is the leader!
Next frame:
0L/1.3m 0L/0.1m 0L/4999m 0L/4998m
Now the third driver is the "leader".
And so on. So the focus will jump to all drivers after the start.
But with all old versions of rdpatch the focus changes after start too.
Now I found the reason.
I use the following algorhytm to find the shortest distance:
The patch compares all the time the distances between all drivers.
At first higher value - lower value. Then the value will be compared with the halv tracklenght. If the value is more than the halv tracklength then Tracklength - value.
There is no mistake in this algorhytm, but I found the track length isn't 100% correct. Maybe this value is the length of idealline or so.
So the distances at the start/finish line are not 100% correct. Later in the race nothing happens, but at the start all drivers have less distances. So it comes to the crazy flickering at the first crossing of the finish line.

But I found a solution...The patch does different things at different times. Now replaydirector_livestream works in this way:
PHASE1 / From the approval of the race (the moment the drivers can click at the green flag and go to track) until 35s: The focus will be set to forewardmost driver. So when all drivers are on track the focus will be at a driver in first startline.
PHASE2 / 35s-50s: The patch will do nothing. So the focus will stay on a driver that was in first startline.
PHASE3 / 50s-60s: Same like Phase1, so the focus will be set at the leader.
PHASE4 / 60s until the forelast racelap: The normal old algorhythm works, but as I wrote some days before the focus will be set at the leader when the adjusting values it says.
PHASE5 / Last lap: Same like Phase5, but the distance when the focus will be set at the leader is very short. So in most races the focus should be complete at the leader, except there would be overtakings or very hard fights.
PHASE6 / if the leader have seen the quecked flag: The same like Phase5, but the leader is only a normal driver, the new "leader" is the second driver. If the second driver has finished the third is the "leader"and so on.
It's a algorhythm like flying to the mars. It makes me perplex. Programmers loves simple algorhythm. But in this case I see no better way ...

BR Olaf

P.S.: You find the 42 now at the beginning of the file (seond sequence) again, Andre'.