Thank you very much indeed for your involvement Gerard.

Using FFTF with a 4K TV @30Hz, frames internally limited at 30, I am seeing a real improvement on busy European airports with the FS Labs Airbus and on ORBX regions when flying VFR.

I've made it simple using the AGL height scenario at [0-5000] and [0.05-0.33].
Quite pricey but very smart addon IMO.

Can someone explain to me what this tool does and why I would find it useful?  It seems a little complex so I don't know if I would benefit from it.

38 minutes ago, David Roch said:

Thank you very much indeed for your involvement Gerard.

Using FFTF with a 4K TV @30Hz, frames internally limited at 30, I am seeing a real improvement on busy European airports with the FS Labs Airbus and on ORBX regions when flying VFR.

I've made it simple using the AGL height scenario at [0-5000] and [0.05-0.33].
Quite pricey but very smart addon IMO.

Which European airports please

1 hour ago, David Roch said:

Thank you very much indeed for your involvement Gerard.

Using FFTF with a 4K TV @30Hz, frames internally limited at 30, I am seeing a real improvement on busy European airports with the FS Labs Airbus and on ORBX regions when flying VFR.

I've made it simple using the AGL height scenario at [0-5000] and [0.05-0.33].
Quite pricey but very smart addon IMO.

Thank you for your kind words David.

Enjoy the flight ...

Because we are talking about running another process in parallel with the sim there are considerations that some might not have realised.

Let's first look at the four core guys - very often they are convinced that they get a better ride reserving one core, usually core zero. Why is that? One reserved core will be talking hits from other processes which does not have part of the sim on it.

After a lot of testing I found that using all cores of your CPU for the sim does one important thing - it shares every core with the system. With all cores used by the sim the results become indeterminate.

Upshot? If you have lots of cores leave one at least for other stuff.

Should it be core zero. Doesn't matter.

Hi,

I know this thread is quite long and feel there is a risk that a Moderator may be tempted to intervene before it becomes too unwieldy. This student is trying very hard to acknowledge his shortcomings and is making an effort to act accordingly! Sometime within the next few hours he intends to bring an apple for his teacher along with a list of requested Scenario loading times recorded under controlled conditions. So, I would be grateful if the topic is allowed to run for the time being. Also, I’m quite sure there must be others who wish to add their contributions, positive or otherwise, regarding their experiences and the effectiveness of the FFTF Dynamic utility.

Thanks.

Mike

Mike, You can measure the time taken to a datum, the point whereby ATIS starts for example. It's more complicated to measure how much terrain beyond the player is filled in. You should see increases in loading speed as you add cores, then as you double up LPs on cores in HT mode you should continue to see small increases. Remember though that a tiny increase in loading speed for the use of an entire core might be counter productive. Hence I mentioned earlier that  there's an optimum core count. I'm pretty sure we don't need a masters in comp science to work on this.

9 minutes ago, SteveW said:

Mike, You can measure the time taken to a datum, the point whereby ATIS starts for example.

Hi Steve,

I believe I may have found a better way, a way which ensures reliable accuracy of measurements thereby ensuring valid comparisons. Three guesses? In actual fact, you should only need the one

Regards,

Mike

5 hours ago, newtie said:

Case closed and sealed.

Amen..........

7 hours ago, Cruachan said:

Hi Steve,

I believe I may have found a better way, a way which ensures reliable accuracy of measurements thereby ensuring valid comparisons. Three guesses? In actual fact, you should only need the one

Regards,

Mike

I see. Well, the graphs I showed are from the stock P3Dv4 about 30 second load time. If you have a shedload of addons and you are looking at minutes then you won't see much in terms of how the different cores are handling things. You'll be looking at a redline. There's no trick involved with the sequence of graphs I did from the stock sim, an easily repeatable result. The graph show the settling time over two minutes as less scenery is loaded at less resolution as it gradually compiles the viewdata farther away from the player. Running a test setup is not easy and can't be rushed into. Imagine trying out a wind tunnel for the first time.Good luck!

6 minutes ago, SteveW said:

I see. Well, the graphs I showed are from the stock P3Dv4 about 30 second load time. If you have a shedload of addons and you are looking at minutes then you won't see much in terms of how the different cores are handling things. You'll be looking at a redline.

Hi Steve,

You are right.....on several counts. Once again, back to the drawing board! Returning to stock is a major obstacle right now so I think I’ll pass and concentrate once more on trying to find a good alternative AffinityMask. What fun!

Thanks and kind regards,

Mike

On 25/03/2018 at 8:13 AM, GSalden said:

@Mike : regarding the amount of cores

Record a flight of 4 minutes over a terrain with a large city and a rural area.

You can play it back with a different amount of activated LP’s and run Fraps for 3 minutes to measure you min/max/average framerare.

Then you can test and see what is best for your system.

Hi Gerard,

Good shout. Thanks! I’ll certainly incorporate your suggestion, along with those from Steve, in my testing scenario. My apologies for not having responded before now - I’ve been somewhat distracted. Steve will understand 🤓

Cheers!

Mike

16 hours ago, SteveW said:

 

After a lot of testing I found that using all cores of your CPU for the sim does one important thing - it shares every core with the system. With all cores used by the sim the results become indeterminate.

Upshot? If you have lots of cores leave one at least for other stuff.

Should it be core zero. Doesn't matter.

I'm on I7 6700K 4 Cores HT On 8 LPs. Currently for P3Dv4.2 I use AM 11.11.11.01 Everything else related to P3D I put on 10.10.00.00 each. 

So are you saying 11.11.11.11 for Prepar3D.exe may work even better in my case? Did I get it right, Steve?

Thanks

On 26/03/2018 at 4:47 PM, tooting said:

Which European airports please

Essentially Fly Tampa's ones which I find excellent (LGKR, EHAM, LOWW) but also some nice others from 29Palms (LGSM) or Taxi2gate ones (EDDM, LFPG) or JustSim's EDDH and LFMN. There are also some nice ones for Greece and Spain from Aerosoft or others serious folks.

I've almost stopped flying heavies in the US, only VFR, because it kills my fps.
In Europe my flight experience is smoother, less traffic, less roads, less accuracy and details also probably and my PMDG + FSLabs birds feel much better!