Ultra Low Latency Mode, how it work?

You can adjust it higher until you start getting oscillation in your wheel, then back off maybe 0.5 or 1% and leave it there.

Due to high torque I am running, mine is at 8%, but for most guys, you should be able to turn it up to 15-16%.

1 Like

Can the new NVIDIA Reflex filter present in iRacing also affect this value? Can you try to lower it by finding benefits or does it have nothing to do with Ultra low latency?

NVidia’s Reflex things is not a filter, and is related to graphics, not game controllers.

I supposed.
Difficult this ULM filter, I feel the difference but I still do not find a definitive value (I can confuse the response of the sc with a response perhaps of the machine due to a wrong setup). Using iR and all TD filters at 0 so theoretically low latency, should I go down as much as possible with ULM?

It is a myth and misconception that filters somehow spoil the signal and increase latency. It is of course true when looking at outrageous / high values - for example low pass filter (torque bandwidth limit) will cause all peaks in the signal to be felt later or not at all, if the frequency in the filter is low. But modest values do nothing but improve the driving feel and remove artifacts from the game-generated FFB signals, thus it is crazy to drive a direct drive system with 0 filtering. Most of the “ripple” that you would feel like that is something that the game developer is not meaning to be able to be felt anyway.

With ultra low latency mode filter, you will be able to go a bit lower on the friction/inertia/damping filters without the wheel starting to oscillate too much. Such oscillation is an unwanted artifact caused by inherent lag in the feedback loop from game engine -> pc-> windows -> USB -> wheel -> wheel position -> wheel -> USB -> Windows -> game engine.

3 Likes

Let’s say that it is not the oscillations that worry me, but a linearity of the steering with a few degrees, because I tend to turn the steering wheel a little.

The filters do not affect steering linearity in any way.

Hey! Hope you dooing good
I remember (or think I read it
somewhere in this forum) that recon filter 1 does sort out signal above 1000 Mhz. This is what bandwidth limit is for AFAIK . Is this total nonsense? Some say that the less one filters the better the “true FFB intention from the game physics” comes through " while others are using huge amount of filters (the constantly operating ones) at full torque to rebalance all. Is it possible that a F1 wheel with 300mm and weighting 1150 grams will need total different settings? Thanks!

Scientia potentia est :nerd_face:

https://granitedevices.com/wiki/SimuCUBE_Firmware_User_Guide#SimuCUBE_Force_Reconstruction_Filter

1 Like

Me, I’m the “errare humanum est” type and quite often part of the “beati pauperes spiritu” fraction.:grinning:

Hello @Mika

Could it be that an erroneous value in ULLM causes instability in the signal and/or variable latencies?

The fact is that for a long time I had noticed instability of the steering wheel between sessions, something like the steering wheel felt lazy and in the next sesion it felt fine, the next day lazy and so on, until last week, preparing for the Bathurts 12 o’clock, I took a couple of afternoons using the WheelTest.exe program:

My first tests with the usual Profile gave me very variable latencies that ranged between 4 and 8ms, with 8ms being something common and 4ms an exception (I add that I have managed to understand wheeltest.exe and perform stable and representative tests)

After lowering ULLM from 15% (original of my Profile) to the current 3%, I have been seeing the same thing that I am telling you about, the latency that I obtained in the tests was stabilizing and reducing to the 4ms that I have achieved by putting a value of 3% in ULLM
Greater values cause instability on Latency ranges Up to 9ms depending on who knows, but just with a 3% value on ULLM I can achieve constant readings of 4ms every test

Lower values than 3% cant cause latency instabilities on my ende but the general latency increases

no, the latency is not variable when using the Ultra Low Latency filter.

I’m not very familiar with that software, maybe it does not measure latency in a compatible way on what the filter is doing.

1 Like

Must be something like that and I found the stability of the test and my SYS when ULLM at 3%, constants results and behavor, but more than 5% the app reads values of 0.0001ms, or ranges up to 12ms or wont star in some cases

By the way, looks like my instability between sessions are ended since I lowered ULLM at ttest values

Indeed, the instability appears to have increased. I had the Podium which suffered a lot from this in my opinion. I have had the SC2 for over a year, the thing I really liked was the stability, I stayed with the same profile for a long time and the steering response was always the same. Lately it’s not like that and I’m not talking about FF. These days I have had the feeling as if the first 20 degrees of steering are too sensitive, then instead it becomes insensitive and does not seem to turn. I don’t know if it’s me or what, but I’ve lost some linearity or stability.

Perhaps what I perceived was due to an error in the iRacing tires, perhaps with the P217 it was accentuated and led to indecipherable reactions in some situations. After yesterday’s patch and having tried various circuits, the steering wheel seems to be back more coherent and linear.
The mistake was this, I don’t know if it makes sense:
Tires

Overall tire rim friction has been reduced slightly. This parameter being too high was inadvertently causing tires to gain lift when rubbing walls, enabling easy wall-climbing

Mmm Rim friction against trackwalls is not something you do on normal driving situations…
It seems that metal2metal (rim VS trackwalls) friction was exagerated before

Then I don’t know the specific reason. Now the SC is back to being linear, predictable and constant. This is after iRacing update. It makes me think that in the last period some updates have not gone the right way. Maybe it’s just due to my way of playing, but for the moment it’s better and I like that.

1 Like

Maybe I need to start experimenting with ULL. I always disable it because the instantaneous torque, power, acceleration, etc. of these wheels can be so unrealistic (plus the fact that an active & position-velocity based system like these behaves completely different to a reactive & torque based system which is what real life is) so I therefore always thought ULL would make that even worse, but you’re saying that applying ULL can actually reduce the hyperness & aggression of wheel? You’d think a setting called “Ultra Low Latency” would increase it. Thanks for your input, I’m gonna have to play around with it now instead of just disabling it for everything.

1 Like

Please don’t do that! Your software is what makes the Simucube2 product so superior. I can imagine that some might have a lot of problems adjusting the device. On the other hand, once you get the hang of it and you have your tools… it’s super easy!

Please don’t kill the TBF!! Leave the True Drive software as it is!

May I ask where the problem with the software is?