Cool sounds good. Thanks for clarifying!
These days how do people know what is the proper NM setting per car in iRacing?
I remember David Tucker saying that he had those numbers, but I do not ever remember seeing it posted.
I am a little behind on the info at the moment as I was not as involved for a bit.
I am back now and eager to catch up! 
The thing with iRacing is that it doesnāt matter what the āproperā Nm is as that would depend on the person driving the car to know how strong they want the feel to beā¦ The Way I racing works is that all cars output is based on telemetry so if the cars model puts out 10Nm it will put out an output of 10Nm if it is 60 it will put out 60ā¦ Then the slider alters that output by a % which I call the Specific output which can be pretty much anything including over 1:1ā¦ So lets say you have the Small Mige and you have Max MMC of 12.86 that would mean the wheel is capable of 20NMā¦ So if you put the racing slider at 20Nm the specific output would be 1:1 and any thing in iRacing 20Nm and below output from the telemetry would be sent to the wheel at what the telemetry says and if the output is above 20Nm then iRacing internally will clip the signal to 100%ā¦ This is why iRacing has the force meter.
So in testing I have found that you can run Just about ANY car within iRacing and have full fidelity with the iRacing slider at about 50Nm, the strongest cars will still clip a bit at this pointā¦ so for a 20Nm Max Power Mige that would be .4:1 or that all output will be 40% less than 1:1 full power. So a Miata may put out 4Nm at the wheel where a 488GTE will put out about 12Nm and the DW12 would put out about 18Nm generally giving you the difference in power feel realtive each otherā¦ If you had a Large Mige at the SAME racing setting the Specific output would be about .558:1 or 55% so the relative strength at the same setting is higher because the Servo is more powerful.
If you like the strength higher you go for a specific output closer to 1:1 so if you ran a Small Mige at 33nm on the slider it puts the specific output at .606:1 but if you rant the Large at 33nm the specific output would be .845:1, but of course running at 33Nm would cause all cars that output beyond that to clip at 33Nm telemetry outputā¦
So this all sounds complex but I think with a picture that Mika Posted all this may get easier as it looks like we are getting an estimated wheel output Number along with MMC. Once we have that Number then it will be really easy to figure out your specific output in iRacing as you can just divide the estimated output by iRacing slider to get the specific output.
Optimally if you like to feel the physical difference between the cars with iRacing you would set the wheel to a specific output you like and then drive any car and it allows iRacing to alter the strength and feel accordingly. Otherwise you could use a different specific output for each vehicleā¦
I generally run at almost .6:1 but that makes the Miata really light so i will tend to boost that car to 1:1
I wonāt say much about iRacing FFB, except that it has the same resolution as any other sim out there, because they are all limited by the DirectInput layer in windows. That means that they are limited to +/- 10.000 values.
So, back to AC which is the sim Iāve explored the most, and the internal resolution of that FFb signal is 64 bits. Think about itā¦ You have 64bits of resolution, pushed into a 14 bit signal. So, bsohn, you do not need to worry about resolution at all from AC, at any gain setting. However, running ac gain at 100% WILL produce clipping in the vast majority of cars and track combinations. I logged the values from the sim at 10% gain, because I had the same concerns you did about resolution. Over a lap of nords I logged almost 140.000 values, and NOT A SINGLE VALUE WAS EVER PRESENT MORE THAN ONCE. Out of 140.000 valuesā¦ This was at 10%. This show you that the resolution of the ffb signal in AC is very much there to run any gain you want without ālosing fidelityā. You WILL lose fidelity by running gain at 100% because you WILL get clipping.
And just to be sure you understand it, iRacing is not magic software that can get around limitations in windows and hardware. There is a finite resolution in there, again limited by the DX layer, not the sim itself I would imagine. And it would be very interesting to run a proper blind test to see if people could detect a difference between 10.000 values and 256 values of torque. I think youād be surprised at how difficult it would be to separate them.
We are at a point in history where the software and hardware resolution is WAY above what our meaty paws can discern and make useful, so Iād say letās not worry about resolution, letās worry about filtering and what actually comes INTO our hardware. That is a much more useful discussion.
Sort of. Lord Kunos ālikedā the post that outlined it, which Iād say is a confirmation. 
Also AlteDreier might know more details about this.
I talked to him about it hes my goto man for ffb in AC
iirc, iRacing doesnāt use DirectInput.
It does. It uses Constantforce effect type for torque commands.
Doesnāt change the fact that 14 bits is plenty resolution for smooth FFB. The hardware is 14 bits, as confirmed by Mika in this very thread, so anything above that is wasted anyway, and could potentially lead to aliasing, harmonics and other undesirable side-effects that you get when dealing with analog stuff. The real world is a bitch sometimes.
hm, I thought thatās why they need special calibration for the pedals because they donāt take the DI values into account (tbh, I never understood the point of this).
Could be though that pedals and wheels are handled differently, of course.
They are usually on the same device. Itās most likely to ācatchā any funky calibrations out there. Idiotproofing if you will.
No it canāt get around the limitation of windows and I think that is where the point is being missedā¦ Most likely all sims have to send a signal through the windows Direct Input layer which would limit their resolution and intern the maximum fidelity that can be achieved.
Not sure why AC would clip at 100% gain unless they intend for their car models to clip but that seams weird to meā¦ I am not familiar with how AC does stuff internally but I am aware generally of how the iRacing system works in relation to the cars and the way it clips and uses its telemetry.
The point is that we are no longer going to be losing fidelity WITHIN the SimuCUBE. Whatever, is sent from the software will now translate in itās entirety through the system at 1:1 EVEN when scaled in power because we are no longer messing with the FFB signal Amplitude we are changing the actual Wattage capability of the servo. This of course as mentioned earlier also means that Filtering by the drive will be applied more consistently. So it is not so much about how it relates to a certain software title it is how it will relate to all titles and that translates to no loss of fidelity regardless of what the initial fidelity is coming into the SimuCUBE and that is big.
On a note, you would be surprised at what would be detected, as lower signal reading 256 instead of 10000 is detectible as smoothness of the wheel. The Same as having a 12 Bit or 16bit pedal set transmits more accurate braking than an 8bit, and a Higher count encoder 2.5K to 10K to SinCOS increased the smoothness of the servo transitions. So it is noticeable but might not be to everyone.
Of course the 8 bit resolution was a gross overstatement, but it got themessage across.
I see your point on not losing the resolution within the SimCUBE, and itās a good thing to preserve it, noone is contesting that. I get a bit riled up when we spend weeks chasing unicorns because of a 0.0044Nm difference, is all Iām saying. That time could be spent developing the recon filter or alternative methods that ACTUALLY make a difference.
Regarding the clipping in AC, that is as designed. The devs are using Vallelunga Club as their baseline FFB normalization. This a track that is short, slow, flat and smooth. So any car driven on any track with bumps or elevation changes, or aero that works more at higher speeds than Vallelunga Club WILL clip at 100%. This is information directly from Stefano, I can find the post if you need it.
Just curious how is the development of 081 going ?
its progressing. #verysoon.
Still fighting with that āgrindingā feeling under load in corners on my Lenze with Simucube.
I am considering going back to my Argon OSW it bothers me that much. Feels just horrible.
I would like to know if this is going to stay forever because Ioni is designed like that or if there is hope a coming additional filter is going to combat this unpleaseant grinding feeling. Again my Argon OSW with MMos does not produce that grinding at all.
If somebody doesnāt understand what I mean: put the dampening Filter in Simucube configurator tool to 8% and move your wheel. That āgrindingā is what I am feeling in every corner under load.
None of the existing filters nor the TBW value does help.
Can you also tell which encoder you have on your system?
As far as I know there is only the original Lenze encoder for the MCS12H15L, 16384 steps.
But a friend uses the small mige with 10k encoder and he is feeling the same grinding he says.
Hi @McErono. As perfectionists, weāre curious of this grinding observation. We will send you PM and ask necessary details so we can try to replicate and understand the issue in here.
Very nice Tero! I will do what I can to help you replicate the issue!
Hi,
Can you test if you can repeat the issue also when just using/pushing the wheel against the bumbstop, or is it just when driving? Is there a specific simulator that produces more or less of the griding feeling?