Tactile Immersion - General Discussion - Hardware & Software

[Mr Latte]

Will check out the latest effects shared this afternoon.
One last question are any of these "G" effects including tyre slip or are you intending to do slip effects as well?

An example: wheelsip can be very active with no or little motion "G" loads

A burnout or low-speed corner exit, so while the "G Load" may not be that active or high the spinning tyre could be highly active and vary based on the car (FWD RWD 4WD)

Is your intention to add another effect or combine these "G" effects with the standard "Wheelslip"?

 

[blekenbleu]

You can also add your own group (folder) to name it and then drag/dop inside related effect layers.
Drag the effects with the icon to the far right.

By analogy, this employs SimHub to generate tactile responses much as does a synthesizer for music, with game telemetry replacing MIDI keyboard.

Take the perspective of a large LFE unit being used with low Hz to generate a deep response in feedback to represent higher slip and the additional loads.

A user could apply 4 corners based effect, but instead of it being routed to 4 independent channels, be sent only to one channel.

Certainly. One (IMO) serious constraint is that SimHub seemingly does not allow a single custom effect to generate different frequencies with the same amplitude or the same frequency with different amplitudes. For music, one can mix some channels to reduce output in a certain frequency band to "make space" for a lead or vocal in that band to be more intelligible, which seems not possible by SimHub (so far as I understand) always changing amplitude correlated to (optional) frequency changes.

My understanding of this is that the user then gets active responses from 4 wheels and with this it will compound the Hz used increasing the amplitude of matching frequencies being generated.

Yes, 4-corner will typically be generating 4 somewhat different (although related, by math and physics) tones. One could in theory rewrite this effect into 4 separate effects, so that an identical value calculated from telemetry could generate different tactile signals. To some extent, that could also be accomplished by performing very different math calculations for each channel in a single effect. For example, output values for some corner could be inversely proportional to telemetry signal values.

A personal constraint is that, while having a somewhat trained ear for music, haptics perception are relatively poor; gradual changes in tactile intensity or multiple simultaneous frequencies from one source are poorly discerned.

 

[blekenbleu]

One last question are any of these "G" effects including tyre slip or are you intending to do slip effects as well?

ALL of the effects have "G" (loading) properties modulating slip.

An example: wheelslip can be very active with no or little motion "G" loads

Indeed. the design is that, with "G" magnitude = 0,
conditioned slip values should be at 1/4 strength per corner.
This will yield a silly result e.g. when sliding on ice, since IRL tires will not squeal,
but will by this approximation. On pavement, slip with zero "G" is unlikely.
Consider that "G" (and loads) are based on changes in speed and/or yaw,
which require some forces (Force = mass * acceleration),
and cars vary speed or yaw mostly by forces between tires and surfaces.
Brief and relatively minor changes in speed or yaw
provoked by accelerating masses within cars are not considered;
rigid body physics are assumed.

A burnout or low-speed corner exit, so while the "G Load" may not be that active or high the spinning tyre could be highly active and vary based on the car

Effect values are products of estimated loads and conditioned slips.
If it is estimated that some tire has zero load (e.g. front wheel of a dragster in a wheelie),
then zero signal. So long as there is some load, then slip generates non-zero signals.
Slip telemetry values can evidently become very large and dominate loaded values.

 

[Mr Latte]

Just a brief check for now, but seems the "Harness Tensioner" effect is stronger than the others.

I am getting very little output on ACC with the other effects, having to increase alot the "Min Force"
value. Will try in AC as well but I don't think it and ACC operate that much differently?

Any way you can increase the levels?

The High Hz you apply for the effects, when used on my TST unit, generates a warbling sound like a ZX Spectrum game loading under water.

 

[blekenbleu]

Just a brief check for now, but seems the "Harness Tensioner" effect is stronger than the others.

I suppose that by "Harness Tensioner" effect is meant
the first "ProxyG" properties SimHub ShakeIt CUSTOM EFFECT.
That only becomes "Harness Tensioner"
when processed by some Custom serial device message Javascript.

That first CUSTOM EFFECT generates property values
which estimate Lateral and Longitudinal accelerations
from changes in SimHub speed and yaw properties.
Changing its Filter values (gain, etc) can increase estimated accelerations,
which should cause other CUSTOM EFFECTS to calculate stronger loads, hence loaded slip signals.

I had not considered outputs of that first custom effect to be directly useful for tactile effects.
Its output could easily be stronger and certainly more nearly constant than that from subsequent CUSTOM EFFECTS, which also depend on significant amounts of slip to generate strong signals.
At least with AC, strong slip signals can be VERY strong, but quite infrequently,
depending on how one drives. Most of the time, with settings as provided,
resulting signals in Dayton pucks approximate tire squeal from AC game audio.

I am getting very little output on ACC with the other effects,
having to increase alot the "Min Force" value.
Will try in AC as well but I don't think it and ACC operate that much differently?

No idea..

Any way you can increase the levels?

How brave do you feel?
In expressions for channels in the bottom two CUSTOM EFFECTS
can be found "Math.log(1 + s)"

• replacing "Math.log(1+s)" by "s" will substantially strengthen slip signal contributions.

The High Hz you apply for the effects, when used on my TST unit,
generates a warbling sound like a ZX Spectrum game loading under water.

That warbling results from amounts of white noise, which can be increased, decreased
or unchecked (yielding quite boring tones less like tire squeal)..

Sorry, but I am away from my gaming computer for awhile, so somewhat limited in ability to help.

On reflection, AC slip properties were never directly histogrammed, so that is something to do.

 

[Mr Latte]

No rush,
We maybe should move this to a thread just on the subject of these effects and tests.
I dont feel the need to go above 200Hz to generate finer detail, the noise aspect being generated to try to emulate tyre squeal, isn't floating my boat.

If you have an effect that works well in motion or for seat tensioners, then why could it not be used with suitable Hz to generate a feel that works in tandem with the belts?

ACC Test:
Im back in testing Lat G (Left/Right) but something is weird with how the Response Curve is working
Even at idle, It's active from upto 7 when the car is not even moving!!!

The effect is nicely active upto Mid 20s with the highest values working up to low 30s and above mid 30s it is not really active hardly at all.

Other than that the operation of the effect seems good.

Appears AC has a similar issue with the response curve operation. Its not working like normal in the way other effects use it.

 

[blekenbleu]

Since math and physics for these is at best dubious, the prototype profile will just be attached here.

Attached profile has been updated:

• typo resulted a wrong slip for one effect channel
• slip signal values have more useful conditioning
• simplified tire load math and physics

 

[Mr Latte]

Do appreciate your continued efforts.
Trial and error isnt a problem

Please though link or attach the newest file in the above post, saves any confusion.


Today I have been comparing the large BK LFE to the Earthquake Q10B
Also, have the Multiplayer bracket attached for the Earthquake unit and the Buttkicker levered via a BK Couch bracket.

TST T429 can be seen hanging below.
I decided to test and compare these 3 Premier competing transducer models before furthering my own build and making some additional improvements to my original installation plans.





Taking measurements for comparison data references via iPad Pro



Seems quite accurate but I will build a proper testbed for comparing tactile for future Youtube videos. Also have various software and hardware options to compare test data.



Not a definitive test as yet but lets me experience their performance differences and characteristics with different frequencies over a really nice new RPM I was working on.




Boring drones for engines are not for me. This has quite a nice mix of very low and midbass frequency layers for this RPM. It comprises excellent idle, inertia and low rev grunt with specific peaks in felt sensations for the cars green and red RPM indicators and finally with the car at overboost.


Initial Impressions:
Q10B seems more musical with having a nicer roll-off in its detailing but large BK seems king for felt energy with frequencies below 20Hz.

 

[Mr Latte]

Attached profile has been updated:

• typo resulted a wrong slip for one effect channel
• slip signal values have more useful conditioning
• simplified tire load math and physics

Okay I went back to the attachment, which I assume is the replaced file?
Just had a quick 30 min session. quickly found good results with some tweaks of frequencies and getting acquainted with this Q10B.

I have a light feel come in to start then a mid feel for increasing strength and a 3rd layer that kicks in only with a high slip angle or the car in a spin scenario.

Nice progressive sensations coming through for understeer/oversteer scenarios as well as burnouts or donuts. Was using 60Hz to 5Hz over the 3 layers applied, each layer with its own threshold and min values.

Nice work in the code, this has good potential from this quick test.

 

[No Depth]

So I’ve been reading this massive thread and feeling a bit overwhelmed.

I have an 80/20 profile rig with an NRG Prisma seat and soon to be getting a Next level motion V3 mover.

I would like to add a Buttkicker Gamer 2 to attach directly to the Motion V3 and a Mini LFE to mount under the pedal deck. Ideally both driven by a Stereo amp to a USB sound card.

Looking at this one which is referenced here and there but have concerns.

S.M.S.L FBA_SA-98E - Black Accordion Accessory (SMSL SA-98E) https://www.amazon.com/dp/B00JESRNK...abc_AP8CTD6ET2G16QM0JDZ3?_encoding=UTF8&psc=1

Namely the Gamer 2 outputs at 2ohm but this amp seems to only support 4-8ohm impedance? What danger is there to having this drive both Buttkickers?

Ideally I want to use an amp of this size(an actual volume dial!) but also to mount in my rig for easy adjustment access. My rig is on casters and i need to move it around when not in use, so having the amp live on the rig means I’m not interested in a rack-size amp solution. Curious on thoughts on this setup?

Really appreciate any help and answers! Thanks.

 

[anton_Chez]

At the risk of being that guy, any chance we could have a peak at that new profile you've got going there, @Mr Latte ? I've pasted together a couple of effects from some of your older tests and I've got it feeling not too bad. Always keen to see what progress has been made.

Not sure what info or intel I can pass on from what I've been finding with my own setup to help you on your journey, but if there's anything you want to know about the Mini LFE + MQB-1 setup I have, let me know.

One thing I can think of off the top of my head is the engine idle seems to miss the idle revs in the cars I've tested in ACC. Even the 488 that it's been designed with seems to completely miss the rev range and is dead when sitting idle in pits. I've managed to alter the curve to get a response, but it's weird that it's completely missing. I seem to be missing an effect for max RPM as well. The effect is there, it's enabled, the hz are in the correct range, but the engine completely peters out once the rev limiter is being hit. The curve looks like it should provide an effect, yet it doesn't. The units just go silent once the revs have been max'd out.

I'm also using an old gear shift effect with the clutch effect. Compared to the newer gear effect it felt a little more rounded out, so I chose to copy paste this one into the 488 profile.

 

[blekenbleu]

With minor success going to my head, lateral tactile energy to simulate acceleration traction loss tire squirm will be attempted. The relatively cheap (but obsolete) Visaton BS130-4, which may be a Dayton BST-2 with delusions of grandeur, claims relatively low (25Hz) resonance.
Presumably, little power should be required to nudge a swivel office chair, which is for the best, since motherboard sound card output appears to be already well down by 25Hz.

 

[blekenbleu]

Smart too late

I have a light feel come in to start then a mid feel for increasing strength
and a 3rd layer that kicks in only with a high slip angle or the car in a spin scenario.

I only just now discovered that one can clone CUSTOM EFFECTS!
.. which is presumably how you create multiple layers.
I suppose that causes multiple identical JavaScript expressions to run, which seems wasteful..
Re-reading SimHub Shakeit Bass Shaker V3 BETA - Discussion helped...

Having options for auto-calibration and auto threshold on CUSTOM EFFECTs would be great,
particularly for the proxy_G properties, since peak values can vary so widely among cars.
Being able to include renamed CUSTOM EFFECTs in the Add Effect list could also be nice..

 

[Mr Latte]

I have yet to still test some of the effects on that latest release.
Will maybe get a crack at it tonight.

The Visaton unit I think is very much similar to the Monacor AR50
I have it but wasn't that impressed with it.

In my experience, the only way you properly get the delivery and power of the lower Hz is with heavier pistons and lots of watts.

 

[Mr Latte]

@Anton
I will maybe try to come up with a small effects batch for RPM in AC / ACC to suit the common units.
Custom engines are made to work with the car they are tested with. IF we try to have a more general RPM then you lose out on matching things like idle or peak rpm fitting in with lots of other cars.

Even with GT3 cars that are similar you still find differences with the changing gear rpm and max rpm. It feels odd if a peak sensation doesn't tie in with the car. Although pretty decent.general RPM is possible with only 3-4 layers.

At the risk of being that guy, any chance we could have a peak at that new profile you've got going there, @Mr Latte ? I've pasted together a couple of effects from some of your older tests and I've got it feeling not too bad. Always keen to see what progress has been made.

Not sure what info or intel I can pass on from what I've been finding with my own setup to help you on your journey, but if there's anything you want to know about the Mini LFE + MQB-1 setup I have, let me know.

One thing I can think of off the top of my head is the engine idle seems to miss the idle revs in the cars I've tested in ACC. Even the 488 that it's been designed with seems to completely miss the rev range and is dead when sitting idle in pits. I've managed to alter the curve to get a response, but it's weird that it's completely missing. I seem to be missing an effect for max RPM as well. The effect is there, it's enabled, the hz are in the correct range, but the engine completely peters out once the rev limiter is being hit. The curve looks like it should provide an effect, yet it doesn't. The units just go silent once the revs have been max'd out.

I'm also using an old gear shift effect with the clutch effect. Compared to the newer gear effect it felt a little more rounded out, so I chose to copy paste this one into the 488 profile.

Pointless sharing it with people not using the larger units capable of producing under 10Hz
I can have a try with doing RPM to work more with popular units in the 30-80Hz ranges but at the moment the focus is on this wheelslip and loads.

 

[blekenbleu]

In my experience, the only way you properly get the delivery
and power of the lower Hz is with heavier pistons and lots of watts.

My impression is that heavy piston units are unhappy with horizontal mounting...?
Absolutely, for some spring constant, lower resonance frequency requires more mass,
and accelerating more mass requires more power.
For a swivel chair, horizontal spring constant is roughly zero.
I'm seeking only low amplitude effects;
large amplitudes not matched in steering might feel confusing.

 

[Mr Latte]

I never spent much time trying the BK horizontally as it's recommended to mount them vertically.
You don't necessarily need large amplitudes to feel low Hz. It's upto the individual what level of output suits their primary role/purpose and user preference.

Although giving more substance to many effects with combined layers using low frequencies certainly richens them.

 

[blekenbleu]

One (IMO) serious constraint is that SimHub seemingly does not allow
a single custom effect to generate different frequencies with the same amplitude

A work-around:

• change Frequencies to based on Input, instead of Output
  • I wrongly supposed that Input meant to expressions; it means input to Response filter
• Crank Response Filter Gamma Factor to 5.00 or so
  • this forces non-trivial amplitudes to nearly constant, allowing more clearly sorting loaded corner(s) beginning to slip.

 

[Mr Latte]

Smart too late

I only just now discovered that one can clone CUSTOM EFFECTS!
.. which is presumably how you create multiple layers.
I suppose that causes multiple identical JavaScript expressions to run, which seems wasteful..
Re-reading SimHub Shakeit Bass Shaker V3 BETA - Discussion helped...

Having options for auto-calibration and auto threshold on CUSTOM EFFECTs would be great,
particularly for the proxy_G properties, since peak values can vary so widely among cars.
Being able to include renamed CUSTOM EFFECTs in the Add Effect list could also be nice..

You should put in a request on the Discord to have such added.
From the latest file shared, I have only so far tried the corner-based effect but I was running all 4 wheels to come out the same channel.

Later at some point, I can compare the frequency/output differences 4 corners generate to a single channel to using only mono mode with the effect.

What I will do is offer up a variant using your effect with 2-3 layers that might work not too bad on typical units. I do wish the response curve was working better for finer control with these but maybe that is something that can be improved?

Will also look into the other front/rear and left/right options.

 

[blekenbleu]

I do wish the response curve was working better for finer control with these
but maybe that is something that can be improved?

Depending on what wants finer control,
perhaps Autocalibration would compensate for differences in cars, tracks and driving styles.
Meanwhile, adjusting proxyG CUSTOM EFFECT gain so that it occasionally reaches 100
should help the other effects.
Considering differences between 4 separately driven pucks
and mixing those signals to mono in several layers for a serious LFE transducer,
I continue to be amazed that it translates at all usefully.