Is there a plugin already that changes a single axis L0 line according to the the used other axis? To my understanding to not slip out L0 has to be reduced in certain areas if an axis other than twist is used. I am sure this can be calculated. Would make Multiaxis compatible to single axis if multiaxis scripting would derive from full L0 that would always exist. Would also make it easier to multiaxis script a single axis script if reduction would be automatically done. L0 and L0 MA. Best would be if L0 would be the same for single and multiaxis but that would involve calculation in the player like MFP I guess.
Never scripted mutiaxis other than twist, I might miss something, so please enlighten me. Seems to me MA is not only taking abysmal time to script but also makes it unusable for single axis users if the scripter wont do a full L0.
MFP has smart limits which will reduce the scripted limits of other axes you configure. For example you can set it to reduce the pitch or roll by 50% when L0 is in the far up or down position but have full range of motion inbetween.
I actually just started making my MA scripts without worrying about limits recently after spending some time properly calibrating my device. Because of the differences in devices/anatomy/mounts I think proper calibration should always be done on the user side and scripters should feel free to use the complete range of motion 0-100 for L1,L2,R1, and R2.
R0 is a little special because some users have more degrees of motion than others.
So as a scripter I think you shouldn’t even worry about limiting L0 length at all, just follow common device speed limits or interval limitations.
Yes (but not yet released) I have created an extension that stunts L0 depending on R1/R2/L1/L2 as well as selectively compensate R1/R2 movements to L1/L2. Will clean up and release within the next few days.
Yes indeed, me and @Yoooi have been in discussion around this. The difficulty is that parameters for the device’s and sleeve’s geometry are necessary to do this, which with the individuality of the devices becomes a huge thing pretty quickly. And maybe there are some other structural things, that make it difficult, but I haven’t looked exactly at how MFP handles the signal path.
Yes, that is a very helpful feature, but limits the angles and deviations possible at the tip in order to not slip out. The other way round, i.e. script the angles/deviations and then stunt L0, such that it does not slip out despite the angles allows for a much better experience. For me at least ![]()
This probably will lead to a better experience no doubt, but I also think it’s a bad move to implement this in the actual creation of the script. This should only be implemented on the software level when the funscript is translated to the device, such as through MFP or other program.
Otherwise you wind up with a script that does work great with SR6 because L0 was stunted due to high surge/sway but isn’t as good in OSR2 which is missing those axes, and really bad with SR1 or the handy where stroke is the whole thing.
The scripts stay highly compatible if we keep limits away from the scripts themselves and implement these limits in the software instead.
Fully agreed. As long as that is not the case, the scripts themselves have to be adjusted though. By starting this discussion with an extension and a few adapted scripts I’m hoping to get the ball rolling. Because until now, it isn’t really a problem, that’s on the radar.
Hopefully we can get some smart heads together and discuss what a good implementation could look like. Especially with more devices coming out that are different in key areas (arm length and firmware parameters influencing the stroke length of the FunOSR Pro for example) and the individual nature of these devices in general.
Indeed. Yet I would say that compatibility is suboptimal at best in any case. If you script for a OSR/SR6 device, then many details and intensity might be on the additional axes. The classical example is the twisting motion at the end of a handjob. Firstly, the counter-movement for the single axis script does not belong in a multi-axis script and secondly the position on L0 might not always be the same. And that’s generalizable to any other 3D movements that are emulated onto a single axis.
I strongly disagree on this one. Its the same discussion as low mast tip scripting. Again it is about the position of the lower rim of ANY sleeve with ANY geometry. All sleeves have a lower rim. That nonslipout ring is set by the user low or high. The rest is to be the same for everyone.
It is all about the ring position and angle, not the user, not the sleeve not the device.
If you can downcalculate a singleaxis L0 to work with more axis so could Yoooi. Which would be a lot more preferable.
This part is near impossible to get accurate. Just because dick length and positioning of the user matter a lot. A device, even if it knows the sleeve does not know how it reacts here.
lying down vs sitting make a lot of angles behave diffirent, in sitting your dick usualy points more in the direction of your chest, while in lying down its often straight forward. This already sets a movement limit just from this starting point.
Add to that that some sleeves have a thin ridge, and some a thick, this can easily be 3cm in just fleshlights alone. While this could be programmed in as static parameter list, for onaholes this gets problematic quickly due to sheer number of options, and the way they are attached.
For this the only way it could work, is if the user would in some way be capable of aligning the sleeves for its deepest point, combined with a numeric value for dick length. Then regardless of sleeve it knows the movement length, and because you calibrated the sleeve, it knows this.
And then someone figures out a new positioning system or whatever, and a new issue gets found that requires even more difficult adjusting.
To me, basic profiles you can have in MFP as dropdown, which you can create/adjust/duplicate/etc yourself is the most versatile. Its not the most convenient at first, but its reliable.
You are misunderstanding my point.
- When you engage angle axes, the sleeve will not rotate around the sleeve entry point. Therefore if different sleeves have different protrusions measured from the rotational axes at the fastening ring, the corrections have to be different.
- If your sleeve entry is asymmetric, then you need to have more safety margin, because especially with twist, it can be in any position and thereby lengthening the stroke even further.
- Individual dick-sleeve geometry, stiffness and other properties can make it easier or harder to slip out depending on the angle, even if they are theoretically touching at the exact same place.
- Different devices have different motors and configurations. If one OSR can do 25°, but another 35°, then the script values mean different real world angles and therefore stroke lengths. The recently released FunOSR Pro also has different geometry (arm length, different pitch lever lengths on drive and receiver side) and therefore different axis ranges. To make things more complex, there are firmware parameters by which these can be further influenced.
There are other factors at play why there has to be correction, if you want to be able to have an non-zero angle/deviation at the tip, but these are the most relevant imo. And not everything has to be considered in a solution either.
In conclusion there are a lot of individual parameters originating from device or user and software and firmware have no knowledge of the settings of the other. Therefore the discussion should be around where the complexity should be integrated to get the best result with reasonable effort and compatibility in mind, as well as considering the state of most devices, manufacturers, firmware/software developers and scripters today as well as in the future.
Everything comes with a trade-off and with complex problems it is always good to have more than one person thinking about it. I too was once of the impression, that the player should have all the settings, but talking to Yoooi he outlined a few issues that should be thought about more globally. So until then I make an assumption about basic OSR2/SR6 geometry for my scripts, because it makes for a way better experience in the meantime.
Agreed. I think the most important parameter that is also somewhat easily possible to measure, is the protrusion under the rotational axis. The rest will probably have to be included within some combination of physical and software safety margins as well as Smart Limits, but I’m open to different solutions.
Correct. It is why I stated in my last script, that it is thought to be played sitting with an upright device. Since the devices’ range is pretty limited, I think Smart Limits at the bottom half of the stroke could handle the downward bend in a satisfactory manner. Theoretically one could “shift” the angles in software at runtime (pitch+surge+stroke) but with the tight device limits, I don’t believe this complexity to be necessary. Also: Tilting the mount would have a probably better effect.
I see where you’re coming from. I myself tend to prefer absolute numbers for parameters. That said, if there would be a possibility to save a parameter set like you describe, that would be of course preferable.
Honestly, this should be a feature of the device firmware. This would at least be the best option for open devices. Some already have axis limiters, but they usually just do the individual axes. It would be nice if you could have kinematic geometry limiters, like a bounding box for the device.
Maybe I’ll look into doing that myself if I feel like it and got some free time.
The ultimate solution would of course be to not control individual axis but the translation and rotation of whatever stroker you are using and handle the axis with inverse kinematics but that’s a bit more complicated.
But in MFP you already have upper and lower range. If you sit and the range changes you set a different range. You can even save it by macro.
I did not assume the lower ring is the turning point but the lower rim is the slipout point that has to be in consideration for all movement in a multiaxis script. That is what you do when multiaxis scripting I guess or what I would do if I would script multiaxis.
Multiaxis is already a bit of a dead horse due to the time it consumes but L0 being incompatible with single axis devices, which are undoubtedly the majority out there, is really bad and needs a solution.
I still dont agree the scripter needs to adapt to the sleeve or anatomy. The script should be for the standard symmetric sleeve and the user will adapt for his sleeve and size. He will have to lower the upper range if he uses an asymmetric one. If you try to adapt to all sleeves at once you loose a good tip scripting. And that is exactly what I correct in 99% of the singleaxis scripts I download.
What I am after with this thread is a regular single axis L0 that is used in the multiaxis. This should have been mandatory from the start and clever would have been a reduction line/axis that could be calculated automatically later. This way multiaxis scripting would be faster and converting a single axis to multiaxis would leave the singleaxis untouched also sparing time. Each axis you add changes L0 at the moment.
On this i actualy ment being able to somehow push the device in position, and then tell the device ‘this is the limit’ through that. This would be a sleeve independant way to get a reliable position.
Every time you switch sleeve you then need to calibrate it again. But as dick length is static, that should be a very simple task.
But its currently that the OSR does not provide any feedback signal here, and pushing on it in the current state would just damage the device.
A slider while it gets close, is still only an indirect way.
Yeah I’ve played with that thought as well. Especially also integrating some interface to help with the setup process.
That sounds like a second protocol in place of T-Code, which until now does the reverse kinematics after receiving the separate axis values to generate independent movements.
I’m also wondering whether Blender could be a better tool to script multi-axis. Basically that would amount to scripting the geometry possibly graphicly and then translating scripts from that. Not sure whether it would be faster than OFS with some help of extensions, but it might be more intuitive and have the same output structure you’re suggesting (basically quaternions) which could then be converted to T-Code.
Strong disagreement. The way I see it, more people are getting into it and the FunOSR Pro just accelerated that process. The future will tell, if that will increase the number of scripts or not. But outright dismissing the whole thing from the get-go?
Welcome to the world of Robotics! Where every added actor (if not in a linearly independent way) will change the drive signals for the existing actors to reach the same point in space.
So basically the way I’ve done it. My script was done with the parameters given, which is the standard TempestMAX device geometry and ranges, and an average value for the Fleshlights I own. The idea behind a discussion around this, is precisely to decide where to include this complexity, so we don’t have a billion script versions because of individual factors and have a basic, non-stunted version and the software(s) handle the rest at run-/playtime.
This simply results in sub-optimal Multi-Axis scripts. It’s a bit of a different discussion, but let me approach it like this:
There are movements, that you can’t map onto a single-axis, yeah? Classic example: The twist. Therefore, scripters add additional movements to the single axis script to emulate that, i.e. the “Fake Twist”. Why would I include that fake movement in my multi-axis script, if I have an axis that can accurately map it?
This holds true for many other movements, where you can either accurately script or emulate movements onto different axes than L0. All the additional details, movements and exaggerations would be missing, if you simply take L0 of a true and non-stunted multi-axis script and try to play it with only a single axis (that is what I’m adding right now to the Anna video). And the other way round, you add fake movements to real movements, which is also sub-optimal and takes away from immersion.
So no. It should not be mandatory to play the same script in single-axis and multi-axis or we might just as well not have multi-axis devices. Not saying that there isn’t a use-case for it, or even that there isn’t a basic script version which can be used as a base to then adapt to the axes you have, but to have that as a general rule, I vehemently disagree with. I can think of several ways to semi-automate the process from multi to single, but in the end both scripts will be fundamentally different.
The setup process should for sure be included in the considerations, since it’s seemingly not thought out at all until now.
Just a possible problem with setting up the device off-center of its range without shortening it: You run into the risk of colliding the top of the fleshlight against the body of an OSR for example. The servos could easily do more than 60°, but I believe TMax limited it for that exact reason.
Trying to not wall of text so its readable.
There is no fake twist. Good scripters dont script a twist it as it is not possible. Twist just often results in a pause in vertical movement and that is what needs to be accounted for or you loose immersion quickly.
Twist is an exception as in combination with L0 only, it does not provoke a slipout. So it is not interesting in this topic.
No, multiaxis script number per month is not growing and it will not grow by the Funosrpro monster.
Multiaxis scripting takes too fucking long and the risk of reducing L0 too much is too high. A bad macro that wants to do it right for all sleeves and devices wont account for it. Scripts will be just boring like not scripting the tip area right.
I still say:
- it should be possible to use the same L0 for single und multiaxis
- reduction of L0 can be calculated for each point by macro or by player.
- calculation by macro is sleeve and machine independent. Overall working ranges need to be set by user for device and sleeve.
- L0 needs maximum possible movement without slipout. Calculation for many different sleeves and machines will make L0 really bad for all of them.
- MA is dead horse if not made simpler to script (automatic L0)