Newbie (couple) wants to know ... why are hugs and kisses so darn hard?

[Katherine Heartsong]

Sigh. Like this hasn't been asked a million times before. Basically, we're newbies, and my sexy, loving partner and I just want to know one thing ...

Will someone, anyone, tell us the best walk together, holding hands, hugs and kisses animations/hud/method for a couple?

My partner and I are struggling mightily with that one. The ones you need to rezz a controller on the ground? They don't seem to work in 99% of places since you don't have rezz permissions. And some others, having to line poseballs up to get the animation anywhere near right? So frustrating as newbies because the whole 3D editing shapes in space thing and saving is a struggle.

I just wanna walk with my man!!!! Hold his hand. Hug him. Kiss him. 

Help!!??

P.S. With the help of cool furniture and some stuff from VAW we have the other stuff sorted out well enough.

[Maitimo]

Huds are never going to work accurately because couples aninations rely not just on your body and their body making the right actions, but also that you are both positioned in exactly the right place and orientated at exactly the right angle in relation to one another. The only way this can be achieved is with poseballs and furniture. Even then, they require both avatars to be exactly the same size as the avatars used when the animation was made, and no-one knows what those sizes are. 

There are "walk holding hands" animation sets that work like a vehicle, you can only use them on land where you are allowed to rez objects (which is almost nowhere). Otherwise, sorry. You can't really achieve that sort of thing accurately or properly in SL and what you've seen already is the best we've got.

Your best option is to look for kiss/cuddle poseballs that you can rez in your own home, or couples furniture with kissing and cuddling animations included. It doesn't have to be beds and sofas; you can get animations like this in almost anything from kitchens to trees. 

[Nalates Urriah]

Maitimo has it right.

Telling the computer how to align two objects of unknown size is very challenging. Seldom does it work well.

There are various ways to animate avatars and position them. The poseball thing has rez-rights problems, as you noticed, but allows for individual avatar adjustment. But, even then avatars look off as you can only line up on one point, lips for kisses, hands for holding, genitals for ... well, the floor or chair when walking or sitting. The Project Arousal system does calculations to line up two avatars but, even it has position adjustments. And if both people do not have RLV enabled it uses a type of poseball.

Some animators provide their animations in sizes. Usually S, M, and L. That helps some, but it is not perfect.

The best I have seen at lining up avatars are the dance machines and other objects that both avatars connect to. But, a 5ft avatar dancing with a 6ft avatar needs special animations to look right. Arms need to reach up for one or down for the other, AFAK no one does that type of animation.

Script and animation-wise it is possible to make scripts and animations that allow avatars to walk hand-in-hand. Adjustable, invisible prims can be placed in the hands and adjusted by the user. The script can calculate the relative positions of the prims and adjust the avatar position for the animation. As far as I know, no one builds such a system as it would be too complicated for most users and support questions would over whelm them.

I suspect the best you can do is demo animation HUDs until you find one that is satisfactory. There may not be one for you.

[Katherine Heartsong]

Perfect, thank you. Some of the ones we've tried work "okay-ish" so we'll stick with keeping on researching what works best for us. Again, ty so much for both your help!

[Lindal Kidd]

First, I recommend that you both acquire a hug and kiss HUD.  You do have to take care to face each other before accepting the hug request from your partner, but in general it's easy to use and an effective way of communicating your affection.

You can find a free one in the lobby of the Masocado Resort, in the College of Movement at Caledon Oxbridge University, and in my Marketplace store.

As for walking hand in hand, that is a little harder.  The most common solution is an invisible vehicle with two "seats".  One person rezzes the thing and sits on it, and then the other person sits too.  The first person is the "driver".  The biggest drawback is that you have to be in a location that allows rezzing.

Another solution is a modification of the BDSM "collar and leash".  In this modification, there is no visible collar or leash, but the same technique is used to get one person to follow another closely.

Edited October 28, 2020 by Lindal Kidd

[Da5id Weatherwax]

The most seamless "walking together holding hands or (whatever)" sets are actually VEHICLES with two "seats" - REALLY old-school but still the most reliable. When you are on them they are invisible but they act like an old-school AO detecting your state and playing the right anims as the two of you move. The downside is that to use them you have be able to rez the set on the ground first before "stepping on"

 

[norajulian]

K&S has an “Overly Attached” HUD (1&2) that gives you various options on how to cling to each other while in an area that allows scripts. It won’t meet your hands together, but you can climb up your partner’s back like a monkey and walk around that way, so that’s cute.

[Alwin Alcott]

for me the most used tool is Meike's hud .. old...free and adjustable
(btw watch out on MP.. some ****** sell that hud)

[Yuumo Ichibara]

1 hour ago, norajulian said:

K&S has an “Overly Attached” HUD (1&2) that gives you various options on how to cling to each other while in an area that allows scripts. It won’t meet your hands together, but you can climb up your partner’s back like a monkey and walk around that way, so that’s cute.

We have that too and it is actually super cute! Originally bought it for my toddler AV so we can “carry” him while walking.

You could also try this one it has 19 animations + 3 hugs + shoulder ride. This one worked for us best despite some negligible issues that could easily be solved with minor adjustments. No holding hands while walking though, but at least you do not need to rez it. You can also teleport together automatically. The only thing we did not like about it was that it can be script-heavy and causes lag sometimes.

Edited October 29, 2020 by Yuumo Ichibara

[Mollymews]

a kinda cute substitute for exact hand holding is wrist particle ribbon binding

the first person wears a bracelet/bangle/cuff on their right wrist. The other person wears it on their left wrist.  There is a script particle ribbon that runs from cuff to cuff

[Yuumo Ichibara]

5 minutes ago, Mollymews said:

a kinda cute substitute for exact hand holding is wrist particle ribbon binding

the first person wears a bracelet/bangle/cuff on their right wrist. The other person wears it on their left wrist.  There is a script particle ribbon that runs from cuff to cuff

Correct!

Here is one example. I do not know about other couple bracelet/bangle/cuff/rings but you need to have RLV enabled in your viewer for this one. You can modify the script particle chain to make it invisible so it just looks like you have matching rings on and always walking closely together. One time hilarity ensued when we forgot to turn it off and my man fell off a deck, he was still bound to me with the chain so it kept drawing him back to me but since he was in midair it looked like he was bungee jumping and bouncing back lol

[Phil Deakins]

I didn't read every post in detail but it does seem that this idea didn't get mentioned.

Rather than trying to find couple animations that fit your avatars, you could find couple animations that you like, and then alter your avatars to fit them.

Couple anims are normally created to fit reasonably average sized couples so be an averagely sized couple. By that, I don't mean n feet tall. I mean a couple where the male is a little taller than the female. Both short or both tall shouldn't make too much difference. And try to get body proportions average too - length of arms, torso, legs, and such.

So find hugs and kisses that you like, and change yourselves to match them. The chances are you'll find that you're reasonably matched to other anims too.

Edited October 29, 2020 by Phil Deakins

[Lucia Nightfire]

If LL would only update the animation protocol to allow either insertable IK's or procedural animation generation, we wouldn't be 17 years behind MMO standards.

[Drakonadrgora Darkfold]

21 hours ago, Lindal Kidd said:

Another solution is a modification of the BDSM "collar and leash".  In this modification, there is no visible collar or leash, but the same technique is used to get one person to follow another closely.

Many collars dont require you to modify them for that purpose, they have a follow feature. It acts like being leashed but without the leash particles being shown. You can adjust the follow distance on them to be very close to simulate walking close to each other. but the other will always follow just a little bit. If set too short, the follower will bounce off the other person when they start and stop moving though.

thing about using a collar, its you can get them for free from opencollar instead of having to buy one from some other place.

Edited October 29, 2020 by Drakonadrgora Darkfold

[Jennifer Boyle]

On 10/28/2020 at 8:48 AM, Maitimo said:

 Even then, they require both avatars to be exactly the same size as the avatars used when the animation was made, and no-one knows what those sizes are. 

 

Why do we have this limitation? Why can't scripts determine the 3D coordinates of the avatars' bones and adjust animations to fit avatars as needed?

[Gatogateau]

Here's another alternative: Embrace (as it were) the weird and hilarious borkeness of trying to hug in SL.

Trying to hug, kiss, pounce hug, etc. other people and vice versa has resulted in some of the deepest, longest... LAUGHING (you pervs, laughing!) in my 11+ years in SL.

It ain't never gonna work well, at least as I think you're wanting it to, because of all of the reasons mentioned above. Enjoy the madness.

 

[Nalates Urriah]

17 hours ago, Jennifer Boyle said:

Why do we have this limitation? Why can't scripts determine the 3D coordinates of the avatars' bones and adjust animations to fit avatars as needed?

The reason is computer speed and power. We simply don't have enough on either the client or server side. Determining positions based on the vertices in the mesh body and clothes is technically possible. But the time needed for those calculations would not allow for a real-time world.

The SL physics engine uses numerous short cuts to figure out when an avatar has walked into a walk so it can refuse to let it pass. It sees the avatar as simplified pill/egg shape and things in-world as boxes. Still with a bunch of avatars moving around in a region the server side slows down. You can really see this slowdown with any Pathfinding toys when there are lots of avatars moving around.

The physics engine and most everything else in SL avoids calculations based on the vertices that make up the avatar. Still put 20 avatars together and frame rates in the viewer and server drop and that is with reducing the number of points representing the avatar to less that a half-dozen. Doing calculations for a cloud of vertices numbering between 50,000 and 100,000+ would drop us into a frame rate of minutes per frame rather than frames per second.

It is simply too much to calc in the time available.

[Arielle Popstar]

6 minutes ago, Nalates Urriah said:

The reason is computer speed and power. We simply don't have enough on either the client or server side. Determining positions based on the vertices in the mesh body and clothes is technically possible. But the time needed for those calculations would not allow for a real-time world.

Imvu has had a slew of default couple animations available that work great with whatever size the individual avatars are, since they started. Surely if Imvu can do it, S/L can have it too. I believe that during the actual hug, the avatars are reverted to a default shape that then works perfectly with the animations with no pose balls required, simply r-click on the other avatar and select Hug or whatever from the context menu.

[Lindal Kidd]

1 hour ago, Arielle Popstar said:

during the actual hug, the avatars are reverted to a default shape

If true, this would make hugging a Tiny a very surreal experience.

[Arielle Popstar]

On 10/29/2020 at 8:13 AM, Lucia Nightfire said:

If LL would only update the animation protocol to allow either insertable IK's or procedural animation generation, we wouldn't be 17 years behind MMO standards.

This is possible to do? 

[Jennifer Boyle]

 

6 hours ago, Nalates Urriah said:

The reason is computer speed and power. We simply don't have enough on either the client or server side. Determining positions based on the vertices in the mesh body and clothes is technically possible. But the time needed for those calculations would not allow for a real-time world.

The SL physics engine uses numerous short cuts to figure out when an avatar has walked into a walk so it can refuse to let it pass. It sees the avatar as simplified pill/egg shape and things in-world as boxes. Still with a bunch of avatars moving around in a region the server side slows down. You can really see this slowdown with any Pathfinding toys when there are lots of avatars moving around.

The physics engine and most everything else in SL avoids calculations based on the vertices that make up the avatar. Still put 20 avatars together and frame rates in the viewer and server drop and that is with reducing the number of points representing the avatar to less that a half-dozen. Doing calculations for a cloud of vertices numbering between 50,000 and 100,000+ would drop us into a frame rate of minutes per frame rather than frames per second.

It is simply too much to calc in the time available.

Perhaps more would be possible if the viewer code utilized the processing power available on most modern PCs by efficiently using multiple cores. This will have to happen at some point if SL is to remain viable. Why not now?

If it's not possible to have on-the-fly computation of avatar size and position, shouldn't it at least be possible to have manual adjustments for avatar height and girth, similar to the commonly available adjustments for position and rotation? That would improve the experience with two or multi-person animations a lot without needing much additional processing power. Or, even better, couldn't a script determine avatar size itself? Doing it once at the beginning of an animation would require only a little additional processing.

 

Edited October 30, 2020 by Jennifer Boyle

[Anuuubis]

There are absolutely ways to make character models "ease" into animations and have limb bones stretch to about where they SHOULD be, it's just a matter of if SL's old engine/hardware can accomplish it. Modern games and even games almost a decade old now like GTAV have IK systems that allow for things like this; watch any video on The Last of Us 1 or 2 or Red Dead Redemption 2 and watch how the character arms stretch out depending on how you're positioned to grab items on shelves, or how models (especially in RDR2) have an animation to adjust themselves and line themselves up with whatever they're about to interact with before moving to interact.

Even SL's current engine has some really primitive IK. Turn your AO off and adjust your hoverheight, or stand weirdly on an incline and watch as the legs bend and contort trying to adhere to the ground without sinking or floating off of it. Surely it wouldn't be impossible to either let scripts take control of the IK system and "adhere" them to surfaces/other avatars when code calls for it and/or update the skeleton similar to the bento skeleton that allows for IK use on hands as well as legs.

Honestly there are a lot of things SL is really lacking but it's unknown if it can even be implemented. Jigglebones, for one. We have physics attachments and sliders, but it'd be nice to be able to rig mesh to predetermined "jigglebones" that utilize something similar to the physics we already have. Imagine bento tails having actual weight and bounce to them instead of relying 100% on animation HUDs? Imagine dresses being able to utilize jigglebones to avoid the awkward clay look mesh dresses have when the avatar walks?

If only.

[Nalates Urriah]

1 hour ago, Jennifer Boyle said:

 

Perhaps more would be possible if the viewer code utilized the processing power available on most modern PCs by efficiently using multiple cores. This will have to happen at some point if SL is to remain viable. Why not now?

If it's not possible to have on-the-fly computation of avatar size and position, shouldn't it at least be possible to have manual adjustments for avatar height and girth, similar to the commonly available adjustments for position and rotation? That would improve the experience with two or multi-person animations a lot without needing much additional processing power. Or, even better, couldn't a script determine avatar size itself? Doing it once at the beginning of an animation would require only a little additional processing.

For the last several years the Lindens have been changing the original single-thread computation of SL to parallel-multiple threads. It is no simple task. Any parallel computing process is complex. The viewer now uses about 22+ threads. There remains the single huge thread for rendering which all other threads support. The Lindens don't talk about what happens on the server side. But we know there are multiple computers running the backend services and I assume all new code is as multi-threaded as possible. We know the region servers, chat servers, and BOM servers run at near capacity as from time to time they lag when things get busy or there are problems. My point is we are close to the limits now and people are complaining about performance.

We have manual adjustments now. Shape siders do affect the shape of the avatar's bounding capsule in all directions.

When the Lindens made their last pass on avatar positioning about 2013 a Linden team spent about a year figuring out how to improve what we had without slowing down SL. We got both the Shape - Hover Height (in Appearance) and Avatar - Hover Height (the right-click on avatar thing).

The avatar height information is available to scripts now. But the problem is not that simple. Head size, neck length, torso length, leg length, and foot size are all variable. The individual parameters are NOT available to the scripts. That is because we already have a data transmission bottle neck. Each avatar in a region increases the data load exponentially. Thus the Lindens are extremely frugal in what data they require be sent. I doubt we will ever (as in the next 10 years) have the info we need to figure out where an avatar hand is.

So, while it seems a simple enough problem to solve, as one steps through and details with all the limitations it becomes much more complicated.

However, the viewer is open source and anyone can write code for the viewer or be creative with scripts. So, you are free to resolve the problem by writing some code. We have had this problem in SL without an upgrade to it since 2013. No one has made an improvement which solves it. We have some bright programmers in SL and while the Lindens get bad mouthed, it is usually by the ignorant, they are pros. But, go for it.

[Jennifer Boyle]

8 minutes ago, Nalates Urriah said:

However, the viewer is open source and anyone can write code for the viewer or be creative with scripts. So, you are free to resolve the problem by writing some code. We have had this problem in SL without an upgrade to it since 2013. No one has made an improvement which solves it. We have some bright programmers in SL and while the Lindens get bad mouthed, it is usually by the ignorant, they are pros. But, go for it.

 Let me ask you a related question that I have long wondered about. We have been told many times that the reason that SL's performance is not better is that all of the content is dynamic and the system must constantly update the data it uses for rendering. Most of the content that I am routinely around is almost static. Most of the non-avatar content in the sim where I  live is structures and furniture, much of which hasn't changed for months or years. I have wondered why locked objects cannot be flagged and  checked much less frequently than others, saving a lot of resources. I visualize the system intentionally taking several (10-30?) seconds to unlock an object and the system examining it at the same interval instead of multiple times per second. Locking objects could be incentivized by giving an LI reduction for locked objects, if just having better performance wasn't enough incentive. My BF is a professional programmer, and he thinks such a system is feasible and should help.

Sometimes, when I logon in my own house, where I spend at least half my time, textures on objects that have not changed for years take as long as a minute to load. Other times it's really fast. I have a very fast connection, powerful CPU and graphics card, lots of RAM, and OS, programs, cache, and data all on SSDs, and cache is at maximum permitted size. Why is there such variation? What is going on to make it so slow sometimes? It's been considerably better the last several weeks. I've wondered if the migration to the cloud had something to do with that.

[Candide LeMay]

17 hours ago, Arielle Popstar said:

This is possible to do? 

It's not only possible but was already in the works over a decade ago http://wiki.secondlife.com/wiki/Puppeteering

We don't have it because LL management doesn't see the value of avatars actually interacting with each other or the environment.