I’m going to do some experimenting as well. It’s possible that the Boom just does what MostHost is proposing.
In my experience, the boom is only useful to allow the character to bang his/her head inside stuff and prevent clipping.
it’s essentially what it’s designed to do…
But you can also use the boom to smooth out motions with the camera lag options. I haven’t played with this anymore since I’ve been focusing on level design for my project, but if you can somehow get the camera lag to smooth out the Z movement, then you’d not have to go through the process of creating custom code or animations to smooth that out.
Lag makes you more sick, so don’t waste time.
1 Like
The camera is attached to the head socket. This can be reduced from Anim graph of character, you have to add transform modify bone just before output pose.
I am facing same problem while making my game. You can watch this video for full tutorial: “https://youtu.be/BNxeXHrBNz0?si=CtJxPGUg_j2c1fho”
It works like a charm
If you use a separate head mesh like metahumans have, you can just check “hidden in game” for the head mesh and also check “hidden shadow.” You wont see the camera clipping through the head, since it is invisible to the camera, but it will still cast a shadow.
With a camera on a spring arm unattached to any socket, you can then lerp between the world location of your head bone and the world location of your spring arm, and the alpha for the lerp will be a slider for how much headbob want.
1 Like
That’s an interesting solution that I’ll have to look at in the future.
The method I settled on was to not attach the camera to the skeletal mesh at all. If you pay attention to games like Mirror’s Edge you can see they’ve actually just added a body to a typical First Person setup, rather than adding a First Person Camera to a third person setup. That way the arms move with the camera and any hand animations are always going to be in the correct place, and then the body is only visble when you look down. In my experience since this original post, it’s a much easier system to work with since the body only needs movement animations.
I wrote an updated tutorial with my results a few years back:
You may want to look up some of the Crysis shots of what characters look like during different actions.
Wither way, true first person with an attached camera is only ever achievable on custom animations specifically adjusted to account for motion.
Anything you try is either a hack or not going to work.
So yes. Detached from mesh is the right the way to go - unless you have a budget or the know how to deal with making your own animations.
1 Like
This works for me, just put a delay on the uncrouch to sync with the crouch animation getting up. No clipping through the character as the camera moves up and down when crouching and yes I don’t like the head bobbing so it’s a case of moving the camera manually with the animation, my animation blend time is 0.3 so I match it with a delay.
May work “for you" won’t work for anyone else due to FPS delay.
If you want it to be somewhat functional, try adding an animation curve with the Z value on the animation (only needs the curve, can be added within unreal) and using the curve itself to detract from the camera Z - that way it’s at least somewhat in tune (one frame behind) but dependent on the playing animation.
I’m still a bit unclear on this subject. Currently using GASP with Metahuman in GASP’s Visual Override. I’ve tested attaching to head socket, definitely doesn’t work as head bobbing is excessive. Now tested the spring arm method attached to the head and although it lessens Z head bobbing, it’s not by far acceptable in my opinion..
Really gotta hand it over to Star Citizen team, that’s what I’m looking for.. anyone knows how to build this?
Easiest solution if you don’t want to try to code a camera smoothing system is to remove the Z movement from the animations. Of course if you’re using the Animation sample, that’s a lot of editing. The other solution is to not bind the camera to the head and to make your True First Person character in separate pieces. I created a tutorial on how to do this, and it’s still my preferred method:
If you don’t want to watch a video that’s more than an hour long, the TLDW is to start with the normal First Person template, add a mesh for the lower body, and a hidden full body shadow mesh that casts hidden shadows and use aim offsets for the shadow mesh. This gives you the best of all worlds and eliminates head bob entirely which you can always add back in as a procedural animation like weapon sway. Then you can multiply it against a value so that it can be scaled up or down in your menu.
2 Likes
Doesn’t look bad, the only major issue is that it’s upstream so it costs compute time. If you make proper animations you have no such cost.
But, you spent 100 hours authoring animations for perfection.
Also doesn’t look bad doesn’t mean it’s good. You’d have to feed it off to someone with bad motion sickness and a VR set to know if actually worked or not.
NOT that they did, but remember that you-tube video can be stabilized in editing as well.
Real first hand accounts are what you are really needing if you intend to make your game playable to those poor f*cks who suffer from motion sickness (I being amongst them).
Hey guys, thanks for the help, but I think this only works when you have the head stable and don’t need to move it (in this example it follows the bike so I don’t think it applies, I tried this method in my case and didn’t work.. I think the only way to do this is by sticking it to the head X and Y values but maintaining Z from the character’s Capsule component. I found that the most logical solution but can’t seem to be capable of implementing it yet..
Well, like I said the only other way is to remove the Z movement from the animations. The better solution though is my tutorial on how to combine 3rd person for the body and 1st person for the arms. I just discovered they changed the original FP template though, so it’s better not to use it. You need to just parent the player’s arms to the camera and then have a detached 3rd person body mesh and use Anim Offsets to make the shadow line up with the camera movement. The only way you can remove the bob on a first person camera is to not have any movement on it or to counteract that movement in code which would not only be annoying to figure out, but also likely introduce more issues. There really is no need to parent the camera to the head bone to get a True First Person effect.
There definitely is a reason to bind to the head bone. It wont be first person without it. It’s some other stuff that breaks when the character does actions like crouch.
The code solution is bad because of the lag introduced by calcs.
Removing Z is bad because you also loose rotation.
The only real way, like it was said a hundred times, is to properly author the animations, which takes ages.
“There definitely is a reason to bind to the hrad bone. It wont be first person without it.”
No disrespect meant as you’re one of my favorite people on these forums, but that’s entirely incorrect. A traditional first person camera setup parents the arms to the camera, and the camera is positioned at the location where the head would be. Then you don’t get any bob without adding it to the camera itself. It’s only relatively recently that people have started trying to solve it by attaching the camera to the head bone, and I can tell you, it’s a bad solve.
The best solution I’ve come up with over 9 years of experimentation is to attach your arms to the camera, and then use a separate mesh for the body and another for the shadow. It works beautifully.
Also, using a third person mesh and animations while simply allowing you to control the camera parented to the head, doesn’t give you proper control over aiming. The only way to get that is attaching the arms to the camera directly. It may not look quite as real, but it’s much better for the gameplay if you ask me.
Finally, when I said to remove the Z, I simply meant to adjust your animations to remove Z movement, not tranlation entirely. In any animation package you can separate the up and down movement from the rotation. I’ve also done this before, but as I said, it’s not worth the hassle.
I meant its not a True First Person camera (thats’s the name of literally the system where the whole mesh is always present).
Your solution works, but it isn’t the same thing.
You can’t see your feet, you can’t see your chest, you can’t perceive the motion of the character which in spite of nausea giving is often the feel of the game.
In TFP you do other things to aim, or remove Bob, or adjust the Z height. When done right its usually great. Problem is almost no one does it right.
Your solution while not true first person is better in terms of end output and cost to work ratio. Less work than producing high quality animations needed to TFP, better overall result. But it’s not the same thing obviously…
Instead if you make a character with a dedicated camera bone and you just animate that camera bone you sort of solve all the problems of TFP (you just need to remember to animate the bone!)
Ps: Also, shame on you, if you like my posts how dare you now read my mind about what I meant to write? 
PPS: I just remembered that I made myself 2 camera bones. One is for the pukingly impaired like me and has 0 motion/rotation (it’s used when the game sets headbob to 0). One has some motion but not all of what the head bone has (mainly, you edit it by making sure that through the animation the focal point of the camera view [10m in front] doesn’t move more than 1cm or so). With full headbob the camera is just attached to the headbone which is also autothered to keep that viewfield within 5 to 10cm.
I would strongly advise anyone going for TFP to do things similarly.