how to Pose 4 arms then mirror to other-side?

If you mean the octopus that comes with poser.

Make sure the 4 tentacles you pose are all on the same side (all right or all left)

Then click on figure menu, pose symmetry, left to right or right to left depending on which side you posed

if only it were that easy. That is what it should do but is not.

True bi-lateral symmetry requires sign changes on X translations and Y & Z rotations (in the simplest form that relates to human figures, at least). Poser knows how to do that, but cannot do so for morphs unless they contain a paired L/R suffix, so it can't tell that SideSideAll is a master channel for z-rotations. If you were to bake those morphforms (a common name for master channels) out to the individual limb actor rotations, then Poser's symmetry functions could work with them. Unfortunately, there is no (simple) way to uncook an egg or replace a set of common limb rotations back into a morphform that controls them.

Looking through the free stuff and marketplace came up with this: NetherWorks' PoseWriter Panel as the only readily available option for morph symmetry. I have written such scripts for my highly customised V4 figure, without the UI, though, and there is a great deal of customisation required to identify trunk and limb control morphforms (master parameters) which are subject to symmetry transfers. The Giant Octopus is certainly a prime candidate for such treatment, but my script would take significant work to bring up to the same standard and general figure applicability as NetherWorks' offering, so I'll leave it at that.

It looks like you're morphing it, not posing it.

No, @EClark1894, the SideSideAll channel on the root of each tentacle is an EZPose type control that drives the zrot channel on each actor of that tentacle. Ultimately, no morphs are involved, but the channels are likely to be targetGeoms, rather than valueParms. The problem is simply that it's no easy matter to identify that they are subject to symmetry without detailed analysis of the valueOperations. That said, it could be done, and Poser should do it natively. I have made this request for subsequent releases, as it's not impossible to do, just a bit involved.

Are you sure? He looks like he's got a value on the second morph dial in that shot and it's highlighted indicating it's been selected. The SideSide dial is reading 0.

It's not a morph. It's a master parameter. I've loaded the figure and there are no deltas. Its valueOperation dependencies form a chain down all the child actors affecting their zrot parameters, which causes the curl of the tentacle. This is not a morph effect. It's purely rotations controlled by a single dial.

click on the morph dial arrow to the right, and select Match Centers to Morph.

Once done, zero your figure and re-save and it will be saved with that parameter.

an0malaus posted at 2:27PM Thu, 06 February 2020 - #4379256

It's not a morph. It's a master parameter. I've loaded the figure and there are no deltas. Its valueOperation dependencies form a chain down all the child actors affecting their zrot parameters, which causes the curl of the tentacle. This is not a morph effect. It's purely rotations controlled by a single dial.

Does figure > rig symmetry > left to right fix it?

Why would that fix it? The rig is almost certainly already symmetrical. You don't reverse the sense of +ve rotations on opposite sides of the body for symmetry, you change the sign of the value. It's not a rigging problem. It's a master channel on each tentacle that applies in a consistent rotational sense, +ve gives a positive z rotation on all child actors, which is what is seen in the posted image. There is nothing to tell Poser that a master channel on an actor on one side needs a sign change to give a symmetric appearance when applied to a master channel on the opposite side, especially if the channel on all actors is called SideSideAll. How is Poser supposed to know just from the name that, like zrot and yrot channels, left/right symmetry requires a sign change on the opposite side of the body. Unless Poser analyses master channels to see whether they ultimately control actor rotation channels with signed symmetry, it cannot apply symmetry to such channels.

Why would that fix it?

I figured it might be worth a shot, not knowing how EZPose applied the rotations to all the tentacles.

A simpler fix might be to create additional master channels on the BODY or head actors that drive symmetrical settings of the master channels on each pair of tentacles, that way one dial will give a symmetrical pose of a pair of tentacles. If Poser's symmetry functions won't do the job, use valueOperations instead.

Say in valueOp syntax:

actor rTentA01:1
{
channels
    {
    targetGeom SideSideAll
        {
        valueOpDeltaAdd
            Figure 1
            head:1
            TentASideSideAll
        strength -1.000000
        deltaAddDelta -1.000000
        }
    }
}

actor lTentA01:1
{
channels
    {
    targetGeom SideSideAll
        {
        valueOpDeltaAdd
            Figure 1
            head:1
            TentASideSideAll
        strength 1.000000
        deltaAddDelta 1.000000
        }
    }
}

I did actually try what you suggested, but it didn't make any difference. Pose symmetry does work for pure rotations on the tentacle actors, but unfortunately, it copies the morph values directly, so lTentA01:SideSideAll = 2.0 gets copied to rTentA01:SideSideAll also = 2.0, but those master channels are additive valueOps on the tentacle zrotations, so both tentacles curl to the left (+ve zrot), whereas, since they need to be symmetrical, there should be a sign change as there is with just the pure zrotations.

With left to right symmetry applied on only rotations, the correct result ensues:

With a master control set on the left, symmetry just copies the save value to the right side, giving non-symmetric results:

To get the symmetric result, a sign change is required. Poser could do this, with a bit of intelligent exploration, but doesn't ATM: