Deformations on top of an alembic cache

24 views
Skip to first unread message

Steve Davy

unread,
Mar 20, 2025, 12:41:12 PMMar 20
to Maya Group
I'm trying to make some fine tweaks to a dynamic simulation that's been cached out to alembic.

Basically I'm attempting to nudge over a few errant verts that are penetrating another surface where they shouldn't, simply by keyframe a cluster that contains them from 000 to (for the sake of explanation) 111 (the fix position) and back to 000 again over just a few frames (a long main render is already done — these are meant to just be fix frames).

When I do this, it seems to work fine, but when I then move the timeline, the position of the cluster pops back to 000 even at the fix frame. Bizarrely, the animation curves in the graph ed still show non-zero values, but the channel box shows 000 and if I render that's also what I get.

I've also tried actually putting keys directly onto verts (messy, not my preferred solution), and got the same behavior.

I'm absolutely sure I've animated components on an alembic cache in the past and never had these weird issues. Can anyone confirm this should be possible, or otherwise?

Steve Davy

unread,
Mar 20, 2025, 6:26:56 PMMar 20
to maya...@googlegroups.com
OK, turned out there was an active Timewarp curve in the scene, so the frames I was setting keys at were then getting offset. Makes sense but not very helpful practically, especially as I now can't identify the problem frames with the Timewarp disabled, and can't set corrective deformations with it on.

Does anyone have a workflow for this?

Seems like there should be a way to identify which absolute frame you're viewing with the Timewarp enabled, or some other way of correlating warped and unwarped keyframes.

This is why I never use Timewarp any more.

From: maya...@googlegroups.com <maya...@googlegroups.com> on behalf of Steve Davy <stevi...@hotmail.com>
Sent: Thursday, March 20, 2025 9:41 AM
To: Maya Group <maya...@googlegroups.com>
Subject: [maya_he3d] Deformations on top of an alembic cache
 
--
You received this message because you are subscribed to the Google Groups "maya_he3d" group.
To unsubscribe from this group and stop receiving emails from it, send an email to maya_he3d+...@googlegroups.com.
To view this discussion visit https://groups.google.com/d/msgid/maya_he3d/DS0PR11MB7768E9B648B51B1E7D217C91DDD82%40DS0PR11MB7768.namprd11.prod.outlook.com.

joiec...@gmail.com

unread,
Mar 21, 2025, 6:41:54 AMMar 21
to maya_he3d
TimeWarp is pretty confusing most of the time, I don't use it if cannot.
You can't find the offending frames with timewarp disabled?

Jim Collins

unread,
Mar 21, 2025, 9:51:26 AMMar 21
to maya...@googlegroups.com
Clusters or containing Vertexsq, which have numbers. And each time you bring in an alembic file, the numbers of those vertexes change instead of using a cluster, try a soft modification which could work on a volume space rather than a select number of vertexes

joie

unread,
Mar 21, 2025, 5:57:54 PMMar 21
to maya...@googlegroups.com

Soft modification deformer es great. Another alternative is a lattice deformer, but without selecting vertices but the whole object. Then you can restrict the influencing area by moving and scaling the ffd objects (one is hidden by default but you can select it in the outliner)


You received this message because you are subscribed to a topic in the Google Groups "maya_he3d" group.
To unsubscribe from this topic, visit https://groups.google.com/d/topic/maya_he3d/TfbJV_goEHw/unsubscribe.
To unsubscribe from this group and all its topics, send an email to maya_he3d+...@googlegroups.com.
To view this discussion visit https://groups.google.com/d/msgid/maya_he3d/CAGXnhFUZVo-Moi_4C9toxG-SykZu9LoBHRRtnrFL4mwK%3DTXKWQ%40mail.gmail.com.

ryan harrington

unread,
Mar 23, 2025, 6:09:03 PMMar 23
to maya_he3d
Hey Steve.  
If you select the timewarp curve and set a key on it, you can select the key and see the re-time look-up value in the graph editor value feilds (or in the channel box if it shows anything.   
However, setting a key on a curve will affect the tangency in and out of that key, so it might affect the retime on other frames. 
So a non-destructive way to find you problem frames would be to key the timewarp curve on a problem frame...look up the indexed value, make a note, and then undo setting the key to leave the anim unchanged.

You could bake the timewarp curve if you are sure it's the final fix, that would avoid issues with changing the tangency and dancing around undoing things.

FAFF. 



Steve Davy

unread,
Mar 25, 2025, 12:09:39 PMMar 25
to maya...@googlegroups.com
Thanks for this... typical of the kind of convoluted workflow you have to adopt in Maya so frequently! I'll make a note in case this ever comes up again, but I'll never use a Timewarp again in Maya by choice (this was an old file). Even baking them, as I recall, is a major headache and involves using animation layers.

But, I managed to get through this project without as many fixes as I was expecting.

From: maya...@googlegroups.com <maya...@googlegroups.com> on behalf of ryan harrington <ryanowenh...@gmail.com>
Sent: Sunday, March 23, 2025 3:09 PM
To: maya_he3d <maya...@googlegroups.com>
Subject: [maya_he3d] Re: Deformations on top of an alembic cache
 
--
You received this message because you are subscribed to the Google Groups "maya_he3d" group.
To unsubscribe from this group and stop receiving emails from it, send an email to maya_he3d+...@googlegroups.com.

ryan harrington

unread,
Mar 25, 2025, 1:45:03 PMMar 25
to maya_he3d
Yup, it's a load of nonsense, and baking timewarps is absolutely horrible. 
It's worth mentioning that you can connect fcurves to 'unwarped time' allowing animation to exist outside of the warp...I'm doing some bullet time shots at the moment and the camera fcurves bypass the timewarp. 
Also Bonus tools allows you to timewarp individual objects, but still relies on the animation layer insanity.

stephen mann

unread,
Mar 27, 2025, 8:33:29 PMMar 27
to maya...@googlegroups.com, maya_he3d
Sort of already been said. But yeah.  The time warp in vanilla Maya affects the time1 node.  

While most scripts (including mine) are post time out. So you can affect individual objects or just specific animation nodes. 

 Since it’s alembic. Id bake that out with the time warp applied. 

 Then in a new scene. Do the tweaks on a fresh alembic import.  Of that baked alembic 


But yes.  Soft mods. And lattices are nice in that the don’t arent vert number specific 

Check out the tools in anim polish to get ideas. 
 Used it a bunch at Psyop.   

 General approach was bake out to cache. 
 So tweaks.   Re cache and send that new cache to render 

Hth 


-=s

On Mar 25, 2025, at 1:45 PM, ryan harrington <ryanowenh...@gmail.com> wrote:


Yup, it's a load of nonsense, and baking timewarps is absolutely horrible. 
It's worth mentioning that you can connect fcurves to 'unwarped time' allowing animation to exist outside of the warp...I'm doing some bullet time shots at the moment and the camera fcurves bypass the timewarp. 
Also Bonus tools allows you to timewarp individual objects, but still relies on the animation layer insanity.

--
You received this message because you are subscribed to the Google Groups "maya_he3d" group.
To unsubscribe from this group and stop receiving emails from it, send an email to maya_he3d+...@googlegroups.com.

joiec...@gmail.com

unread,
Mar 28, 2025, 12:59:51 AMMar 28
to maya_he3d
I'm not in front of MAYA right now, but I remember you can use "Ctrl" or "Shift" keys while creating a keyframe in the curve editor, and it will NOT change the tangents in any way. Pretty cool tip there.

Steve Davy

unread,
Mar 28, 2025, 5:21:18 AMMar 28
to maya...@googlegroups.com
Yep, I'll never use it voluntarily again if there's any alternative, and there usually is.

With Ryan's workflow tip I could probably have found most of them, but in the end only really needed to make one fix which I managed to find by scrubbing.

From: maya...@googlegroups.com <maya...@googlegroups.com> on behalf of joiec...@gmail.com <joiec...@gmail.com>
Sent: Friday, March 21, 2025 3:41 AM
To: maya_he3d <maya...@googlegroups.com>
Subject: [maya_he3d] Re: Deformations on top of an alembic cache
 
Reply all
Reply to author
Forward
0 new messages