Project Bento User Group update 26 with audio + hints’n’tips

Bento: extending the avatar skeleton
Bento: extending the avatar skeleton

The following notes and audio were taken from the weekly Bento User Group meeting, held on Thursday, September 15th at 13:00 SLT at the the Hippotropolis Campfire Circle. and chaired by Vir Linden. For details on the meeting agenda, please refer to the Bento User Group wiki page.

Note that this update is not intended to offer a full transcript of the meeting, nor does it present the discussion points in chronological order. Rather, it represents the core points of discussion, grouped together by subject matter were relevant / possible, whilst maintaining the overall context of the meeting.

RC Viewer Release

As indicated in an official blog post, and in this blog, the Bento viewer is now at release candidate status with the release of version – which can be obtained through the Alternate Viewers wiki page. This means that TPVs can now officially adopt the Bento code and release it in their own viewers. Cool VL Viewer already has Bento merged into its experimental branch, for example, and Firestorm has the code merged ready for their next scheduled release.

There are a number of small updates in the RC version of the viewer, chiefly typos in the sliders which were preventing them from having symmetrical effects. These affected the lip thickness slider, the square head slider, and the body thickness slider, together with a slight tail bone issue.

The RC Viewer and “Distorted Avatars”

If you opt to experiment with the Bento RC viewer and Beno demons etc., keep in mind that any Bento-enabled items to you wear will only appear as intended when viewed from another Bento capable viewer. Anyone using a non-Bento viewer see you is at best going to see things incorrectly positioned on your avatar, and at worse, see whatever you are wearing utterly deformed.

Bento creations will only appear correctly when viewed in a Bento-enabled viewer, as shown in these two pairs of images. On the left: a Bento head seen in a Bento viewer, then in a non-Bento viewer; On the right, a Bento avatar seen respectively in a Bento and a non-Bento viewer (click for full size)
Bento creations will only appear correctly when viewed in a Bento-enabled viewer, as shown in these two pairs of images. On the left: a Bento head seen in a Bento viewer (note; the jewellery is out-of-position on the first image as I didn’t bother repositioning it to fit the head for the photo) and then on a non-Bento viewer – note the head is now located in the small of my back (arrowed). On the right, a Bento avatar seen respectively in a Bento and a non-Bento viewer (click for full size)

Obviously, as the Bento code is more widely adopted and reaches release status, these issues will decline – but for now, if you’re using the Bento RC viewer (or other viewer with Bento support), do keep this in mind when venturing out in public.

In extreme cases, older viewer versions may crash if a Bento avatar / Bento content is encountered; but these cases should be rare. The Lab added code to the viewer some time ago to specifically prevent Bento updates from crashing the viewer, so as long as a viewer has these updates – as should be the case with all currently maintained viewers – it should not crash.

Reset Skeleton

The Reset Sketon option is available from the right-click context menus for both avatar name tags (l) and avatars (r)

There are occasions (when changing between avatars, for example) when those who are on a Bento viewer may see themselves or another Bento avatar distorted. This is due to variances in how appearance updates are received / handled (and even with non-Bento avatar models has in the past required a re-log to fix).

To overcome this in the Bento viewer, there is a Reset Skeleton option, which can be used on either yourself or other deformed avatars in your view. This can be found in the avatar name tag context menu, or the avatar context menu, either of which can be access by right-clicking on the name tag or avatar respectively.

Vir notes the solution is not ideal, but the only other way to fix such issues would be extensive re-working of the viewer code – and at least this approach avoids the need for a re-log to correct matters.

Time Frames for Release

Questions were asked on when the viewer might go to release. this is actually dependent on a number of factors, including:

  • The viewer’s crash rate compared with the current release viewer and other RC viewer in the release channel
  • Whether exposure to a wider audience of users uncovers non-bento bugs or regressions which require additional fixing
  • Relative priorities between Bento and other projects.

Avatar Vertical Position (height above ground) Calculation

This has been a running topic for some time. In short, a n avatar’s vertical position relative to the ground is determined by a complex calculation which involves using a number of joints running up through the body from the left foot to the skull. The idea is to present a consistent view of an avatar standing on the ground, rather than in the ground or over the ground. However, if any of the joints used in the calculation are changing position unintentionally when the calculation is made, it can result in the avatar seeming to bounce up-and-down (see here for more).

To help overcome these kinds of issues, Vir has now documented the body height calculation bones, and the details can be found here: Avatar body size

Joint Position and Volume Bones

Visualising volume bones

It’s been noted that zooming in on a Bento avatar or attempting to click on a part of the avatar can be difficult. This is generally because the collision bones are not set-up correctly. The positions of the volume bones can be seen using Advanced Menu > Avatar > Show Collision Skeleton.

This will overlay the avatar with a series of oval shapes (official viewer – other viewers may render the collision skeleton slightly differently) which show the position of the volume bones. The closer these match the shape an avatar (something which may not always be possible, depending on the avatar type / size), the easier it will be to zoom in on the avatar and / or click on it.

Tapple Gao’s avatar testing and visualisation kit of meshes and animations can also be used to show the collision skeleton.

Setting volume bone positions can either be done within the mesh model or via animation, although there was some confusion whether collision volumes now get included in joint offsets.

Gestures and Outfits

This is not always realised, but gestures – such as for speech – can actually be stored as a part of an avatar’s outfit in the Outfits folder, allowing them to be automatically activated / de-activated. This isn’t done as a part of the Save Outfit function, but can be achieved by manually copying a link to the gesture into the required folder. Replacing outfits should then trigger gesture activation  of those being added, and de-activate those gestures associated with the outfit being removed.

MayaStar and Avastar Updates


Cathy Foil has been working hard to get slider support fully implemented in MayaStar, her rigging system for Maya users. She believes she’s 90% of the way there, and the remaining slider support will be completed in the next couple of weeks.


  • There will be one more Alpha release of Avastar 2, to fix some minor skeleton problems after which things will move to beta
  • The beta will have more documentation and an improved UI, but beyond possible re-targeting enhancements (see below), the capabilities will not different extensively from the alpha version
  • Avastar now has much improved support for re-targeting bones (e.g. for use in motion capture, or for converting models a creator has animated for use in one system for use in Second Life, etc)
  • Import of animations in much improved, including support for batch import of animation files
  • The hope is for Avastar to be at a release status within the next six weeks, with a beta version out hopefully in the next 14 days.

In addition to the above, Matrice Laville would like to hear from anyone involved in motion capture and animation on tools used, and how Avastar might better support them (right now Avastar only recognises SL .bvh files and Carnegie Mellon files.

Additional Links

One thought on “Project Bento User Group update 26 with audio + hints’n’tips

Comments are closed.