Here are links to the v14.0 Alpha 2 Installer.


Windows:

• v14.0 Installer (2007 - Oz CD ) (11m)
• v14.0 Network Installer (Hasp Requires 2006 subscription) (15mb) ]

OSX (Universal Binary):

• AMv14Installer.zip (23m)

Fixes since the last release:

• 0003703: [Bones/Rigging] feature or bug? Distortion boxes not directly possible in Characters (robcat2075)
• 0004055: [Interface] interface tweak: defineable color for keyed bones (robcat2075)
• 0003970: [Rendering] Gels on volumetric Lights render incorrectly (JohnArtbox)
• 0004081: [Constraints] Dynamic constraints don't work any more. (PianoMan)
• 0004065: [Net Render] render nodes getting Stuck inThrottle Delay (Jack West)
• 0003934: [Net Render] Netrender Throttle Delay is Triggering Falsely (rickh)
• 0004053: [Timeline/Channel] action shortcut ease and blend ratio don't interpolate with action cache window open (alweb)
• 0004072: [Hair] Exception Message Box in Hair/Particle System when exiting renderlock (Glen Clarkson)
• 0002829: [Animation] Blend method's blend ratio extends actions instance croprange. (sbk)
• 0004070: [Animation] Moving bone in action window with chor window open moves wrong bone sometimes. (bobcroucher)
• 0002823: [Animation] Extending action shortcut length doesn't update result until frame change. (sbk)
• 0004069: [Animation] Hand Gizmo doesn't refresh properly in action and chor. (bobcroucher)
• 0004042: [Interface] Absolute paths from pc to mac and back to pc cause open request to fail (Noel)
• 0003274: [Rendering] render to file shaded, crashes (Steve)
• 0003537: [Rendering] Frame Burn doesn't work on mac (text is garbled) (jandals)
• 0003426: [Hair] AM crashes when dynamics/collision dection turned on. (stuartcalvin)
• 0003878: [Rendering] Renders will crash or not update if Particles/Hair in Render options set to ON (CRToonMike)
• 0004016: [Constraints] Dymanic Constraint Options cannot be set (pwknox)
• 0004061: [Hair] Made densityfactor and the realtime settings visible on the instances of hair materials. (bobcroucher)
• 0004050: [Net Render] error loading string 5923 (Jack West)
• 0004058: [Rendering] Improved escaping during rendering. Shadows Hair (Noel)
• 0003931: [Rendering] Post Effect - Frame burn hangs AM on render to QT movie (wsuckow)
• 0004049: [Animation] invisible objects set to inactive don't interpolate the active Property. (Noel)
• 0004023: [Relationships/Poses] Paste Mirrored, not working with simcloth applied to the model or its groups (pixmite)
• 0004044: [Interface] Absolute paths from pc to mac and back to pc cause open request to fail (Noel)
• 0004045: [Interface] Invalid filenames sent to openrequest will be detected and openrequest will retry (Noel)
• 0004038: [Hair] New option on hair so it can be opaque in reflections. This can significantly decrease render times. (Noel)
• 0003988: [Bones/Rigging] Null visbility can no longer be toggled on or off on a null under a model cache. (mechadelphia1)
• 0004009: [Hair] Deleting hair emitter crashes AM. (PianoMan)
• 0004001: [Modeling] Finding patches when modeling in wireframe and particles are OFF. (bobcroucher)
• 0004003: [Interface] drag drop hang model over model in action. TWO: 2_8_13 (martin)
• 0004004: [Interface] Embedd all now has a Yes/No dialog explaining its use, with a don't ask again option (Noel)
• 0004000: [Modeling] Groups that point to deleted cps, due to corruption or svn problems, warn on load. (Noel)
• 0003686: [Plugins] FrameBurn PostEffect Mac, and PC's black bar (Noel)
• 0003999: [_Other] Save version and platform info in file (Noel)
• 0003998: [SVN Data Repository Interface] SVN conflicts that are accidently commited now warn better when loaded in A:M (Noel)
• 0003790: [SVN Data Repository Interface] conflicts with driver "Name =" being removed. (Noel)

Make sure when you report bugs to A:M Reports, and that you pick the v14 project in the upper right.

No more Yeti installer? Also, the link to the network installer doesn't seem to work.

Looks like with the new year has come the time to upgrade.
Drats... and Hooray*!


*'Cause I plan to upgrade tonight.

Done. Come on Mailman!

Yes I just ordered the upgrade tonight

man thats allot of bug fixes. Nice! Wasn't SSS supposed to make it into this release to? or did I miss it in the list.

looks like I'm logging on to try my upgrade now to.


edit... the check is in the binary code on your server ;-)

Yes, it's in there, Colin.

Click to view attachment

Thank you, Yves and all at Hash Inc!

I'm not able to see the surfaces of the patches, they're not showing up in the realtime render though they seem to render correctly.
I'll file a report.

I have that too paul. I already filed a report too. I have an ATI mobility x1400 256 mb with up to date drivers. If I tried switching to direct3d it crashed the program guess i gotta wait til next release to play with sss.

Yep. Ditto on that.

If you press "T", to turn, and then drag you will probably see the object flash into view briefly from time to time but it looks like the screen is being cleared after a redraw.



I think you should still be able to render some tests with sss. The quick render that I tried looked okay. You just won't be able to see the patch surfaces with the realtime renderer. That still leaves some room for play while we wait for the glitch to be fixed.

I get the same results with both nVidia and ATI cards.

Link to Net2007.exe is broken....

No luck rendering sss here. I just get a black surface.

Just ordered my copy.

Question; is old Yeti compatable with the new version?

I just get a black surface with sss too.

Edit: Well it rendered in final render to file. It looked like this with the default settings and not much different with any other setting. It's fine if I was painting with a palette knife but not what I thought. This is done with image maps, could that be the problem? I would love to know yves settings.

Click to view attachment

The last time I tested it was. (about a month ago)
I assume you are talking about file formats.

See this topic for my settings. I use the values 4.5, 2.5 and 1.6. It looks like I forgot to change the default settings. Also, there a known issue with short extinction distances (the current default values are quite short) and I am currently working on that.

Are you sure that you didn't set up something else to get those results, Yves? A buffer enabled somewhere or something else?
I have tried the chickenskin2 settings on several objects, with various arrangements of the default lights, and I just keep getting these very specular patchwork finishes that look like over-heated copper castings.

Click to view attachmentClick to view attachment

When I render to file with sss I get exception #010 and am told to back up my work.

I am using a 3.4 gig Intel.

I tried setting the number of threads to 1 (it is hyperthreading)

Paul,

I am totally guessing from this point (for now..)

Did you try a different light-type and or light-angle or maybe ambient-occlusion?

Niels

Exception #10 may be an error-indicator but it gives very good advise



From my own experience this far, something I think of is; only change the number of threads when no project is open, I hope you did just that...
Hyperthreading is a physical-unit on the processor but not a second processor, so threads set to 1 is the right setting.

A new project, a new simple model and the standard chor will give you probably more clearity, it being a static happening or not.

Niels

The v14 network installer link is not working. I am getting page not found.

Try this link:

http://www.s1.hashmirror.com/ftp/pub/updat...pha/Net2007.EXE

I think the original link had Net2007.exe intead of Net2007.EXE and the site is case sensitive.

Richard Harrowell.

Is anyone else seeing that when you open a model and go to shaded mode and then zoom in a little the model disappears?
If I do a preview it looks fine. I am also finding the same issues of disappearing model in the chor with shaded mode on.

I had a problem with a model in v12 where the model bone was very far from the center of the model (I'm not sure how it happened in the first place). Moving the model bone back to the center of the model did the trick. I don't know if that is the case for you, but just a suggestion for something to double-check.

Mousman,
Thanks for the suggestion. I went to go check the model bone location and what is happening for me is that when I go to bones mode the first odd thing I noticed is that all bones are un hidden and when I click in the model window to activate the model bone the program locks up.

That was the setting that I missed.
With threads set to 1 I get much better results:

Click to view attachment

The lighting here is just the default rig with the rim light intensity increased to 84% and it's colour changed to a pale yellow.

Click to view attachment

Oooh... yuo got it working. Not fair

It looks nice though. Particularly on the ear. Looks like it is missing from the lips though and possibly the patches around the left eyebrow.

Now what is it about my computer that prevents it from working?

Yeah that did it.

I really like this feature a lot. It is the best thing ever! I need to try it out with better lighting set up and stuff to get full effect and play with the settings. But I think this skin here looks pretty cool for a first shot.

I tried to apply sss to the shirt too but the shirt just turned black. i removed roughness settings and it was still black. So it is just a diffuse color. Don't know why.

So when i turned threading to 1 this means i am not using my dual processor for rendering now right? Even so this render was very quick. Maybe 30 or 40 seconds.

Click to view attachment

Yes, there was a group with specular values set for the eyebrow area and the lips groups was overriding the skin group.

Example with these things fixed.

Hey, Dennis, that really makes a big difference doesn't it! So cool!
Average render times for Woot were 43 seconds per frame for a single pass render at 640 x 480 pixels.

Outstanding render. Wow The improvement is amazing.

With a certain amount of wrangling I got there. Here is a summation of tips:

> Render with 1 thread only.

> Render to file first to cick it into action first time and whenever you change settings.

> Make sure that normals are facing out.

> If you are getting a black render, render to file and if that doesn't work, save and re-open.

> If you are getting squares on the object then whack the Reduced extension coefficient (REC?) up.

> Expect the render to take a little while to quit.

Cross fingers.

Here are some findings for sample density:

It seems that setting the sample level to higher than 100 can have a beneficial effect. Here you can see that it resolves the issue of a sharp demarcation in the gradient in the thin area of this test model.

Also, I would have expected the smallest nodule of the model to be brightest (following the patturn) but it is darkest.

One thing that might resolve this would be to have the sss sample an ambient environment colour. That way, small objects would emmit light that they are recieving from all sides. Maybe what is happening here is that the small nodule is only getting a small amount of light to spread around - too small for the sampling to pick up in these examples.

Dunno.

I know that this is a bit of an impertinent question, John, but are you certain that your light source is actually illuminating the very top of this object?

If you look at the stem/neck, that supports the dark finial at the top of your object, it is showing the correct effect, allowing more light to pass through thinner volumes. Then the bulb at the top suddenly goes dark, as if it is just outside of the cone of light. I guess I should try this myself.

I have rendered with sss and 5 passes without any problems so I don't think that the number of passes is an issue except for the added render times. I guess multipass will be neccessary to smooth out all the jaggies on the edges of objects.

Hi Paul. It did uccur to me that the top bit might not be lit but I used a bulb so there is no reason why the top shouldn't be illuminated. Since loading the project again, the result is better but still not completely correct as far as I can see (I might be wrong of course).

I have attached my project for you to look at.

you are right that it doesn't have to be one pass only. I think that was a superstition brought about by the fact that rendering was taking a while to quit and I thought AM had hung. Putting it at 1 pass means you are less likely to try to quit the render. I will ammend the list.

The neck of the very top bulb shouldn't be that dark; it's in direct light.

Click to view attachment

I think we are seeing a similar thing going on at the end of Jims' models nose here:

http://www.hash.com/forums/index.php?showt...5867&st=120