1  2  |   Order by index

Jun 2, 2002, 2:06pm

It is certainly not system dependant ... its something to do with how the
sequences were made ... that much is VERY clear. I'm still trying to
identify which sequences were made in which software that still work. So far
seqs made in lifeforms, accutrans, and TS4 no longer have vertical
translation. Some of the much older AW seqs still seem to work but no one
seems to be able to tell me what they were made in.

Opinion about how good something looks is mere opinion. It doesn't
necessarily indicate something is working as it was designed to work.

Glitter


[View Quote]

Jun 2, 2002, 2:12pm

An amusing thing I found today was someone who swore blind that after
updating to 3.3 everything looked exactly the same. I spent a great deal of
time trying to understand how he was getting everything as it was before the
upgrade. He kept saying it was my system (and everyone elses that was wrong
..... 9 in 10). After about 20 minutes he discovered the upgrade had not
worked and he was still in 3.2. This is beyond my comprehension but
nonetheless occured .... it at least gave me a smile for the day.

Glitter

[View Quote]

Jun 2, 2002, 2:25pm

Of the six world owners I've spoken to in the last hour, ALL of them have
said that ALL of their visitors have trouble with the majority of seqs
(except the really older AW ones) ... now this a pretty broad cross section
if you ask me ... and certainly enough to support my observations...
especially considering the complaints are all of the vertical component.

I would like to hear from anyone that has made a seq themselves that
demonstrates vertical motion of the hip. I'd like to know what world I can
view it in and what version and title of software they used. So far no one
has been able to prove this to me ... opinions don't help solve the problem.
Of course if you don't want a solution to this problem that 9 out of 10
people seem to have then just keep out the way while the rest of us try to
figure it out ... I'll be sure not to share the solution with you and waste
your time.

Glitter

[View Quote]

Jun 2, 2002, 2:46pm

its a feature in the feature list that says there is a new animation engine
and dequanes will be diffrent. obviusly somethings different about YOUR
computer becuase your one of the FEW who mysteriusly dont see 3.3 or havent
actualyl upgraded as sosmoen said below me

--
-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-
A message from Zeo Toxion
-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-

[View Quote]

Jun 2, 2002, 2:52pm

I won't bother with another reply after this. All the sequences work fine
for me. I chose the world "Fogland" randomly and was able to test all of
the sequences. No problems.

[View Quote]

Jun 2, 2002, 2:53pm

I'm running version 3.3.

[View Quote]

Jun 2, 2002, 3:25pm

Hey ya Mauz!
My .02 on all this........

In LifeForms there's a 'Pin to floor' option. The shappy.seq is pinned to
the floor. In the Lifeforms stage window it shows the Pelvis (center of
avatar) as being on the same level as the stage. If 'pin to floor' is off
then the avatar feet are on the stage. I'm trying to remember exactly the
problem I was having when first testing Accutrans for Wayne. If I recall the
pin to floor option had to be on for my seqs to work correctly. I was having
the same type of problem where the pelvis would stay in the same spot if
'pin to floor' was off.

Glitter, in short, put the pelvis at stage level and see if it works.
Casay




[View Quote]

Jun 2, 2002, 4:24pm

Glitterbug:

"An amusing thing I found today was someone who swore blind that after
updating to 3.3 everything looked exactly the same. I spent a great deal of
time trying to understand how he was getting everything as it was before the
upgrade. He kept saying it was my system (and everyone elses that was wrong
..... 9 in 10). After about 20 minutes he discovered the upgrade had not
worked and he was still in 3.2. This is beyond my comprehension but
nonetheless occured .... it at least gave me a smile for the day."

thats my reasoning...so anyway if they work fine for you then thats good.

--
-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-
A message from Zeo Toxion
-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-

[View Quote]

Jun 2, 2002, 4:37pm

> Opinion about how good something looks is mere opinion. It doesn't
> necessarily indicate something is working as it was designed to work.
>
> Glitter

Opinion about how bad something looks in mere opinion. It doesn't
necessaruly indicate something isn't working as it was designed to work.

Jun 2, 2002, 4:37pm

Hiya Casay,

Great idea but I tried that first off. The old 2.2 browser wouldn't work
without that pinned. I wondered if the step back (to a new engine) had
caused them to go back in that respect to. I have however just worked out
what is wrong and YES there is a problem with how AW interprets the seqs
now!!!! I'm somewhat reticent to off this information due to the attitude of
the community, but I found the solution when I compared the 2.2 era of qjump
(yes I still have it all burned on CD) to the more recent one offered for
d/l in AWGate ... they are different. How so? Two ways:

1. the vertical scale is 100 times out ... import as 0.1, export as 10x to
fix ... looks like someone in AWC messed up their math ... of course they
changed their seqs or wrote around it but didn't tell anyone!!

2. you now need to specify rotation, facing angle, altitude and location as
keyframes wherever a keyframe exists for one of the others. Eg; you can no
longer assume default is zero but you actually have to specify a key frame
value (albeit zero) for it to work. In3.2 you idnt have to specify defaults
if they were not used ... now you have to specify every piece of information
for each key frame ... making the sequences 4 times the size!!! ...
fantastic upgrade huh?

Oh well ... there's the BUG (not feature) fix for those who are interested.
if anyone wants me to explain that further then they can telegram "Glitter
Kitty" and I'll walk them through accutrans ... considering the complexity
of some of the older seqs however I'd say most of them will take a looooong
time to fix.

Casay ... maybe we could ask Wayne to put in something such that if a key
frame is specified for one dimension that keyframes are automatically set
for all the others (or turn them on as default or whatever).

Glitter


[View Quote]

Jun 2, 2002, 4:39pm

> Opinion about how bad something looks in mere opinion. It doesn't
> necessaruly indicate something isn't working as it was designed to work.
>

when something doesnt look as it was intended then that's bad ... at least
from an artists point of view

Jun 2, 2002, 4:42pm

Maybe you don't know how to make seqs, and now that you are trying, they
aren't working? :D

[View Quote]

Jun 3, 2002, 2:47am

well, before, the engine flipped a step from -180 to 179
through 0, so the rotation went backwards for a moment -
and, that's the strange thing, it did that only sometimes.

Now the engine seems to recognize that 179 is the direct
neighbour of -180 and continues rotating in the "right"
direction.


[View Quote]

Jun 3, 2002, 3:36am

When you have some time, go to MoonBeam or Labirint and try the
"Philiris" avatar. It has a "fly" sequence that went up into the
air about 3 meters in 3.2 / 2.2 - it now doesn't do anything.

The new engine is definitely messed up, the vertical movements
need to be fixed, or AW needs to supply a tool to create/modify
the sequences.

You cannot supply a viewer for something where you don't supply
a creator for the format, especially if it's proprietary. AW
cannot always rely on others to do this work for them. This is
a wrong conzept and will never allow AW to set any standards
in 3D and VR.

[View Quote]

Jun 3, 2002, 4:55am

this problem can be completely removed by having a key frame at 179 and
at -180 ... then there is no redraw between them and no sudden flip ... eg
no intermediate frames.


[View Quote]

Jun 3, 2002, 5:25am

> In LifeForms there's a 'Pin to floor' option. The shappy.seq is pinned to
> the floor. In the Lifeforms stage window it shows the Pelvis (center of
> avatar) as being on the same level as the stage. If 'pin to floor' is off
> then the avatar feet are on the stage. I'm trying to remember exactly the
> problem I was having when first testing Accutrans for Wayne. If I recall
the
> pin to floor option had to be on for my seqs to work correctly. I was
having
> the same type of problem where the pelvis would stay in the same spot if
> 'pin to floor' was off.

In accutrans, pin to floor just brings the av (lowest contact point, feet
for example) to the stage and sets the altitude. In lifeforms the pin to
floor brings the hip to the stage. The pin does not get saved in the seq at
all for either .. its just a method of altering vertical position. I
understand you can change how pin to floor works in lifeforms to make it
total av pin or hip pin.

Glitter

Jun 3, 2002, 5:59am

[View Quote] Hmmm... from my experience with Lifeforms the pin to floor option does make
a difference and must save some setting within the seq file. Pre 3.3 if an
seq that I created in Lifeforms didn't have pin to floor on then at times
when the seq played in AW the avatar would go below the ground level even
though it didn't when I played it in Lifeforms.

You're correct, in Lifeforms it brings the 'hip' to the stage. Actually
though I think it's halfway point of the bounding box of the avatar in the
seq. When you look at an .seq file in Lifeforms all it shows is a box. Pin
to floor always puts middle of the box on the stage. It's kind of like the
bounding box determines eye level, 3/4 the height of the box no matter where
you put the eye in LifeForms or Accutrans.

I haven't played with 3.3 at all yet. I sure wish Roland had asked me to
beta test as this is what I would have worked on and had figured out already
.. Looks like you really hit on something with the scale though. I only wish
that AW would provide some specifics about how the new engine works, what
program(s) can generate the seqs (aninations) and such. I'd be very
interested to know if seq's created in TrueSpace still work like they used
to. I never used TS to make them so I wouldn't be one to test that aspect of
this. Hamfon might be the person to ask?

If you want to send me any seqs you'd like me to test in Lifeforms to cross
reference your results and / or save from Lifeforms directly, I'd be happy
to. I use the studio version.

Casay

Jun 3, 2002, 6:21am

> Hmmm... from my experience with Lifeforms the pin to floor option does
make
> a difference and must save some setting within the seq file. Pre 3.3 if an
> seq that I created in Lifeforms didn't have pin to floor on then at times
> when the seq played in AW the avatar would go below the ground level even
> though it didn't when I played it in Lifeforms.

Thats usually a scale factor between COB and RWX

>
> You're correct, in Lifeforms it brings the 'hip' to the stage. Actually
> though I think it's halfway point of the bounding box of the avatar in the
> seq. When you look at an .seq file in Lifeforms all it shows is a box. Pin
> to floor always puts middle of the box on the stage. It's kind of like the
> bounding box determines eye level, 3/4 the height of the box no matter
where
> you put the eye in LifeForms or Accutrans.
>
> I haven't played with 3.3 at all yet. I sure wish Roland had asked me to
> beta test as this is what I would have worked on and had figured out
already
> . Looks like you really hit on something with the scale though. I only
wish
> that AW would provide some specifics about how the new engine works, what
> program(s) can generate the seqs (aninations) and such. I'd be very
> interested to know if seq's created in TrueSpace still work like they used
> to. I never used TS to make them so I wouldn't be one to test that aspect
of
> this. Hamfon might be the person to ask?

I wish exactly the same too ... both that they asked you to beta test AND
some specifics! I tried in TS but the problem there is I dont make my seqs
at 1:100 scale to it makes getting scale tricky to get right. Maybe Hamfon
could take a look at seqs too?

> If you want to send me any seqs you'd like me to test in Lifeforms to
cross
> reference your results and / or save from Lifeforms directly, I'd be happy
> to. I use the studio version.

Excellent thanks ... I now have lifeforms as well :) ... its pretty obvious
its not the software and _IS_ bugs in the new engine ... I'm still at a loss
as to why they actually needed to change engines ... I don't see why they
did it in the first place. Was it a lisencing issue? Something else? Why
exactly?

Glitter

Jun 3, 2002, 7:53am

Another observation. If the seq has an accompanying end seq and that end seq
doesnt work then the original which otherwise would ... now does not. Eg; if
you fix a walk sequence to work in 3.3 and you test it before fixing the
endwalk then you wont see the fix. I found I had to delete the endwalk from
avatars.dat before I could bug find the walk seq.

All of these little fixes do not fix everything. I still have some sequences
which don't have vertical movement but all things are set correctly.

Another interesting thing was I fixed a jump sequence and it worked as
normal. When I lengthened it by 5 blank frames without changing anything
else. It no longer worked. Take out the extra five frames and it worked
again. Absolutely no explantation for that one!!

Glitter

Jun 3, 2002, 3:03pm

I went into MoonBeam and tried the "Fate" avatar (couldn't find Philiris). I used the Jump gesture and it seemed to move vertically.

I have, however, noticed that the Walk SEQ doesn't seem to be doing it's usual bobbing thing on any avatars I've tried that usually
did bounce a little...

-Agent1

[View Quote]

Jun 3, 2002, 3:54pm

That's exacly how I have it, but it still it flips though 0
in 3.2 (pelvis tag, flip over Z axis).

I haven't found out yet why it worked on some models, where I
used the same (-180=>179 with no frame between) for wheels,
balls and stuff like that, but not on this one. Maybe, because
those didn't use Pelvis and Z but used either different tags
and/or the Y axis.

If you don't believe it, try LemurAnn with "Joy" in MoonBeam
or Labirint or whereever you find it - once in 3.2 and once
in 3.3 :)



[View Quote]

Jun 3, 2002, 3:56pm

Philiris is (with 1152x864) somewhere in the upper third of
the second column :)

[View Quote]

Jun 3, 2002, 4:49pm

<snipped>
"> Casay ... maybe we could ask Wayne to put in something such that if a key
> frame is specified for one dimension that keyframes are automatically set
> for all the others (or turn them on as default or whatever).
>
> Glitter
>
<more snipped>

I'd sure hate to ask Wayne to put something extra into Accutrans to 'fix'
what AW broke. Yes, it's obviously a bug. Wayne has done so much for this
community yet no one wants to pay the nominal $20 fee to register the
program. Maybe he'll make something for the people that have paid for it.
Who knows. I've always and will continue to try to talk him into making the
program expire after 30 days and have people then pay for it. He has done so
much to support us yet no one wants to support him. :-(

My .02 is all..
Casay

Jun 3, 2002, 5:23pm

> I'd sure hate to ask Wayne to put something extra into Accutrans to 'fix'
> what AW broke. Yes, it's obviously a bug. Wayne has done so much for this
> community yet no one wants to pay the nominal $20 fee to register the
> program. Maybe he'll make something for the people that have paid for it.
> Who knows. I've always and will continue to try to talk him into making
the
> program expire after 30 days and have people then pay for it. He has done
so
> much to support us yet no one wants to support him. :-(
>
> My .02 is all..
> Casay

I still remember when you walked me through the beta version of it so long
ago. I think I paid $69 for it back then. You are right he does do a lot ...
I'd lost sight of that for a while. A short term trial makes sense ... or a
limited version.

Glitter

1  2  |   Order by index