[Ldraw-content] Forwarding thoughts about the MOTM/SOTM contest

[Timothy Gould]

OK. Some more responses.

> > But there isn't and probably never will be a reason for them to start.
>
> Prizes?

I really doubt that would be enough unless we were talking about
seriously expensive prizes.

> > There is no way that LDraw competitions will ever be as prestigious as
> > real brick ones because too many people have an animosity towards it.
> > The amount of criticism I used to get on CSF (and still do to a lesser
> > extent) is extraordinary. Look at point 1) here
> > (http://fbtbforums.net/viewtopic.php?
> > p=223624&sid=b5909870de9ea024caf8e5c2f3d8ef08#223624)
> > and see the comments of a long term and quite respected AFOL (Dave
> > Eaton). The only theme which seems relatively accepting of CAD seems
> > to be trains.
>
> Well, actually, I build physically too.  I thought everyone did (I
> know Burkhard "Primus" (brickcommander.com) doesn't though).  Dave
> Eaton's points are exactly why I would want a panel of judges to go
> over an entry.

I'm not sure it's entirely due to ignorance. Prejudice doesn't go away
so easily.

> I think animosity is directly proportional to quality.  The higher
> the quality of the entries and winners, the higher the prestige and
> acceptance by non-LDraw users.

I agree that it makes a difference to some people. I know from my own
work that if I build things which are miles above the typical quality
of stuff on CSF I can get some people out of their anti-render shells
but it's very rare. When I post in real brick it almost guarantees me
more responses even if the model is inferior.

> >> And to complicate matters more, Mladen, IMO, could win every month if
> >> he submitted his models.  Again, this goes to show the disconnect.
> >> Why isn't he submitting his models?  He is both a genius MOC builder
> >> AND an LDraw user!
> >
> > Mladen hasn't built anything for a very long time and isn't likely to
> > in the near future due to his studies. He's not a renderer either so
> > although his stuff is amazing and very well modelled he's not likely
> > to be entering the SOTM any time soon.
>
> Well, he's posted 4 MOC's and one LDraw file from this year:
>
> http://www3.sympatico.ca/mladenpejic/MMP.htm

I've only seen the MMP this year. I'll ask him next time I speak to
him if he'd be willing to enter but he isn't even building in brick at
present.

> Certainly a contest winner.

Agreed.

> >> And personally, I know the idea behind the simple unappealing model
> >> renders is to make them fair, but they are not fun to look at.  I
> >> think they need to be ray traced with complimentary scenes and
> >> lighting.  And it wouldn't hurt to have 360 degree spins of the
> >> models, and maybe even take apart spins too.  Again, I think this
> >> will increase the prestige.
> >
> > I agree with this point but it ends up being a lot of work for the
> > people running it. I recently had to coordinate a set of standard
> > render settings for a large group of people and farming out is not so
> > easy and ends up taking a fair amount of time for each model.
>
> I have about 400 computers I use to render my POV-Ray animations.  My
> last animation had 6000 frames (averaged into 200).  I render Anton's
> animations.  I've offered my grid to other users (just send me the
> projects to render).  I can't take hordes of render requests, but I'm
> nowhere near that now.  It has sat idle for about 2 months.

OK. Well that takes away my argument there. We just need to develop
top quality settings (based on Koyan's tutorial would be my
suggestions) and we're good to go.

> >> And yet, if those things are allowed, then they should also be part
> >> of the judging.  For example, a scene that uses fog poorly should be
> >> docked.  What if someone works hours and hours developing their own
> >> light effects, and someone else just uses the POV-Ray lens flare
> >> macro?  Are these taken into consideration?  This provides further
> >> evidence that the scene contest is too unfocused.  It needs some sort
> >> of guidelines to take these variables into account.
> >
> > I don't think either should be judged on POV-Ray technical merit but
> > on quality of the scene. Otherwise you are measuring people's
> > patience, research skills and computing access rather than the model
> > and layout of the scene. If people are really interested in
> > complicated rendering then they can enter a "proper" raytracing
> > competition.
>
> Ok, the "hours and hours of writing a lighting effect" is probably a
> bad point.
>
> But how is technical merit and quality different?  If the scene
> contest isn't a raytracing contest, what is it?  Anyway, I discuss
> what I think should be done about this below.

It's a CGI scene contest based on Lego bricks. I don't see any reason
for it to change from that. I'd rather see something great with
hand-editing in PS than something boring but perfectly rendered with
no post-proccessing.

> Ok, now I'm just going to list specific contest rules and proceedures
> that I think need changing or adding.
>
> ---
>
> This is from the contest rules:
>
> > Entries must use your own model. You must have express permission
> > to use someone else's model(s).
>
> So I can use my own model *and* someone else's if I have their
> permission?  This rule is vague.

If this is for SOTM then here is an example
http://www.brickshelf.com/cgi-bin/gallery.cgi?f=158316

All the modelling is by me but the rendering and post-production is
all done by Adrian. This amounts to essentially a joint effort.

> I'm just curios why the following rule exists:
>
> > Images may not include any represention of the trademark LEGO(r)
> > brand insignia.
>
> Is it to avoid trouble with TLG?  All of my renders include the logo
> and I asked Jake McKee about it years ago and he said it wasn't a
> problem as long as I'm not making money off of it.

What Jake says isn't neccessarily a guarantee that the TLG lawyers
won't complain. I agree it is unlikely but unless we get something in
writing I would keep that rule.

> ---
>
> New rules applicable to both contests:
>
> All entries must provide ldraw files (even scenes) and allow them to
> be downloaded by the public.

I don't agree with this. I think the scenes should remain as is.

> All entries to both contests need to be unpublished LDraw files or if
> published, be published within the last year (or something).  MOC's
> that were not in LDraw form but were published as pictures may be
> submitted if the LDraw file is new.  (I think this makes it fairer
> for people who don't have a huge arsenal of LDraw models, and it
> gives people who haven't digitized their models a reason to do so--
> the ultimate goal is to get more LDraw files.)

I'm not sure this is neccessary. The continual improvement in the
parts range and building techniques means that most older models will
find it hard to compete anyway.

> ---
>
> Changes applicable to both contests:
>
> Add 2nd, 3rd places along with honorable mention and include all of
> them on the webpage with the 1st place winners (or something).

Agree.

> Space the contests out, one every other month or one a quarter.
> Using honorable mentions will increase the number of "winners", and
> by decreasing the contest frequency, it will level out more.  Another
> useful side effect is that busy people have more time to work on the
> models.  This also sort of solves the problem of submitting models
> month after month.  This may mean "of the month" may be dropped from
> the titles.  Allow each participant one entry for each month of the
> competition (2 month competition = 2 entries).  Or something.

Definitely agree.

> Voting doesn't start until after the submission deadline.

Agree.

> Set a theme for each contest and give it some specific qualities it
> will be judged by.  Some theme specifics could include buildings,
> civilian space, racers, etc.  This has a lot of potential to generate
> interest.  Possibly    choose the themes by voting or discussion on
> Lugnet.

I think I've been convinced that this is a good thing.

> Ask some of the prolific LDraw users: Mladen, Burkhard
> "Primus" (brickcommander.com) to submit their models.  Even if they
> win month after month, someone will eventually come along with
> something that challenges them.  I think this will increase the
> prestige of winning.

Still not sure I like this idea.

> Young authors (if there are any) get their own categories as well.
> Maybe.  Just an idea.

Not sure that's neccessary. The only difference normally between the
quality of younger and older builders MOCs is typically availability
of parts. Since CAD removes that issue I don't see any need for it. As
an example Mladen is 20.

> Give out prizes, preferably official Lego sets related to the theme
> of the contest, or just a gift certificate to shop.lego.com.
>
> ---
>
> MOC changes:
>
> I think the "Model" contest should be renamed to MOC and make it
> clear that it is a MOC contest.

Agree.

> Set a specific size for the contest, or choose winners for certain
> sizes.  Either way, make it so that someone with a 10000 piece MOC
> isn't competing with a 25 piece MOC.  If there are dozens of small
> entries, possibly give out more honorable mentions or 2nd and 3rd
> prizes.  Give an honorable mention to everyone who submits a 10000
> piece MOC unless it was obviously thrown together.  (I'm using sizes
> very loosely because I have no idea how many pieces a 4 foot model
> would really take)

Make them part count categories. I would suggest <150, <1000 and anything goes.

> Build a set of preset POV-Ray scenes users may choose to have their
> model displayed in.  Just a simple backdrop gradient of various
> colors, with the choice of hard or soft lighting.  Use the same
> camera angles already on the rules.  However, along with the entry
> defined camera angles, let them define custom lighting for those
> renders as well (even if it means that the MOC's have lighted elements).
>
> Render all models in POV-Ray with aliasing and radiosity.  Build 360
> degree views of the models.
>
> Voting is split between a panel of judges and the public.  The judges
> will analyze a MOC and will give it points for:
> - Instructions built into the LDraw file (does the LDraw file produce
> instructions, and are they effective instructions)
> - Posable or animation ability (are movable parts in sub-models, are
> they centered around origins, and is it nested correctly)
> - Realistic construction (is it a physically realistic model that can
> be built and shown at BrickFest?)
> - How closely it matches the theme and stated goals of the theme
>
> The public would vote on how cool it is.

Agree if possible.

> ----
>
> New scene rules:
>
> The Lego models in the scenes must have source LDraw files.  (and
> must be made available to be downloaded by the public)

Disagree.

> That way these are disqualified: http://forums.cgsociety.org/
> showthread.php?t=243823 and http://forums.cgsociety.org/
> showthread.php?t=111101.  Anton's library is still an option since he
> has an LDraw converter.  Unfortunately, this disqualifies some of the
> best looking scenes I've seen (Anton's, and another user whose name
> eludes me, as well as some of my own...).

Which is why I disagree. I would rather have a best scene competition
based on Lego than a more limited competition.

> Providing scene project is not required (the goal is to get LDraw
> files, not POV-Ray or other ray tracer submissions).

I would rather see the POV source than the LDraw source but I'm happy
seeing neither.

> ---
>
> Scene changes:
>
> Possibly have a of panel judges vote in addition to the public (like
> the MOC competition).  Judges would award points for:
> - Quality of lighting (is it low contrast?)
> - Quality of render (is the image grainy?)
> - Realistic effects (is the fog a blanket or can you see wisps?, does
> it use depth of field or motion blur effectively)
> - Textures (chrome, metal, rubber, index of refraction on transparent
> parts)
> - Scene quality (is it balanced, does the eye flow well, is the
> camera placed at a good distance to avoid distortion)
> - How close the scene comes to the stated theme
> - MOC qualities (the same qualities a MOC is judged by: instructions,
> animation, and realistic)

I really strongly disagree with these criteria. Judging on technical
merit is not something I am interested in. If the techniques make a
better scene (as they usually will) then people will use them. I think
it should be based purely on the final result and not how it was made.

As a previous winner of SOTM I would stop entering if the criteria
were changed to bias more towards technical issues. I am happy to use
them but only insofar as they help me achieve an image that I like.
I'm not going to spend days (nor can I) rendering a model just to
achieve something POVray pure that I can do in 2 minutes in GIMP.

Not to mention that it might discourage creative use of "bad"
lighting. Would this model
(http://www.brickshelf.com/gallery/timgould/Wallpapers/mms_wallpaper.jpg)
lose points for the harsh lighting?

> Some of the things that are easy in Anton's library, but I have no
> idea if this is easy using L3P.  Maybe it could spur some development
> activity...
>
> Public votes on coolness.

I think this should be the only vote.

> Listing things models and scenes are judged by by also gives would-be
> competitors something to strive for.  And I would LOVE for people to
> strive to make their LDraw models animatable!  I can provide
> tutorials better than my current one to help people with this if
> needed.  I still haven't given up hope of getting models from the
> community that I can use.

I think that setting up models properly for animation is a good
section of judging for MOTM. There should be a balance between
technical and model skill.

> Obviously this is all very ambitious and we are all volunteering our
> time.  So I understand if this is asking for too much.  I know I
> can't pull it all off.  But this is what I would like to see happen
> if changes do happen.
> James

Fortunately a lot of these changes can be made with what is currently available.

Tim