The Undertals Production Focus

Vishka Stu­dio pro­duc­tion focus on the “The Under­tals” and 10 pro­duc­tion tips.

This arti­cle is about the mak­ing of The Under­tals pre-production and we have also included some pro­duc­tion tips which we hope will be use­ful to our readers.

Pro­duc­tion Focus

(Fig. 1: final pro­duc­tion image of the Undertals)

The Under­tals 01		The Under­tals 02

The pur­pose of cre­at­ing the Under­tals was to have a pre-production pack­age to present to even­tual investors and pro­duc­ers in order to estab­lish a co-production oppor­tu­nity and pro­duce “The Under­tals” as a tele­vi­sion series.

• The first thing we did was to define the scope of the task and set some objec­tives, some of which were the following:
• Tar­get Group: we decided to aim for a rather wide audi­ence group of ages 5 to 12 while still mak­ing the series acces­si­ble to other ages.
• Tar­get Qual­ity: although the objec­tive was to cre­ate a series for broad­cast it needed to be on par with or have a bet­ter visual qual­ity than some of the works seen in the inter­na­tional markets.
• Fea­si­ble Project: while aim­ing for a high-quality visual we also had to design and plan for a very tight pro­duc­tion sched­ule and high out­put that could cre­ate the nec­es­sary mate­r­ial in a tight deadline.

Tip No. 1: The scope of your project is cru­cial to cre­at­ing a high stan­dard of work. Never stop plan­ning and try to fore­see and plan as many details of your project as pos­si­ble. Some of the most impor­tant steps are the project scope, file-naming con­ven­tion and your project direc­tory struc­ture. We have tried sev­eral com­bi­na­tions of these and we are still opti­miz­ing them for every project.

Tip No. 2: Put every­thing on paper. We tend to think that we will remem­ber where we put our files and what we named the. Believe me your mem­ory will not retain 80% of what you are doing in two weeks (and I’m being opti­mistic!). Write it down and if you are col­lab­o­rat­ing with a group try to cre­ate processes that would allow them to write down what they have done about the project. Word, Excel and Project or what­ever soft­ware you tend to use are your friends; USE THEM.

With these goals in mind we started the task of research and look devel­op­ment by look­ing at a lot of children’s series and gauge the visu­als and sto­ry­lines they had. We also scoured the inter­net for images of envi­ron­ments and char­ac­ters that we felt were unique and could be close to our style.

(Fig. 2: sam­ples of images from the internet)

After that came the writ­ing stage; we wrote exten­sively try­ing to flesh out ideas and come up with a story that would please us, writ­ing a story, script, and syn­op­sis as well as char­ac­ter and envi­ron­ment descrip­tions. With this infor­ma­tion in hand we were ready to start production.

Tip No. 3: Even if you are work­ing alone the writ­ing stage should not be omit­ted. What is impor­tant in this process is that even if you are not a great writer just sit­ting down and try­ing to shape your story or char­ac­ters will help you in find­ing ideas and inspi­ra­tion that would not have been pos­si­ble if you just started sketch­ing or mod­el­ing. Writ­ing will cre­ate ques­tions in your mind and find­ing answers to these ques­tions will make your project more robust.

We pro­vided our 3D artists with detailed descrip­tions of our char­ac­ters and envi­ron­ments and went through a series of sketches, prob­a­bly 3 to 4 revi­sions for each object. Although we wanted the con­cept designs to be very inspir­ing we knew that we did not need to make them per­fect since it would be eas­ier to final­ize them in 3D. Sto­ry­boards were also cre­ated in this stage since we were plan­ning on cre­at­ing a trailer for the series.

Tip No. 4: After each stage of pro­duc­tion is com­pleted cre­ate a pre­view of that stage. Ben­e­fits: you will have a mak­ing of your project, you will know exactly what assets have been cre­ated and at what stage they are, and you can show it to col­leagues and friends for feedback.

(Fig. 3: inter­me­di­ary sketches of char­ac­ters and environments)

Dogholoo sketches	Mooboland sketch
Pre­lim­i­nary island sketch	Pre­lim­i­nary island sketch

(Fig. 4: final sketches of char­ac­ters and environments)

Topol color concept

Dogholoo color concept

Before this pro­duc­tion we had had a few prob­lems with our pro­duc­tion pipeline and our file and asset man­age­ment, so we decided to define a clearer process and a bet­ter file and folder nam­ing con­ven­tion. We some mate­r­ial on the web con­cern­ing pipeline def­i­n­i­tion and file-naming con­ven­tions and although they were just a few arti­cles, it really helped us on mak­ing some decisions.

Since we were work­ing in Maya we needed to use the Maya project fold­ers sys­tem as much as pos­si­ble to pre­vent break­ing links between assets and ref­er­ences. We also decided to use file-referencing, which we had already used exten­sively in Sof­t­im­age suc­cess­fully, but was rather new to us in Maya. The fil­ing sys­tem we decided on finally, with great help from dif­fer­ent forums was based on the fol­low­ing system:

Most of the names are self-explanatory except for Library and Build. The library is in fact a con­tainer for fin­ished assets such as mod­els, shad­ing, rigs, etc which will be ref­er­enced into the dif­fer­ent shots. The build folder con­tains works in progress of our assets and the Shots folder will con­tain all the shots in sub­fold­ers (shot001, shot002, …) each in turn con­tain­ing fold­ers for ani­ma­tion, effects, …. Each of these fold­ers will in turn be a Maya project con­tain­ing spe­cific data for that process (for exam­ple “effects”).

The pro­duc­tion started with the mod­el­ing process where the main empha­sis was on fea­si­bil­ity; we needed to have believ­able mod­els with minute details but we couldn’t allow the meshes to get too heavy since in a series pro­duc­tion you don’t have the time to ani­mate very dense meshes. We even decided to forego cer­tain fea­tures such as dis­place­ment which we believed was not cru­cial to our mod­els’ visu­als in this project. The crea­tures and their envi­ron­ment were mod­eled in Maya and some of the details were done in ZBrush but we tried to keep these simple.

Tip No. 5: Units! If you skip this part you will regret it later. Define your units and stick to them. Defor­ma­tions, hair, rigs, dynam­ics and cloth­ing will all depend on this small issue and if you don’t think about it in the begin­ning you will suf­fer the con­se­quences later when you need to resize that rig and it just breaks. Cre­ate a unit file where you cre­ate a phys­i­cal ruler as well as a cube rep­re­sent­ing your unit (1x1x1 m) and place all your char­ac­ters and other objects in this file to ascer­tain their size and com­pare it with the rest of the project.

(Fig. 5: pre­view of models)

Topol model preview

Dogholoo model preview

Since the look of the Under­tals was sup­posed to be very car­toony but fleshy all the tex­tures were painted by hand and we decided not to use a sin­gle ready-made tex­ture. We used Pho­to­shop, ZBrush and Body­Paint 3D to paint our textures.

The next stage we tack­led was the char­ac­ter and envi­ron­ment shad­ing. There was an empha­sis on sim­plic­ity here as well and we tried to avoid cer­tain time and cpu-consuming such as FG, GI and sub-surface scat­ter­ing but dur­ing our tests we came to the con­clu­sion that we were los­ing qual­ity by avoid­ing all these tech­nolo­gies. We finally came to the con­clu­sion to use sub-surface scat­ter­ing for the char­ac­ters’ skins and FG for the envi­ron­ment since the increased ren­der times were min­i­mal after some tweak­ing and tests but the rise in qual­ity was quite noticeable.

Tip No. 6: Sim­plify! Nowa­days 3D soft­ware has become extremely sophis­ti­cated and you have a plethora of won­der­ful tools at your reach. DO NOT use them all. Just because GI ren­ders are cool you should not try to use them in your project. Always try to use the sim­plest tech­nol­ogy that can pro­vide you with 95% of what you need.

We ren­dered our assets in sev­eral passes, using stan­dard Maya shaders, occlu­sion and other stan­dard passes and tweaked them exten­sively in com­posit­ing to obtain the look we were try­ing to obtain; about 30% of the final look was obtained in post-production.

Tip No. 7: Always use the 80/20 rule in your projects. This rule states that 80% of each task should take up 20% of the time and 20% remain­ing should take up the remain­ing 80% of your time. What this means is that you should first try and very rapidly give shape to things and keep the bulk of your allo­cated time for tweak­ing and work­ing on small but sig­nif­i­cant details.

Dur­ing all this we were also try­ing to improve our cur­rent pro­duc­tion pipeline and cre­ate a more robust work­flow. Another objec­tive was to speed up our pro­duc­tion by test­ing new ways of doing things. One of the issues we faced was the rig­ging of the char­ac­ters which took about 5 to 7 days for each char­ac­ter and was not an opti­mal time for us. We tried a few plu­g­ins but none of them gave us the con­trol we needed and most of them had small issues or large learn­ing curves which would have been pro­hib­i­tive for this project.

Tip No. 8: Ref­er­enc­ing is your friend. By ref­er­enc­ing your files the stor­age space used by your projects will decrease by an order of mag­ni­tude. Prox­ies and ref­er­ences will allow you to use the cor­rect assets in your shots with­out hav­ing to man­u­ally import them into every shot. Although ref­er­enc­ing may have a few issues, if done right it can save you a lot of valu­able time.

We decided to try script­ing the process and were able to speed up some of the rig­ging issues but we were not able to auto­mate the whole rig­ging process in a way that would save us a lot of valu­able time. This is still one of the sore points in our pro­duc­tion and although the rigs we cre­ated for the char­ac­ters were rather thor­ough, we are still work­ing on automat­ing the rig cre­ation process.

We nev­er­the­less cre­ated very sophis­ti­cated and com­plete rigs with squash and stretch, IK/FK switch­ing that used soft bod­ies and Maya mus­cle, par­tic­u­larly for their fat bel­lies, to allow them to be ani­mated eas­ily in spite of their short stature and very short and stout limbs.

Tip No. 9: Try to auto­mate repet­i­tive tasks as much as pos­si­ble. Your time is extremely valu­able; don’t waste it by try­ing to rename 1505 meshes in Maya. Either learn some script­ing or try to ask a friend to write you some code in exchange for a free model or a sam­ple rig, it will be worth your time.

(Fig. 8: movie of rig in action)

We also cre­ated some turnta­bles and test ani­ma­tions for the characters

(Fig. 9: images and movies of char­ac­ters animated)

One of the main parts of this whole process was the com­posit­ing process which really gave life to our ren­ders and pro­duced the final look of the Undertals.

(Fig. 11: images of ren­der layers)

Tip No. 10: At the end of your project cre­ate a solid archive of your project and delete your ren­dered passes and any other inter­me­di­ary files such as incre­men­tal saves; this will save you lots of space on your com­puter. Cre­ate a final post-mortem doc­u­ment describ­ing your project out­come, the issues you had, the plu­g­ins used and its short­com­ings. Write what you will do for your next project to address these issues and read this doc­u­ment at the begin­ning of each new project.

We are cur­rently cre­at­ing a one minute trailer for The Under­tals which will be ready in a few weeks and we hope to be able to present it to the CG Com­mu­nity in the near future.

Tip No. 11:Self-improvement is vital if you want to become a pro­fes­sional in your field of work. 8 or 9 years ago it was very dif­fi­cult to find tuto­ri­als or learn­ing mate­r­ial any­where but now the wealth of infor­ma­tion you can find on the inter­net and sites such as Dig­i­tal Tutors and Gno­mon are stag­ger­ing. Always try to learn new ways of doing things, improve on your processes and opti­mize your work­flow constantly.

The pro­duc­tion team con­sisted of:
Exec­u­tive Pro­ducer and Project Man­ager: Reza Ghobady
Project Spon­sor: Vishka Assayesh
Orig­i­nal Idea: Ali Chenari
Direc­tor: Ali Chenari
Char­ac­ter Con­cepts: Hamed Kamali
Envi­ron­ment Con­cepts: Kian Kiani
Mod­el­ers: Bahram Najand, Faraz Sayadi, Reza Ramezani
Tex­tures: Bahram Najand, Faraz Sayadi
Shad­ing: Faraz Sayadi
Rig­ging: Faraz Sayadi
Ani­ma­tion: Abed Nat­taj, Alireza Shahram­far
Light­ing, Ren­der­ing, Com­posit­ing: Faraz Sayadi

We loved work­ing on this project and hope you enjoy it as much as we did.

You can see some test ani­ma­tions on our YouTube chan­nel: http://www.youtube.com/vishkastudio as well as on our CGPort­fo­lio page: http://rezaghobady.cgsociety.org/gallery/

Our web­site: http://www.vishka.com

Our blog: http://www.vishka.com/blog

Line Break

Copyright 2010 - Vishka Studio. All Rights Reserved

• Projects
• 3D Models">3D Models
• Concepts
• Shorts Stills
• About Us
• Contact Us