Friday, 27 May 2011

so so

animation deadline week for the deep, my team is extremely busy with their animating and i today collected all the finished shots from people and gave extended deadlines to get the excess shots finished.

this chart shows in colour form what we have completed what is being worked on and what is in need of attention.

i believe we have made great progress and am looking forward to the end product but i am constantly aware of the amount of time we have left to achieve this.

rig demo

here is a demo of how the rig works and the controls involved.

i am extremely happy with the result as rigging is probably my weakest area. but after this i feel much more confident and feel that i will be doing a lot more in the future.

rigging process and breakdown

so here is a brief description and breakdown of the rig that i produced for Patience, i created the skeleton and all the controls, i also painted the weights for the character.

i found the process of creating the FK Spine the hardest but most rewarding as through this process i was able to further understand the task of rigging and i now have a new set of tools to use when i rig in the future.

Rigging notes

so i thought that seeing as i noted my progress and techniques whilst i was doing the rig i though i would post them on here to show my own process.

patience rigging

so i offered to help Chris Caldow's film in the form of rigging. they have their character modelled and UV'd all that is left is for it to be rigged. as they were struggling to find a suitable artist i offered my services alongside chris learning to rig himself.

i began by looking into what the character would be doing, mainly just simple walking, sitting down and use of arms hands and fingers, so i set out to create a simple humanoid rig with some more attention placed upon the hands.

this video was my reference for the characters spine, in order to get a nice arch and amount of control for when the character sits down.

this video was my template for the head and neck

Friday, 20 May 2011

the code that saved the film


mel script:
// Batch rendering workaround V3 - John Mather (NextDesign) // Author page: // Check for updates here: string $filename = "render";  int $startFrame = 1; int $endFrame = 24;  string $directory = (`workspace -q -rd` + "images/");  // check for render panel. Found here: string $renderPanel; string $renderPanels[] = `getPanel -scriptType "renderWindowPanel"`;  if(size($renderPanels))      $renderPanel = $renderPanels[0]; else {     $renderPanel = `scriptedPanel -type "renderWindowPanel" -unParent renderView`;      scriptedPanel -e -label "Render View" $renderPanel; }  for ($i = $startFrame; $i <= $endFrame; $i++) {     currentTime $i;      renderWindowRender redoPreviousRender renderView;      string $concatFilename = $directory + $filename + "." + $i;      if (`getApplicationVersionAsFloat` >= 2011)         // Thanks to nowayfra on creativecrash for his workaround         catch(eval(renderWindowSaveImageCallback ($renderPanel, $concatFilename, `getAttr defaultRenderGlobals.imageFormat`)));     else         renderWindowSaveImageCallback ($renderPanel, $concatFilename, `getAttr defaultRenderGlobals.imageFormat`);      print ("Saved " + $concatFilename + "\n"); }  print ("Completed rendering of " + (($endFrame - $startFrame) + 1) + " frames.\n");

next morning.

so up bright and early at 8am with fresh ideas of how to tackle this problem. i decided to export all the models and import it into a fresh scene such as how we solved the occlusion problem. this did not work either and im at a loss through scouring the internet. i have decided that it is either there is a big problem with maya 2011 batch rendering mental ray, or that the project file i have from keith is full of imported objects trying to link back to their original sources. obviously if they are trying to link to their sources they cant and that is why the render fails, but it doesnt explain why the render works as 1 frame.

i phoned keith to consult him and arrange to meet on monday to try and use the render farm. it was in this discussion that i discovered keith had had exactly the same problem. he had managed to find a code though on a forum that forced maya to batch render through the render view, in othere words bypassing the batch render altogether.

SUCCESS!!!! finaly after 12 + hours of trying to get this darn thing to work. the only draw back is time as it take s long er as maya is producing a fully rendered preview each frame. but it does work and i am currently rendering batches of frames on 2 computers.

the other problem is that after about 20 frames-45 mins- some of the textures turn bright red and u have to preview the rendered images make sure their all ok then restart maya and restart the batch render from a new start frame. as a result this means that i have to render about 50 frames a time and keep an eye on the render viewports to make sure the render hasnt corrupted.

slow but definately getting there :D.

long night

soooooooo, i came home from work to find the render was useless. it had rendered all that i asked but only the first frame was textured all subsequent frames lost their textures.

so i ran a test render to see if i had done something wrong...nop didnt work.

i then alterered the frame padding and numbering as my frames are rendering from -140 i thought it could have been a computing error due to the minus numbers. nooo didnt help.

i looked into reducing the render size as we are rendering 1280x533.... no

i removed final gather from the render luck.

i began to research online and found that if u were using an ocean shader in 2011 from a previous version of maya that could cause trouble so i deleted it and no result.
then i came across a tab in the render settings that was advised when using an ocean shader to stop crashes. something to do with using mayas base shaders. no did not work.

by this point i had reset my preferences in case a bug had found its way into my program no avail.

i uninstalled maya and re-installed but still no luck.

after more hunting and more testing i came across the hotfix 3 for maya and the mental ray stand alone renderer. so i downloaded and had to then un install and re-install maya again.

no louck still. so i installed maya on a different computer and attempted the render only to have the same problems. now maya was crashing the mental ray render after the first frame causing hours of acheing in my head.

although i persisted and attempted to reload the very original scene that keith had given to me, but this did not work so i tried a ball moving in an empty scene rendered fine in maya software and just as fine in mental ray.

it is now around 3:30 am and im losing the will to live. so i decide to call it a night and tackle the problem in the morning, dissappointed that i could not render the scene.

tuesday blooming tuesday

so big 'ol meeting with dan and the whole cohort to find out all that we have to submit and how, twas very useful but highlights the fact i need to do more.

on the other i hand i was able to collect the rendered frames of shot 1, unfortunately there were some irreparable problems that has resulted in a re-render.
the mountain range and sky were not rendered seperately to the town so they mountains and sky would not be editable in the composite. the occlusion included the skyplane which means the sky when composited would be plain white. so the only way to work around this would be to rotoscope out the entire sky and then seperate that way. the main reason tht i want these layers seperate is that the mountain range was not colour matched completely so it needed editing in post.

so to save time i offered to take half the file to render myself and keith would do the other half.

when i got home to my computer and opened the file i realised there was a lot more wrong with the scene. there were floating buildings that casted unwanted light, floating objects, but more importantly where i had asked keith to reduce the poly count in the scene he had deleted unseen faces. evidently unaware of the cameras view there were huge parts of the scene where u could see straight through buildings or that a building had been cut in half. annoyingly i contacted keith and informed him to stop rendering while i re-arrange and fix the scene.

i had about 2 hours to fix the scene and get a render running whilst i went to work. keith did offer to help fix the scene but instead of adding another layer of confusion i decided to do it myself.

so i managed to fix buildings textures and models then set a render to run. planning for a long night ahead to get the shot done.

animation so far

this is a rough playblast edit of all the animation so far from across the whole group!

this is very encouraging as you can begin to see sequences being formed, very nice for me to see my vision unfolding.

shot progression list

so this list is my way of keeping track of our progress it is very similar to the layout of shotgun from term2 but i have total control over this now and it only shows what i want.

so it has the scene number with shots in the scene.
each shot has its number relevant to the scene and also how it appears on the storyboard.

there is the initial of the member who is animating the shot.
A= andrew(me)
B= ben
C= carlos
E= emma
F= fahran
K= keith
M= mel

then in the relelvant columns such as animation the "y" stands for yes or completed.

there is also a line called "in folder" which is for me to see which shots are set up and ready to be rendered out. this is important as i can prioritize my rendering.

there is also a playblast column this one is important as i am editing the playblasts together to give to our sound guys so they have the timing to work with without having to wait for final renders.

i am intending on giving the sound guys a finished playblast scene at a time so they can produce a completed soundtrack for that scene.

shots 19 21 22

so been busy animating and this is the result

roughly spent a day and a half on each shot but again i am extremely pleased with the level of animation given the time restraints i have placed upon myself.

so shot 19 was cool but i feel that maybe the pacing could be improved. shot 21 , the face ECU, i feel is great and i decided to exagerate the size of the eyes in order to obtain more expression from the somewhat lacking face.

22 is again good but could be faster maybe.

i animated this scene as virtually a continuous shot as the movements all flowed into one another. this saved me time when it comes to starting and finishing. then i simply pick the timeline to render the desired frames.

to render

so with this established i set keith the task of rendering the first shot, either through the render farm if this was possible or locally on his computer/laptop. big up to the keith man.

occlusion render problems

so the proposed layers for the film are as such

full light
ambient light

full light - retains all texture and light information with shadows.

ambient light - is flat light with no shadows to enable the compositor to enhance base colours, or wash out the shadows.

occlusion - add depth and substance to the object.

the occlusion with the town does not work for some reason all we get is a stark bright white screen. after messing around with the amount of samples and max and min distance we establish it is nothing to do with the shader as the objects do not hold the characteristic jet black appearance in the orthographic viewports. so what to do......

we had decided that the scene had probably become corrupt due to the amount of shaders and imports involved with the scene.

we discussed the issue among us and still no resolution until clym came up with simply exporting all of the objects as a .obj then re-importing them into a new scene. this worked!!!!

by doing this we were able to get rid of all of the shaders and let maya concentrate on just the town and the occlusion.

so with this i sat down and spent the morning exporting and importing geometry chunks at a time in order not to crash maya, which happens frequently on my laptop. :D


so we are deleting all of the elements from my animation that do not need to be there such as the town that is there for reference and shaders and such. in otherwords we are cleaning the scene before we import it to try and avoid unnecessary junk in the scene.

so the kid imports successfully yay but we now have the task of reassigning textures to the kid. we began with fahran and started reassigning the textures, in doing so we discovered that several nodes and shaders had either dissappeared or had been corrupted so fahran kindly redid the shaders in the morning.

this then lead to the thought of having to create a new shader network for everyshot that included the kid. fahran discovered a tool called export shader network this enables us to create a set of shaders for the kid that retain the texture information and in an empty scene we can simply import this shader network scene and it attaches itself to the character. this is extremely useful and will hopefully speed up production when it comes to rendering.

town progress

i am pushing keith really hard to get the town sorted out as this is the most heavy scene in terms of poly count and stuff thats going on.

he has finished with the location of models and such all that is left is to import the character. booom

week 3

so had some more meetings today and discussed progress with the group which has now come along leaps and bounds we are scheduled to have 17 shots finished by the end of the week at present we have roughly 7 with most of the others being blocked out.

everyone is putting in a lot of work and its really showing :D.

so for me i have animated shots 16 and 18 this week and here they are. with my personal animating i am planning on sticking to my schedule so i can get the bulk done and out of the way, then if we have excess time at the end we can re-render the shots with better animation but at present i doubt we will have time for this, fingers crossed.

im really happy with the animation i am producing as i would say that animation is my weakest skill. shot 16 has some particularly nice pace and weight to it and im happy with the way the character appears to be moving through the water.

shot 17 is good as well but lacks the anticipation, part of this could be due to the very restrictive blendshapes but i have to make do with this.

Wednesday, 18 May 2011


so the group wanted to be assigned their shots for the whole film so they could work on multiple shots not just the ones assigned that week so i divided all the shots up and everyone was mucho happy.

6- titles
40- anyone and everyone

week 2

so after week 1 and went and spent ages setting up the project folders for everyone to access their animation scenes from. it toook a long time but it was 100% necessary to provide clarity and continuity to the animation.

keith is still refining the town and it is looking great. and people are producing some great work
emma has pretty much finished shot 07 and has blocked out 08 i have given shot 10 to keith as he needs some animation to do ben has produced a nicely paced distant swim for 10 and 11 fahran has produced some animation for hots 4 and 5.
so all in progress.

problems it has come to light that the seagull that i have created and rigged in 2011 is not at all compatible with previous versions of maya. the rig is distorted and destroyed. i am using 2011 but the rest of my team are using 2010 as they argue that 2011 is broken and inferior to 2010. so to stop all the hassle and keep the project moving the seagull has been cut with the intent of adding him back in if we have time. dam u maya

after 1 week

damn after 1 week of animating i was able to finish shots 1 and 2 to level of animation that i am happy with.
on the other hand tasking the animators with such a tight schedule has not paid off and the rest of the group are still in the block out stages of animation. the town is taking longer than expected to complete as well and i am personally behind on creating the master projects for people to get their next weeks worth of shots.

meeting with ben

i met with ben a week before we are due back at college, easter break, and we broke down all of the shots and how long each one would take and who should animate it . this was a great move as it put the whole project in perspective and enabled me to prioritize the animation.

so according to my schedule

shot 1 and 2 will be animated by me.
shot 3 will be Mel
shot 4 and 5 will be Fahran.

scene 02

shot 7 and 8 will be Emma
shot 9 and 11 will be Ben
shot 10 will be me

this is 1 weeks work and will be the basis of a weeks workflow rota.

members get 1 week to animate the shots assigned then the shots are collected on the friday sent to the render farm over the weekend and then the frames are collected on the monday. the animators will be assigned their new shots on the monday and the rendered frames will be composited that week. this will create a rolling animation and rendering pipeline that will increase productivity.

the master plan

the kid is finished and i am now consolidating all of the assets so they are workable upon return to college after the easter break.

so i have set up a scene that contains all of the characters and the environment is being laid out by keith. i have the intention of simply importing the characters to the environment scenes.

my intention is to set up a project for each scene in the film , so 6 projects in total. within those projects the individual shots will be saved as scene files and will have the characters in their starting locations with a camera set up for the desired rendering view. there will be no textures in these projects so as to keep the projects low in size for transfer. the animators will work with the un textured models then delete all everything apart from the character so it can then be imported to the environment scene and textures can be re-assigned to the character. this will hopefully eliminate the need to re-assign the textures to the entire environment each time. this is an issue we have all had over the last 3 years of using maya.

ok so project Scene_01 will have 5 scene files, shot_01 through to shot_05.

test renders

so these are two test renders of the film showing our two main environments. all is good in the hood. the environment above water is very model heavy whilst the underwater is all a matte painting with just the character being 3d.

the town

the assets we will use can be seen in this version of my showreel at 1:05 to 1:10

showreel 2010 from andrew on Vimeo.



so a longer intro was introduced to express the the kids environment with the hope of defining the kids world being a built upon warm rural environment contrasting with the cold emptiness of the water. in effect placing the kid completely out of his element in order to achieve his photograph.

i used an old scene from last year where i had preoduced a pirate town consisting of 2 tudor like buildings with a hillside and crates/barrells. this piece was done as a portfolio booster for last years showreel so we will use this as the base to form the town. the pontoon model that keith has produced will slot into place and the town will be populated with more buildings, a few more built and many duplicates. the plan unfolds and we all get to work.

time and cuts

so after assessing that we were roughly 3 weeks behind schedule i had decided to cut the kraken from the film as it was still not textured or rigged. i estimate that if we were to include the kraken it would take a further 4 weeks to complete due to its level of complexity and the amount of research i would have to do to make the rig. so the film would in effect be cut in half the relationship between the eel and the boy is reduced to predator and prey and the story is more slapstick and "animationy".

i approached my team with this idea and it went down extremely well by producing this alternate story we reduce the shot count from 60 to 40 and we lose the hassle of animating the 5 armed kraken. mel in particular was extremly relieved to not be animating the Kraken.

so decisions made the kraken was cut.

the meetings

as it happens keith had arranged a meeting with clym, myself, and ben keswick to discuss the e-mail as he sensed that things needed to be discussed.
as it happened fahran was in college as well and he joined the meeting

we met and begun discussions

i felt it was rather important as we were able to discuss the issues we resolved problems and focussed on the completion of the kid character. we discussed that it was key for both of us to be updating work and that i was aware of clyms commitments to other projects and that any week that he was un able to work on our film that he send a simple message saying that he was working on another project so we know that time is not being wasted.
so we arranged for the kid to be done by the 18th of april. giving clym around 3 weeks to complete it. everyones happy.

after this we clym, tom rotchie and i met with mike smith to discuss the rigging progress on further films. mike suggested that the time line of my film be cut by half to account for time lost and that we needed to produce a render test of a scene a "beauty shot". agreed job done.

so resolution

i went to college the day after the email was sent and continued work as normal arranging for a group meet on monday,, today being a tuesday.

i was approached by my tutor Dan to discuss the e-mails that clym had been sending around the tutors. we first discussed when problems had arisen , which was at the time when the character had become confused between clym and fahran.

Dan then asked why i hadnt approached him about the confusion and i simpply stated that i didnt think it was a big problem back then.

it was only with the e-mail that i realised what situation we were in.
we then broke down clyms e-mail and took points from it where clym had criticised me, it was apparent from our discussion that most of the points were vague and inaccurate the main issue that arose was keeping people constantly informed. communicating with a large group is tricky but that is the area i need to work on.

so the main points were as such:

level of pressure and anxiety - we discussed that it is acceptable to be stressed when in charge of such a large project.

lack of knowledge of the subject area - it is unclear what subject area he is stating but we decided that my role as director is to ensure the narative is good and that the reason for employing students with specific knowledge base is to ensure that they have the skills and knowledge that i dont. it is unfair to assume that the directors should know how to do absolutely everything.

Lack of positivity to getting tasks done - it was noted here that i should be more encouraging and compliment people more for the hard work that they have put in.

constant changes to the look and feel of the film - the film has changed in feel but it hasnt changed in look or style that has remained constant throughout.

changes to the story - as director i have complete freedom to change the story as and when i wish as long as i am not hindering peoples performance or work. i must ensure that if i change the story it will still utilise characters and set pieces that people have made so as not to waste their work. this was ever present in my mind anyway the penguin that was cut was cut in the design stage so no modelling was lost nothing has been cut that people have already produced. plus i must get the best story i can.

updating my group - it was raised that the group had not seen enough of my work to notify him of my input to the project. this was noted but at the same time a lot of time is devoted to reviewing the work of people that turn up to college and then assigning tasks to them. but i will be better with producing updates.

that my project is the most important- it is to me and that is fair enough i guess.

these were the main overiding points and there are a few in there which i can work on to make me a better director and group leader. my next step is to arrange a meeting with clym to discuss the e-mail and get both our points of view out into the open. i will then have a group discussion to highlight key points raised i will then assess the project and our status.

the e-mail

i was sent this e-mail by clym as he replied to a tutor. he added all of the members of my team aswell.

in plain im not happy. this e-mail i can only regard as a personal attack and as such will not reply, the most worrying thing is the fact that he has added all of my team, by doing this he could directly effect the morale of my team and dissent could be inevitable. there are some points i can agree with but most of it is exageration. the only good thing that can come of this is me personally becoming hardened to criticism. i will consult my group as a whole and discuss this e-mail to discover if feelings are mutual and if they are then i shall work to improve upon them.

but all in all i was disappointed by this message.

[Hi Mike,
Since your previous email I have been in contact with Sophie, Depa, Tom and Kofi regarding rigging for their projects. We discussed communication and they have stated they are completely happy with the level between myself and them regarding my role and the completion of tasks.

I would agree with aranging a group leader discussion however I do not feel it is entirely necessary and would just prolong tasks longer.

I do, however feel, that there is an issue with the project leader of The Deep - Andy. I feel that the level of pressure and anxiety is building up on him, especially due to lack of knowledge of the subject area and processes as mentioned before - which I agree is a learning process however this doesn't mean it is an excuse for poor managing. I personally feel a lack of positivity towards getting tasks done and a sense of constant changes regarding the look, feel and story of the project has caused the whole group to feel lost and unsure. As a group we have definately sent updates and keeping him informed but likewise we too must be informed of what he is doing and how that is coming along which I have to say is far and few between.

I personally feel a lack of encouragement towards the work i have either solely done or contributed to, from Andy and I know that there is a general feeling of this from other group members. He also presumes that we are working soley on his project alone and that there will be "fresh" results every week, when he himself has not produced anything that excites nor stirs us forward. I know for myself I am definately not committed to just this project, and he presumes that I am, that his project is the "most important" of everyones.

I know that a few other members definately feel that he a) constantly changes things b) is not aware of the production process c) presumes much without communicating via email, and this is why I have attached the group members themselves to this message so they can add anything they feel comfortable to.

I hope this answers a few more questions and hope to hear from you soon]

a heck load of confusion

so i have been very frustrated lately with the progress of the kid it has now taken 3 weeks and the progress is slow and garbled. i assigned the job to fahran and clym is taking it upon himself to complete tasks. lack of communication between us as a trio has meant that work that fahran has done, blend shapes, was done on a character that was the wrong scale so the blend shapes need re-doing. miscommunication is the key here but i just feel confused and let down. so much so i actually shouted at clym for his extremely relaxed approach to college, he didnt bring his laptop in to work on and joked about not having anything to do. i should not have snapped but i felt that if i hadnt said anything matters would have continued to pee role along.

so whats the status. the kid is not finished and i dont know who is doing what.

test vid

thar ya go a 5 second clip of an animation that took like 5 minutes this vid was not ever intended to show off my skills as an animator but it is there to show that the rig is usable and some of the functions . pleased much = yes :D

seagull is next

so i decided upon a background character to add populous without doing too much

so the model took about a day in all as it is a very simple smoothed shape the beak and eyes are separate to the mesh so they can be easily weighted.

the rigging took a further 3 days.
it wasnt too challenging as i rigged the character using the same techniques that i have done for simple humanoid rigs before. so the legs are ik and the wings are an attempt at FK which i havnt done before. i believe it works really well and will upload a video shortly showing a demonstration. the weights were the hardest thing to do and as such is where the rig falls down as any extreme movements will cause a lot of tearing. i am very proud of this though as i have wanted to test my skills this year and as such i have produced a usable character.

features include:
reverse foot
fk arms
ik legs
fk spine and neck.

problems - when i skinned and weighted the character i had not textured the character yet. i went to texture the character after the rig was complete and to my horror after UV'ing the character the textures would not stay attached to the character so i proceeded to do the usual delete history which inturn deleted the weights that had taken 2 days to produce. noooooooo.
to get around this i decided not to back track due to time but instead i compromised. i simply selected polygons and created simple block colour shaders with no textures. this will work due to the simplicity of the character. hopefully but lesson learned

UV map the character before you rig as the weights are painted on to the uvs lesson learned.

the kraken modelled

so here is the final model that i have produced using maya and zbrush its great and exactly what i wanted. im elated at what i have managed to produce the only problem is the fact that it is 5million polygans which means it is completely impractical in terms of the film. an error such as this is a great draw back and i will discuss this with my team when we next meet.


fahran was not in this week but clym was so i spoke to him about the situation of the character. we created a list of objectives that needed to be achieved for the character and i rescaled the character to a shape that was more like the character that i envisaged.

after the list was completed i asked clym to give fahran the re-scaled character and also the list of changes to make for the character. clym was more than happy to pass this on.

cool i am greatful that the character is well under way and tasks such as this will push the whole film forward and also spur the team on.

the kid

fahrans first pass of the model was a little bit dissappointing as it had strayed from the concept in terms of shape, size and form.

i had found out that he had sent the model straight to clym to start work on the rig which was good and bad as it will speed up the rigging process but there are still a lot of things that need refining in terms of the model.

Kraken Progress

so i have created the low poly mesh of the kraken, which is now ready to be exported as an .obj to zbrush.

this will be a big test of the model as any innacuracies with the topology during transfer will appear as holes in the mesh in zbrush.

greatfully there were no problems with transfering the kraken across programs and all is well and i am ready to start detailing the model.

lighting and environment concept for the cave.

im really happy with this and i hope that it will be of some use to carlos when he is modelling the cave.


so i have capped off my concepts as im pretty happy with the shape and design of the thing and intend to use the following process to model the beast.

i intend to do a low poly model of the kraken in maya and then export the mesh to zbrush to then add extra detail and make the character better in general, i will then export the altered mesh back to maya to texture and rig.

whats going on

so fahran is now modelling the kid, main character and i am currently modelling the kraken.
rest of the world is slowly ticking by...
mel is doing test animations and keith is modelling the pontoon for the film plus texturing.
carlos is now modelling the interior of the ice cave.
ben is modeling the camera for the kid.