All first and second round compositing is now COMPLETE and is ready to go over to Sushma and Kaitlin for editing. This is also the first time the 2 of them have looked through all the composited shots which is VERY good for one main reason - we get a fresh pair of eyes on the animation as a whole.
As the rest of us have had a hand in compositing we've seen what effects and changes have been made etc, but this was the first time the both of them would be seeing this, and therefore could give some initial first feedback!
As a result we've tweaked a few shots to make sure they stay in line with the rest of them - but also to avoid them looking too after-effecty. Kaitlin made a good point about not hiding behind the compositing. We should be showing of the animations and the work that has gone into the models and textures.
There has been some initial hiccups in getting all shots from After Effects (as they are) into Premier Pro for editing - we're wanting to import the shots this way so Amit has more control in the final round of compositing (which will include grains, vignette and all that snazzy stuff) - but after hunting down Malcolm and stealing his After Effects / Premier Pro knowledge all is once again right with the world.
Sushma stealing Malcolm's mind
One key thing we've all probably learnt from this project is that it is important to find a clever (quick) work around (/cheat) for any visual issues that may arise. Our favourite case study is one again the Troll. Reverting back to a much older model to quickly render off the final couple of shots meant that a re-test of that rig wasn't done, and with textures turned off (due to lagging) a couple of deformations were missed in his lower half.
BEFORE!! We see booty
Luckily this was only really obvious in one shot. Firstly Yannis attempted to block out the issue with some of the surrounding plants on the bridge (which is something we've done in a previous shot for a different issue), but this didn't look right. So we turned to Photoshop!
Bimpe took the effected rendered TIFFs and painted over the problem areas (100 frames.... fun times), and thanks to the magic of copy and paste this shots we're cleaned up in 30 minutes.
After - Booty Free!
Success!
Now all that's left is for Yannis to finish up the credits sequence which is still looking awesome - Cristina has also added a great personal touch #nospoilers
Yep, that's right. we have a week to go until the final deadline and things are still ticking along.
The last couple of weeks we've really been bogged down with rendering - renders taking much longer than expected, renders coming out incorrect or renders not coming out at all (being held captive by a hungry render farm).
Where are the rest of our indirects sob sob
But our spirits have remained high! This is a really important stage and one of the key things to making sure we complete this with some sanity is to stay calm and full of humour (which we definitely are - there was also a game of chess to stretch our creative muscles and assess our next plan of attack).
All the scenes are almost rendered - although due to time constraints we've had to drop the AO passes - and for some scenes where we haven't had time to re-render the diffuse/indirect passes we've had to just use a beauty pass - which sadly gives us a bit less control during the compositing stage - but we're still trying to ensure that it all looks good.
In terms of compositing, we've got a little pipeline going on, so were we're going through 3 key stages:
The initial file set up where all layers are applied
The more snazzy effects like heat haze are then added
Finally after editing, the whole thing will go to Amit for one last round of colour corrections and a couple more snazzy things
Upon reflection there are definitely a few things we would have done differently had we had the time - such as making firm decisions on what would be rendered out (for example where there is no camera movement we only need 1 still shot, not up to 100 extra frames of the same thing :( ). But these are definitely good lessons to take with us for future works.
But on the great plus side, we're still coming up with new and simple ideas to create short cuts for ourselves. For example Ritchie had the great idea of creating 1 generic After Effects file with the heat haze effect to be imported into whatever scenes need it. This definitely saves us a lot of time, especially for those who don't yet know the technique for creating it!
Natasha's still working on pulling together the 2D references for the characters, so we had ourselves a good old fashioned photo session to help with referencing the larger movements of the troll and the mariachi's.
For the troll we were focusing on the scene when he jumps over the bridge. So we put Amit's parkour skills to the test to capture these.
Natasha will flip the images when she's drawing them so they fit the trolls entrance on the storyboard.
Yannis will be animating the mariachi's running away on at the end, so we did a few test sequences with them.
We felt that a couple of the earlier sequences were a bit noisey - there was a bit too much going on which would make the scene look a bit too confused and cluttered, and would be tricker to animate! We would have run the risk of models merging into each other.
Now we think we've pretty much got the sequence we want, its a bit more linear, but gets across what we're trying to achieve - and its funny. Huzzah.
Amit's been working on a couple of tests for our scenes waterfall and river! The river has definitely been easier to achieve - this was made with an animated bump map, he also threw in some test rocks to see what the effect would be.
We're all really happy with the results on this one! Amit's going to make a couple of tweaks (colour wise etc) and play around with the lighting. We'll also test it against the rest of the assets and characters to make sure its not making too much 'noise' and distracting from the story.We're not to worried about where the river currently starts as we'll be adding in a misty effect!
And on to our glorious waterfall.
After a group chat about it we still think theres a bit of work to do on this one - again we want to make sure it doesn't distract and we need make sure the water flows in a way that it looks like it's coming from a distance (so the splashes shouldn't be to prominent). Yannis also pointed out that we need to tweak the edges of the landscape to reflect this aswell and create more of a perspective.
Also Amit made this using Maya 2015's Bifrost, so we can't really use this as we're working in 2014. There was the suggestion of rendering it off and adding it to our scene but there were scaling issues when Amit first tested this, and mountain and waterfall would end up tiny tinytiny...Touche Maya touche
All in all it's definitely a really good start, and getting an early look at these tests mean we can start making the relevant changes to fit the rest of the scene.
Today's eye candy is a rough comp that Amit's pulled together of a quick render of our landscape.
The character modellers (Amit, Yannis and Ritchie) have been having a little get together to do some clean up on the models. This also involved a couple of eye tests.The troll and mariachis all have circular eyes and we decided we wanted to do the same for the goats for style consistency. So the guys have been working on the best way to go about this so it fits with the model. They tested different eye styles (flat circle, spherical etc) along with different textures. For the flat dot they sharpened the spotlight and added the reflection values. They also tried to add a transparent eye skin on to make it look more artificial.
They also tested out giving the goats a socketed eye, which all though hilarious, was pretty much very creepy and strayed a bit from our visual direction!
So we've reached our first milestone for our Going Live project and it's all really starting to come together.
Pre-production is almost complete with the exception of rigging and final textures, which we'll be working on over the next couple of weeks.
We'll also be reviewing all of the models and assets to make sure we have style consistency and can make any tweaks if needed.
There's been a lot of work done over the last month and we thought it would be important to log this milestone (and survival) and showcase everyone's contributions up until this point.
Also our famous last words before we descend into animating chaos.
Kaitlin
I’m working as project manager/director. During initial brainstorming, I worked
on a storyboard and took the concept home to write up a script.
I did initial research and mood boards on Mexican art and plants, as well as low poly style that has since been updated.
We chose two plants to model from the Southwest that I created in Maya. I did some initial concept art on the landscape and mariachis.
I’ve been working with Yannis to get a band lined up. I have a local connection and hopefully we’ll be able to connect with a musician this coming week.
I also worked with Amit to pull together the production schedule for the team to make sure we were completing all tasks and hitting deadlines. I've also done some research into naming conventions and best methods of practice for streamlining our pipeline, by making sure everything is being set up to be rigged and textured at the same time.
Amit
Due to previous experience in this, I initially worked together with Kaitlin to help pull together the production schedule and production pipeline for the team.
For the production side of things I was in charge of modelling, texturing and rigging the mariachi band.
For the texturing side I used mudbox because of its capabilities of connecting with Maya - that way I was able to jump back and forth between the two programmes at a click of the button. I felt this was the best workflow because it allowed me to see what the textures looked like in Maya and to quickly make changes if I needed to.
The whole team is steaming ahead nicely and we are even ahead of schedule. I have finished texturing two out of three mariachi band, and since they are quite stylized and simple I am getting through them quite quickly. I should have the final one finished for the beginning of this week.
That means I will have finished the texturing a week early which will give me plenty of time for the rigging. Below are some images of the tall mariachi in the texturing stage.
I also pulled together our 3D previs based on the storyboard
Bimpe
Going Live has taken over my life and I've decided that I don't mind one bit. Yet.
Soooo i'm working as secretary/assistant for the team, and first and foremost i'm maintaining this glorious blog on a regular basis to track our progress and make sure we're not missing anything. I'm also forwarding all notes and action points from meetings to the rest of the team. Huzzah.
In the early stages I worked as part of the art team and directed what areas needed to be addressed/designed and made sure that our 'stylistic vision' was coming across on paper.
The project as a whole is very stylised which allowed me to break out of my comfort zone and work on some heavily stylised character designs. I worked on a first draft for the ogre and the mariachis, which Yannis tweaked for the final design.
I was keen to work on the storyboard as I want to expand my experience in that area, and after some really useful feedback from the team (and Phil/Barry), I worked on a second draft that really focused on the camera angles AND incorporated our final character and environment designs.
Natasha and I worked together on the final tweaks to pull together for the animatic.
I wanted to do a bit of modelling because I feel really comfortable with it now, but also felt like I wanted to try something different so opted for environment and asset designs - things I haven't done yet. So i've been working on the bridge, the terrain and the bougainvillea.
I wasn't particularly happy with my first attempt at modelling the bridge and terrain, as i'd managed to make them overly complicated, so I was more than happy to scrap them and start again, this time with those issues in mind I was able to pull it together a lot quicker.
Also I've created quite a few bougainvillea which could slow down our render time, so these may need to become 3D looking objects on a 2D plane...
Although i've keep the models simple to fit in with our style, i've also realised that the bridge may need resizing at times when animating due to the angle of our shots (something Amit also realised when he created the pre vis).
I'm going to be working with Yannis, Sushma and Natatsha on the sound design due to my work on the storyboard/animatic and previous experience! And rigging the Mariachis because oddly enough i've really grown to love the process.
Cristina I had worked on the initial designs for the goats. I drew inspiration from pictures of goats and piñatas to try and determine a style.
After group feedback I decided to use early designs from Natatsha and Ritchie's design of the big goat as references as there were some great details that really caught my eye, and I wanted to try and pull all these ideas together.
This eventually became our final style design!
I also wanted to challenge myself and work on areas that I haven't had much experience in, so I put myself forward for the environment design and matte painting.
I looked at a lot of pictures for reference, and like the goats I picked out elements that really caught my eye - like the waterfall. I also took on board aspects that other team members had talked about in regards to the character design. Natatsha had been looking at Mayan stones when it came to designing the ogre, and I thought it would be a good idea to incorporate this into the environment.
I'd also been working on some initial concept art for the story
I've had lots of really useful tips to help with the environment design process! For instance, initially painting everything in black, grey and white to help set up the levels and shadows. This was especially useful when it came to changing the colour palette - meaning that I only needed to drop in the colours without really starting again / making drastic changes.
Over the weekend I worked on the matte painting in Photoshop. I've had some great advice from the team, and Natasha really helped in terms of establishing more depth and perspective.
My first 2 versions
Current version
I also wanted to do a little bit of modelling so i've been modelling the instruments for the mariachis. Luckily due to their simplicity, I haven't run into any trouble modelling them!
Natasha
Initially I worked in the art team on character design and concept. I did quite a bit of research and pulled together designs for the ogre/troll. We're really keen for our low-poly aesthetic to stand out from the others so I looked into Mayan patterns and influences for the design.
I tried out different colour schemes which I felt really helped, and after some group feedback we picked some of the elements we liked best from these designs and came up with the final design one. I also did a bit of work with Yannis for some final tweaks.
I wanted to work on my modelling skills so put myself forward for asset and environment modelling. Over the last couple of weeks i've been working on the sweets, stones and mountain.
It took me a couple of tries to get a 'low poly' look with the sweets. I started with the gummy bear as it would be the most complicated to model and achieve the right balance of rounded gummy and low poly. I rounded the edges and scaled a couple of the features until the team decided it looked right. I then moved onto the rest of the sweets using the gummy as a threshold.
I kept the shapes as simple as I could and tried to stay away from using spheres. I had to change the orange lolly though as the square egg and condom wrapper analogies were made. I decided it best to stick with round lollies from then on though using cubes to make them. Yannis suggested I resize certain areas, especially the lolly pop sticks, thus making them more stylised and able to stand out at a distance in the pile of sweets.
For the stones, I referenced drawings from Cristina and her environment deisgn, First I tried to model the shapes into the face of some of the rocks, I felt this would create a good look and cast subtle shadows with the lighting. The results were okay but it took a long while to achieve a carved rock look with just modelling and the texturing will have to do the rest of the work. So I then tried with mud box. The team's comments seemed mixed but I agreed that the rocks looked too smooth and were not low poly like at all. Also I was not happy with the carvings as they were not as distinct and seemed ugly.
So I modelled simpler and more solid shaped rocks and I will try to see if the look can be achieved with texturing and a little modelling. Also we all decided it best for stones in the distance to be left up to matte painting.
With the mountain face I used the concept art as a base to start modelling a rough shape. Though I felt the need to continue carving with a mix of the interactive edge and edge loop tool, to try and get a worn away look for the place where the river and waterfall will go. It still needs some work and tidying up but the team seemed positive about it. So once every thing is approved I shall delete unneeded faces ready for it to UV mapped and textured.
I also give Cristina a few tips for the matte painting to give her piece a bit more depth and scope.
Ritchie
I worked on the early concepts and design of the goat for our short film. Due to the style being low poly, my early designs were very simple and stylised, and I though quite early on about the ways the goats could be animated.
Once we had the final designs from Cristina I started working on potential textures and colours schemes for the goats that would fit in with our environment, but also stand out. I also started work on modelling the 3 goats.
The goats were designed so that they would all stem from one main model which could be duplicated, scaled and edited to create the next goat - cutting down our overall modelling time and giving the goats consistency as a unit.
Working on these I realised that our low poly models would actually be quite high poly - as I needed to add in more edge loops in order to get the boxy look we wanted.
I have now finished modelling and am making the textures. I completed UV mapping before the weekend, and have been testing ways to draw the texture. I firstly tried to draw it with brush tool in Mudbox, but didn't feel this was accurate enough. So I decided to try Photoshop to draw the texture on the UV map.
Although the quality is better, some of the patterns aren't aligning correctly with each other the other side. I plan to fix this by using the projection tool in Mudbox - although I need to do more research about it. I have tried once but the quality was quite blurred!
Sushma
My main skill set lies in the animation process, and for this production I really wanted to expand on my research skills. So during the pre-production process I have been researching and gathering as many reference as possible to understand the movements of goats and ogre.
Although I have animated a quad character before, definitely more of a challenge as these quads are piñatas and very origami in style. Understanding the basic body movements of a character is essential, which makes animating a four legged creature quite tricky.
To further understand my research i've been reading Animal Locomotion by, Eadweard Muybridge, and it has definitely helped me understand a more about animal gaits.
For Ogre and mariachis I will be shooting video reference to help us understand the movements and body language when it comes to animating them.
From next week I will be leading the animation process and covering the basics with the rest of the team who are new to animating.
I have also be assisting Yannis with working out and acquiring our sound assets. Once the script had been complete we were able to assess what sounds would be needed and at what points we would require narration.
Yannis
For this project so far I have been focusing on character concept art and design, modelling the troll and sound design.
For the mariachi's I took Bimpe's initial design and made some changes to these. I wanted to give each of them a distinctive personality visually, so there's the charming one, the happy one and the weird one. I felt this would help them stand out more and really add to the stylisation of our animation.
I also worked on some troll designs, although these were not used there were elements that Natatsha was able to incorporate into the final design.
I have been working on modelling the troll which has been quite an interesting challenge. I feel like I have really pushed myself with this, especially in terms of keeping it stylised and creating the low-poly look.
I got some good advice from some of my team members on certain aesthtic aspects like the teeth and eyes, but also the modelling process itself and the clean up.
This week i'll be meeting with a musician to get started on on our animations score. We already have a rough idea of the sound we're going for, and now that the animatic / pre vis have been completed we have a better understanding of our timing and the mood we need to create.
Ready for phase 2, where tequila shall be consumed by all!
