Supervisor Meeting 22/03/11

Today was the last supervisor meeting. We talked considerably about the user evaluation and I showed Richard my mocked up version that I have been working on. Some of the questions I think I can drop. For example, I had a question which asked the user if stop motion complimented the story or would the animation would have been more suited to film or 3D. As Richard rightly pointed out, what relevance does the answer have?

The good news is that I don’t have to do any re-arranging of the Literature Review. I was worried that because I had exceeded the specified word count that I would have to shift a lot into the appendix to fit within the guidelines. So the submission should read: Literature Review, Development Report, Evaluation (my own and then the users’), Appendix. My own evaluation will essentially be a condensed version of this blog, as it will document the process of creating the animation and the conflicts I encountered along the way.

The second piece of good news is that the poster, for the poster session, doesn’t have to be as text-heavy as I had been originally informed. While it must give an outline of the project, it should also have a nice aesthetic to it. I’d love to spend a month on this, but as Richard pointed out (as he often does) that perfectionists can’t control time. That is until we perfect the time control machine..

The End Is Nigh

As a quick test I decided to render out a second of footage to get an idea of how large a file the different codecs produce. I first rendered in Animation at 100% Quality which gave me a .mov file 316.4Mb in size! I Exported that from Quicktime then as 1080p which dramatically reduced it to 2.5Mb. Secondly I exported using the H.264 codec and this gave me a file 21.6Mb in size. That’s much more manageable and I think this’ll be the one I go for.

The morning was spend implementing the mocked up photos of the child and the wedding (I used a stand in photo here as this has not been shot yet) into the animation. The scene where the patient is having a seizure and looking at the pics shaking took a loooong time as each of the 100 frames had to be treated separately, twice. I used the Corner Pin effect as well as the Box Blur effect (to match the blurring in select frames) to fit the images into the animation. The Blue and Green keys I tried to pull were useless as the footage blurred a lot due to the original movement when taking the shots. This is why I had to manually place the images. It took so long that I decided that it might have been quicker if I had taken the photograph myself, printed it and included it in the shoot a few weeks ago.

Using the Corner Pin effect to place the images into the animation

I’m starting to edit the nitty gritty bits of the animation now. This is where the perfectionist in me will want to change every single aspect of the shots. I used After Effects Motion Stabiliser tool to smooth out the opening shot which I used the glidetrack for but ended up quite jerky. I used the tool to analyse the footage as a whole and then restricted the stabilisation to the Y-axis only. This allowed the footage to move perfectly along the X-axis.

Using After Effects' Motion Stabiliser to smooth out the panning shot along the X-axis

Out of curiosity I counted up the shots I’m just using so far in this session (more to come for the wedding shots). Funnily enough the total came to exactly 6,300. I’m not sure if I’ll get around to counting up the total amount of photographs taken overall but I reckon it’s at least 10,000.

This evening I focussed on collecting and building up a library of sound effects that I could work with when creating the soundscape to the animation. As well as using sounds from my own collection, I used the following free sites to complete the effects I needed:

SoundBible
Freesound
PacDV

I will export a rough version of the animation and use ProTools to arrange the sounds and music to fit. I might have to go back and forth to make alterations to the video so I won’t export at the highest quality just yet.

I would like to have the sounds and video ready for user testing no later than Wednesday. This gives me tomorrow to shoot the final scene and begin editing the sound. The bit I’m worried about is the extremely slow rendering out of After Effects so I might have to render it in sections so I can be working on one bit while rendering another.

The Masked Avenger

Things were going swimmingly until I installed a plugin for After Effects which somehow corrupted the program and prevented it from opening. I wasted the whole evening/night uninstalling and reinstalling it a few times and trying to fix the problem. Turns out the Magic Bullet plugin had installed itself incorrectly and was wreaking havoc with the root files within After Effects. I simply found the troublesome folder and Time Machined its ass back to yesterday, where it never existed. A simple solution but only after 6 hours of panic and panic and panic. and panic.

The scene from above the suspended fac e needed some attention. Due to my flimsy tripod balanced on a less than sturdy surface, the face of the machine moved slightly from frame to frame. To fix it I used my new trick of placing a still frame in a layer above the animation and using a mask to let the animated bits be seen. Simples.

Using a mask to hide unwanted movement

I used this technique again in other places. I noticed that in the scenes I had already used it in (the other shots from above) that the table cloth jittered considerably from when I walked past it each frame to move the machine. The light reflected on the IV bag shifted quite a bit too so I essentially froze all these elements in place during this shot. As I’m blending both layers for the animation, this meant I had to import a frame from each set of exposures and apply identical masks to both.

Using multiple masks to freeze elements

I had to stop the numbers at 0000 in a couple of the takes. I had done it by hand during shoot but I had to hold onto the numbers to prevent them from advancing so they moved slightly each frame. When played back the numbers had the jitters..

Numbers frozen at 0(000) degrees

When editing the piece today I had to be ruthless in the cut I made. I’m all to aware of how long certain shots took and the amount of effort I put into each and every frame but I have to start thinking what works for the actual story and how long each shot should be. I’m the first to admit that I shot way too many frames for this, but that’s mostly because I wasn’t sure what I’d need for the putting together of the film at the end. I’m glad I did as now I have choices of different angles and lighting arrangements to choose from. Still, having to cut minutes (which equates to days of shooting) out of the film has undoubtably been the most disheartening activity yet.

 

I also had a bit of fun making the photo of the kid, which triggers memories of a happier time. I just played around with a few images of my own and came up with a cheesy tourist snap which might work well. I’ve been researching motion tracking in order to get these new images to follow the blue and green mattes. I may have a bit of trouble getting a key from the blue piece of card though as the images are fairly blue-heavy in their overall tint. This is something I will have to deal with when and if it becomes an issue.

Saturday, Sat here (all) day

According to author Malcolm Gladwell, it takes 10,000 hours of practice to become an expert in a select field of interest. Going by those figures, I should be an expert stop motion filmmaker roughly by the time I’m finished this project.

Whilst rendering out of Bridge I have decided to get a jump on the sound design part of the project. There are some amazing websites that offer high quality recordings for free and I shall be using them when building up the soundscapes for this animation. I’ll also be recording my own sounds, mainly the tape machine sounds which are quite object-specific and would not be able to locate online.

More importantly, I have loaded all the treated JPGs into the After Effects session.  Now I can begin to edit from within AE and get the timings figured out. I was surprised to see that the session is 41.63GB big! I’ll soon find out if the laptop is up to the job. There will probably be a bit of back and forth between Pro Tools when I begin adding audio as I want the music to sync up to certain sections. This will be fine tuning later on though and for now I will work on getting the rough structure and sfx in place so I can get some user feedback.

By far the most time consuming edit I made today was the second of the ‘above’ shots. As with the first sequence, to get the proper balance of light on both the subject and the surrounding room I shot each frame with two distinctly different exposures. With the first sequence I explained how I imported the brighter set of exposures into After Effects and then a single frame of the darker exposure was placed in a layer above and a mask was used to display the more exposed version of the patient with the darker surroundings.

I did the same procedure here, only as the patient in this sequence is convulsing in the bed each frame was very different to the next. I had to import both full sets of exposures and apply a different mask to each frame as the patient moves slightly from frame to frame. This was quite laborious but in the end I was pleased with the result and the shot won’t look out of place when played alongside the better naturally-exposed shots.

mask placed on the darker exposure to reveal the brighter exposed layer below

due to the movement, a unique mask had to be created for each frame

Result

Business as usual

Nothing to say about today’s work really. Just trucking on with the usual errors/crashes. I have to be careful of what programs I have open whilst Bridge is batch exporting as it has a tendency to crash if there’s too much going on. I’m quite pleased with the look of the piece so far and hope to have all the footage loaded into After Effects by Sunday. Then I’ll shoot the wedding scene and try and have the whole video part edited by wednesday. Then I’ll start on sound and following that, I’ll carry out user evaluations.

I spent some time trying to figure out how to create a .PDF wit editable text fields for the user evaluation form. This way I can email people the link to the video and a simple questionnaire that I’ve made, and not have to direct people to some horrible online survey site. Turns out I can do it with Adobe Acrobat Pro.

Boring yet productive.

Exile on Frame St.

Happy St. Patrick’s Day! The only difference between today and every other day is there is more noise outside the window than usual. I know that this dissertation is more important than a night out and the like but when a global (??) day of celebration and merriment is going on right outside my window and I can’t join in, it is indeed vexing.

Anyway. Today I bulldozed through a lot of the shots (not the alcoholic sort) and have currently fed close to a minute and a half of footage into the After Effects session. I’m getting faster as I’m getting accustomed with using Camera RAW and I know how to quickly edit the shots to get the look I want. Controversially, I have started working with RAW files. I had a niggling doubt in my head that I would rue not using the best quality shots so I made the change. I won’t have time to go back to the other JPGs and swap in the RAW files but hopefully there won’t be a noticeable difference in quality.

Things were going swimmingly until I hit the scene where the EKG monitor increases in speed. When I originally shot the EKG scenes I knew I could edit out the paper overhang in post production using a Photoshop mask or something because the camera would remain in the same place.. What I forgot about was that in this shot the focus pulls from the patient in the foreground to the monitor in the background. Each frame is going to have a slightly different focus so I can’t simply paste a ‘clean’ monitor frame over the live footage to hide the paper. This had to be done on a frame-by-frame basis.

Frame 10/102 - Shallow DOF with patient in focus

Frame 101/102 - Monitor in focus. the frames in between display the monitor with a slightly different focus making it hard to batch edit

Result

spinning plates

Today I split between the User Evaluation, designing the poster and working on the edits. The user evaluation was really just about making up a mock Lorem Ipsum version. I’ve not really settled on any questions in particular but I hope to have it all ready to go by next week.

The poster, like everything else, will be ambitious and over-complicated. I popped into the School of Computing to have a look at previous year’s posters to get an idea of what was required and I was pretty disappointed with the unpleasantness of them. Whilst a good amount of text is required there’s no need to make the whole thing corner-to-corner text boxes. I had an idea that the centre of my poster would display the tape machine and title of the project and be connected to various sections of information via electrical circuit schematics. Maybe I’m going loopers but I decided to trace (in Illustrator) the schematic diagram of the UHER 724-L which I found online. Four hours later I had managed to  draw a third of it. I though I’d stop there as I wasn’t really sure I’d need the lot anyway. It’s nice to step away from the editing and do a bit of design work for a change. Here’s a teaser of how the poster is fitting together..

Working on an idea for the main pic

I'm very very proud of this...

 

In the afternoon I set about working on the ‘reveal’ of the patient. It’s the shot from above the bed looking down on the tape machine. As I only had two lights which I was using to light the room I shot each frame twice with different exposures in the hope that I could combine the brighter version of the patient with the darker room. This was a good idea when shooting but I really should have gone and done a proof of concept. I spent the entire afternoon and night trying to figure out the best way. As I had 206 images of each exposure, it was efficiency I was looking for.

First I tried selecting the same frame from both exposures and used Bridge’s ‘Merge to HDR Pro’ function. The result was okay but not really the look I was going for. Then I though I could simply use a mask in Photoshop on the dark layer to reveal the well lit patient in the layer below. I didn’t have the time to painstakingly import and set up each of the 412 images so I imported the first exposure set of 206 into Photoshop as separate layers. I then set up an action to apply the mask and batch save the result.

This would have been fine but for the fact that my laptop is getting fed up with me and has decided that it’s going to have a rest whenever I ask it to do some heavy lifting (and bear in mind I’m working with JPGs here. I gave up on the notion of using the RAW files days ago).In other words I had an idea well above my computer’s station. I just don’t have the hardware to deal with what I would like to ideally do. I feel like how James Cameron must have felt when first thinking about making Avatar but I don’t have the luxury of waiting until a better computer has been built (and I’m sure it has considering my Macbook is over 3 years old now) so I’m constantly trying to think of the most processor-friendly way of doing these things.

In the end I applied colour correction to both sets of exposures respectively and then I imported the lighter set into After Effects. From here I ran those frames as an animation, but I used a single frame of the darker exposure placed in a layer avove and with a mask punched through so the bright patient exposure could be seen.

I’m confusing myself now. I’ll make a long story short and upload the pictures.

Darker Exposure

 

 

Brighter Exposure

 

Final combination with colour corrections

Using a mask to combine both exposures in After Effects