C-Studio Project IV Process Notes
10/04
I started with a list of phrases that summarizes Gill Sans’s key features that I’d like to highlight in the animation. Items on the list shall be organized in a certain order that allows the video to ‘tell a story’. This means there should be some logic that links these bits of info together, rather than having them scattered around.
From the reading…
- If Contents don’t follow the right>left, top>bottom reading order, viewers might find the message not coherent; at least a consistent ‘kinetic pattern’ should be created so that viewers know where to look. However, too many monochromatic movements could make the animation boring. To what degree shall the patterns be consistent is tricky.
- Good transition aid communication and enhance the message.
- These are more ways to control hierarchy when it comes to videos and each frame kind of their own hierarchy! One major controller is pace. Text fade quicker shall be less important compared to those who get a pause and stayed on screen longer.
- The sound should agree with visuals and aid communication (light text= light and soft music, heavy impactful font=heavy beat).
- Depth in the animation: closer objects move faster and become blurry when they get too close. Further images shall have less color contrast and clarity.
Since Gill sans is designed in the 1920s, I feel the super smooth and modern ‘fade’ effect is doesn’t fit the typeface. And too jumpy transitions such as flip and rotate are also not ideal. I am thinking of using old photographies and videoclips as the background and context provider of font animations. But before getting into AE, I’ll have my script and storyboard ready. Also, I’ll be looking for 1920s delighted jazzy music (better if it has a modern remix version).
Viable music (and their pros & cons)
From CupHead, BGM during The Root Pack boss fight. Jazzy, feel delighted, doesn’t feel monotonous. Although I can find some ‘light’ and ‘heavy’ moments in the music, these moments are not in the same 1minute… Plue, I found it hard to put this melody with photorealistic visuals
Swing dance Jazz; stronger beat and slightly slower compared to the previous one. Again its longer than a minute, so some editing is needed. And these are some speaking in the background, which could potentially cause distraction.
The first draft of the script
The script covered most if not all of the points I wrote at the beginning of the post. I wonder how am I suppose to know the approximate time/duration that these bullet points are going to take? Although I am going to create storyboards for these, the static images still give me no clue of the duration.
Writing the scrip gave me some ideas about the video:
- keep visual simplistic, minimize the usage of the photo (although it is dominant in the spread)
- Probably the electrical swing (2nd choice) is better for these content since there is more slow-down/speed-up built in the music.
- Repetitive phrases: Probably want to put make them occur at the same location of the screen.
The storyboards
An experiment
Since Gill sans is a humanist sans serif, and my script (and also spread from project 3 ) is largely about Eric Gill, I think it would be nice to include an actual photo of him. I decided not to use Illustrations because the style would be drastically different from the spread, and it probably won’t allow me to easily introduce depth to the video.
I tried to use: ①puppet tool: allow me to show body movements of the figure. ②3D object+ liquify: close-up facial expression changes and shows the depth of the scene by shifting camera backward on timeline. ③ vanishing point (vpe): mimics a 3D box; attached a special 1st person feel of the animated photo. But this requires a photo that has obvious perspective. This means I won’t be able to use the photo I used for my spread.
Editing the music
To provide an impactful viewing experience, I should be mindful of altering the fast and slow pace. My chosen music has a pretty nice pace shift, but these parts will not perfectly match up to my content. Overall, In my opinion, it is the content (what I want to say) that drives the pace, not the music.
The music itself has a fast and very rhythmic beat in the beginning, which feels really nice. I do not want to slow the music down, but then the visual needs to be at the same pace so that the two senses can tell the same story. I’ll have to figure out a way to keep the music as it is but slow the visual down.
10/06
I digitized the storyboards and found the ‘zoom one letter in a word and drift others ’ ideal isn’t working so well. Another major is I decided not to show the name of the font until after the autonomy part. Here are my digitized storyboards:
Wondering if I have a suitable amount of content and if the pace works well with music, I compiled these frames into a video. The music still needs some work. And since I only have static images to work with (except two animated -photos showed earlier in this post), some animation moves are hard to test and show. (Some explanations are included in notes in the video) I probably will test-make these more complex clips rather than start making the video in chronological sequence.
My thoughts and a couple of feedbacks from Sherry:
- The pace of the font autonomy part could be slower.
- The structure of the script: I originally plan to reinforce the features of the font at the very end of the video. This leads to a sudden jump from the brand logo to autonomy again. The flow gets interrupted and audiences might be confused. I’ll just remove this part when I move to AE.
- The 3D plane switch for the title might be confusing; the time allowed for viewers to process the info is too short. Plus after this section, I have the animated photos, which, form the hierarchy perspective, should be the most significant part of the video and should be the most catchy part (‘oh this is interesting’ moment). So content right before this section should be on the quiet and stable side.
10/0910
Overall the storyboards transferred animation pretty well. I slowed the overall pace down (especially the letter anatomy part). Here are some feedbacks from Tuesdays’ class:
- The music and color are fine; fits Gill sans. The yellow sometimes I noticed the color (unintentionally) shifted on the big screen. I’ll check the color profile of the app file.
- Another thing I didn’t notice at the computer-screen scale is: some text feels horsy (‘A_____Figure’ part). Just like doing tests print for the spread, it’s better to frequently hook the laptop to the bigger screen to check the video in the size and scale thta it is going to be eventually displayed.
- the pace is readable but still on the fast side. Slow down.
- ‘Kill the daring.’ The video can be impactful with very restricted movements. The goal is effective storytelling and communication and celebrating the font. There are some nice moments in the animated photos, but I might be over-doing it. I’ll figure out a way to focus the message on the typographer and font rather than catchy animations.
- Vicki mentioned some scripts are hard to understand. English is my 2nd language so I didn’t realize these issues, so I asked Mason to check the scripts.
Adjusting the script
- almost a circle → very circular
- geometric overlay → geometric structure
- ’A Sculptor.’ (The capital AS and full stop feels there are too many attention points in a short sentence. As such structure hets repeated for three times)→ A sculptor, calligrapher, and typography.
- A ______ figure. (similar issue and the keyword in the middle (in yellow) feels descending.)→ curious (just the keywords in middle, color change to black).
10/11 A full draft
Changes
- for the opening part, I used to match one bar of rhythm to a letter. In the current version, I matched 2 bars to one letter so the script becomes much more accessible.
- Only kept the most interesting of the three animated photos. So viewers can be drawing t the animation and then pay attention to my message (rather than keep thinking about the animation).
- Made the last part fade slowly, cueing viewers: we are going to end soon.
10/12 Issues in the 2nd draft & Fixes
Reduce bouncing and distortion
I (and the viewers commented) noticed the slot machine effects feel quite dizzy. And it took lots of time to reach the final position. I thought this is because the offset value is too big (rolls too quickly), but later Sherry pointed out it is the bouncing and distortion of the words that made people feel dizzy. As shown in the 2nd gif above, after I reduce the distortion and vibration of texts the clip is more comfortable to look at and read. Although the first one did mimic the slot machine better, here, the purpose of the effect is to communicate (at least shouldn’t prohibit communication).
Logo display
In the 2nd draft, the why I display logo simply has them pop-up one by one and then zoom+ blur them to mimic the impulse/heartbeat effect. Firstly the blur feels unnecessary, hence it should be removed (same for the letter autonomy part). Secondly, although I intended to inform viewers that the video is ending and slowed the pace down, the logos don’t seem to need that much time. Viewers quickly grasp the logos and expect to see something else. They’ll quickly get bored of nothing new comes up. In the newer version, I continued the dot’s morphing motion, trying to reveal the logos in a more interesting way. however, I now have the exactly opposite pace issue, I cant fit all the morphing and display in the time I assigned to logo display. I’ll be trying other ways to display the videos.
Test print → Test render
A software issue: When the computer is lagging, Ae doesn’t give the real-time preview. So the pace became hard to control. Also, some glitches (like the one on right ) become less obvious when played in slower motion. Andrew suggested, in order to get an accurate preview, frequent export the animation with Media Encoder. This is like doing test prints. We can sometimes fail to see the issues in the project files.
Adding Snap sounds
Sherry also mentioned the gap before the title part is hard to notice. If notices, viewers will likely to interperate it as a mistake rather than an international gap. I tried to make the pause time longer, but then the video feels like ‘unintentionally gets stuck’. To tackle this issue, I tried to add snap sound after letter M (the 2nd and 3rd clip in the video; 1st clip is a blank pause).
Overall I think this is the right move. And the sound effect could be applied to more places of the animation where there are hard/chrisp/ fast pace transitions. The sound would prime viewers, foreshadowing a big change (or accompany the changes) and could enhance the beat of the music.
Testing the pace…
I asked my roommate to read the script as I play the animation. She hasn’t seen the video before, so if she can read all the lines as the video plays, all content should be readable. Unfortunately, the ’43 font’ line and the logo part seems to be too fast. Although in the real show, viewers will be able to read to themselves in their mind (which is quicker than reading out loud), I’d still further slow the pace down, especially for the two parts mentioned above. I did the test with some other peers. They also commented the last bit (slow fading) is really nice, allow them to see what they just saw. And the do prefer the concluding sentence I added to this draft. It is this bit that informed them the video is reaching the end.
10/15 3rd draft
- I reversed the figure-ground color of the ’43 font’ slot machine part. Hierarchy, this line is more connected to the following content (which talks about the typeface rather than the typographer). Applying the same color scheme, I tried to tie this line tighter to ‘ideal for body…perfect for display”. Also I personally really like the snap sound I added to this version. It’s making the transitions more lively and crisp.
- The current logo display part still feels quite fast, I tried to ease the dizzy feel by color everything else other than the current logo on display black. Since distractors are muted, people are likely to get the logos in a shorter amount of time. I could slightly slow down the music, but the pitch adjustments would be tricky.
- When rendering out, both the media encoder and Ae shifts the yellow color so the yellow became desaturated. Currently, I tried to enhance the color with the hue panel, but the result is still not ideal.
10/15 Refinements
Spacing/placing: There are few spacing/position issues that need to be fixed. For instance, the big gap between ‘ERIC’ and ‘GILL’, and the O in the letter autonomy part isn’t perfectly centered with the yellow circle, and alos the comma issue above. Viewers might not notice them, but they do have the potential to distract viewers if audiences caught them.
Color: The color shift issue mentioned earlier seems to be caused by the quick time player’s auto color correction rather than Ae’s problem. I compared the grey used in my spread and in the video, and they are quite different.
In the 3rd(last) draft, the grey is darker. And I was concerned such grey won’t give enough contrast to allow viewers to see letters in black. After comparing the colors to the spread, I decided to slightly lighten the gray up, to match the spread and avoid the contrast issue.
I also tweaked the yellow color in a couple of places. For instance, when yellow is fading on a light grey background. Although such an effect is consistent with the word ‘hidden’, the color combination is making the content harder to access. I enhanced the color contrast of the latest version.
Logo choice: Andrew suggested although the Sherlock series is famous, it is a brand under BBC. Plus Vicki and a couple of other audiences don't know the tv series. I replaced it with the logo of the British railway. Although not as famous as the sherlock logo, people who don't know the company can still easily tell that it is an English brand. Hence I can smoothly transit to ‘the British gem’ line.
Issues identified by test - play on the big screen
*Andrew mentioned the glitch might be caused by how the big screen is set up. I added a shape layer to block the glitch, yet the yellow strip still shows up. And we are using a VLC player to show all the videos so the color won't be an issue.
Revisiting the spread
To establish a stronger relationship between the spread and the video, I darken ed the grey of the spread. And I proofread the test again and fixed a couple of typos (‘foreverremembered’, ‘squarecapital’, etc).
Reflection upon workflow
I didn't tweak some issues mentioned earlier (such as the puppet animation) since later in the working process, I directly alter the pace on an exported draft (with time remap). Although really convenient to make pace adjustments, such work flow’s disadvantage becomes obvious when I’m trying to fix a puppet animation or script issue. Due to a problematic workflow, fixing these issues now might take hours :-(. Fortunately, these are not fatal errors that dramatically impact the viewing experience. Next time when I work on a video project I’ll never make edits on exported drafts. if I’m testing different versions, I’ll just version my original aep file.
