![[personal profile]](https://www.dreamwidth.org/img/silk/identity/user.png){kind=small}
thuviaptarth) wrote2021-02-03 01:51 pm
Thousand Eyes, Song Choice, and Pacing
I'm interested in how you pick your vid songs and moods in general, because Thousand Eyes was such a fascinating, and I think brave, choice, because it took such skill and confidence to navigate the pacing. It's one of those vids I love more every time I watch it. I want to know everything about how you made it.
First, sorry for the delay!
Second, thank you so much!
Third: wow, this is even longer than I expected.
This is hard to talk about because most of the decision doesn't take place consciously for me.
I've come to realize that, even though it's usually the lyrics that crystallize the realization for me, that make this song snap into focus as the right vid song, this realization only happens after I've already determined that the sound of the music is right; that the sound fits the mood that I associate with the source, or that I want to evoke in the vid. When I am deliberately looking for a song, I'll put together a playlist of songs or albums with a certain kind of sound and then listen for the right lyrics; sometimes I'll search for songs by keyword, but that's not as successful. But for most of my vids, I didn't deliberately seek out a song/topic; the topic associated itself with the song as I was listening to music as part of the normal course of my day.
With "Thousand Eyes", what made me think "This is a Mei Changsu vid" was the instant association of a line with an image: the line was "It feels like thousand eyes" and the image was the Lin family memorial tablets of the Lin family overlaid with the image of the Chiyan army. Part of the association was the memory of something Hu Ge said in an interview: that Mei Changsu isn't just Lin Shu; he's Lin Shu plus the spirits of the 70,000 Chiyan dead.
(It's pretty unusual for me to have that clear and specific a visual image that early. Most of the time I associate a character or a set of scenes with the lyric and then discover the specific images on the timeline.)
But the line only hit me because the music already sounded like Nirvana in Fire to me: the deceptive quietness, the epic scope, the whirlwind action. The idea of MCS as the storm came after the idea of the beloved dead as the thousand eyes, but the feel of it was implicit in the feel of the music all along.
In a presentation this year, absolutedestiny suggested an app or website that let you overlay source/trailer footage on YouTube with the audio of a separate YouTube music video, to get an idea if the song would work. This fascinates me, because it seems like a great idea, but … I almost never wonder if a song will work. I know when I listen it that either it is right or it is not. ("Right" subjectively, of course--other people have made vids I love to songs that I personally could not have vidded to that source.)
For vidding I also like songs with a lot of internal contrast and variation. About 40% of my vids have been made to songs that, instead of following the usual ballad structure, have a second half or long outro that is dramatically different from the first half. This isn't so much a deliberate choice as … I just really like songs that do that in general, and some of them turn out to be vid songs. But the advantage for a vid is that it makes it easier to maintain viewer interest over the course of the song because the style of the music--and therefore the style of the visuals--changes so much.
(I had to rewatch the vid to talk about this and was delighted to discover that YouTube has fixed the bug that was messing up the audiovisual sync, and I no longer need to cringe when people are linked to that version.)
So … pacing. I'd say it wasn't a brave choice, because I was never worried about the slowness? I had been thinking for a while about cutting more slowly, giving people more time to absorb expressions or feel reactions. I think the standard pace of cuts has accelerated over the time that I've been vidding, irrespective of the pace of songs, and I know I personally have a weakness for crowding in too much information into too short a time, even when the vid would be better served with more time to breathe. And, as with sound, the key to pacing is variation: even very fast songs have slower bits and even very slow songs have faster bits; I tend to drift from vids as much when they're cut too fast for the music as when they're cut too slow.
For "Thousand Eyes" in particular: I wanted to show off the beauty of the cinematography; I wanted the viewer to slow down and look around and like it. But I was also concerned about needing to convey a lot of information to people who I assumed would be unfamiliar with both the source and its visual and cultural conventions, because I did want it to be a recruiter vid. I needed to be sure people would recognize the Emperor as a kind of king or authority figure, even if his throne isn't visually demarcated the way a European throne would be, and even if they didn't know the hat with a bead veil was the equivalent of a crown; I needed people to recognize that the memorial tablets were commemorating the dead. The deliberate pace gave me time to linger on setup shots and/or intercut more symbols that are familiar to a Western audience (kneeling, gravestones).
The music also does a lot to keep the viewer's attention, even in the slower first half. While it's obviously quieter than the second half, the percussion has always felt suspenseful to me, maybe ominous, in an understated way.
I actually never got the pacing for 2:17-3:20 where I wanted it to be; in the final version it's still too slow. The pacing had to be faster than the first half and slower than the climactic bit, and I never quite nailed it. I often reach a point in long vidding days where my sense of pacing gets really messed up and I can't tell what's too fast or too slow, but my time/pacing sense gets reset the next day, after a break. For this vid it felt like I was stuck in that zone for two solid weeks.
That section was by far the hardest for me in terms of both pacing and content decisions; I kept reworking it and reworking and finally gave in and posted what I had, because none of the changes were making me any happier than the previous versions. It is the part that is the most different in the download than the version I showed at VVC--the storytelling is better, but the pacing is worse, I think. The version at VVC also had an additional measure cut, which I probably shouldn't have added back in, but see above re: my besetting weakness of always wanting to stuff more into a vid. In retrospect, I think I should have framed that entire section with Liyang's walk; cut the major confrontations faster; and probably leaned a little more away from the plot and a little more into the character interactions, particularly the slashiness. I'd originally intended to have more MCS/JY slashiness, because I thought that would probably be the most effective at recruiting people, but later on I felt I needed more time spent on plot (and MCS' plotting!) and on making sure the audience could tell the unfamiliar characters apart.
Despite this, people have singled out the pacing of this for praise more than for any other vid I've made, even in the longer version. I assume this is because the pacing of first half and climax-to-end do work well, the latter in a particularly showy way.
I'd love to hear about how other vidders think about song choice and/or pacing, if they feel like posting.
no subject
no subject
no subject
Like you, most of my vid ideas come from music that I'm already listening to (and I've never vidded a song that I discovered for the purposes of vidding), and the idea crystalises around the sound of the song and one or two key lyrics. I feel like most of the experienced vidders I know start with an idea and seek a song at least a good fraction of the time, but this is pretty unusual for me. I guess if I participated in exchanges more I might do this!
YES, I have been getting really into longer clips lately. I think it helps that so much source now is really scenic and beautiful, with interesting internal motion or other things that make it visually dynamic.
no subject
So much source is beautiful or beautifully framed! But also I have been making more of an effort just to hold on faces and expressions, instead of objects or motion. I have always really focused on continuity of motion when vidding, because (paradoxically) it was one of the things that was least intuitive to me when I started. Now I default automatically to clips with a lot of movement and away from faces without more visual contexts, and I think it hampers emotional connection. So I've been trying to counteract that.
no subject
Now I default automatically to clips with a lot of movement and away from faces without more visual contexts, and I think it hampers emotional connection. So I've been trying to counteract that.
Yes -- last year at VidUKon I presented on ohvienna's wonderful Xena vid, The Lost Sky, and one of the main things I learned from studying that vid was just how effectively she vidded small changes in facial expressions, eye movements, etc, to the music, and how much it put me into the interiority of the characters. I remember when I first started vidding one of the things I learned was to show characters doing things together, rather than just showing them looking at each other, and now I think I took that advice much too far in some of my vids. So this has been very much on my mind as well.
no subject
The most important thing for me musically is that it BUILDS and has affect which means my vids tend to be quite long! As a result I always have this fear in the back of my mind that I will lose viewers but for me 4 minutes plus seems to be the sweet spot for really driving the story home.
no subject
no subject
I would also say you don't need to ruin the build if you select the right edits, and/or sometimes the song's build isn't right for the story you want to tell. Both "Light of Day" and "No Skin" are cut down from songs that over six minutes long, and "No Skin" in particular is completely restructured. "Hey Ho" keeps more of the original structure but I swapped sections inside verses/pre-choruses and actually did cut the musical climax of the song, because it was in completely the wrong place for the story I was telling.
no subject
That's so interesting re changing the song struccture - that's something I would dedinitely be keen to play with.
no subject
Also, I have watched your vid a few times and never found a place to comment on it so here's some thoughts just about the vid! The way I see it, I really love how you combine the fact that MCS is the storm that has arrived at Jinling but also he has been carrying the weight and wounds of riding this storm himself for 13+ years. I also love the shots of the faces of the people who love him realizing it's him and how he lets go of the sleeve, letting go of himself as this person they love because he has to concentrate on other things.
no subject
I don't think it's so much that different shows support different pacing, although that may be true, as that different songs do. Especially for shows with more than, say, 16 hours of footage, there's bound to be a lot of different types of scenes and a lot of things you can do. The NIF vid I saw most often when I was looking for NIF vids, for example, is Mayday, which makes it look like nonstop action! And
Completely agree on how beautiful NIF is and how the acting makes you attend to every detail! I have seen very few other shows on its level.
Also, I have watched your vid a few times and never found a place to comment on it so here's some thoughts just about the vid! The way I see it, I really love how you combine the fact that MCS is the storm that has arrived at Jinling but also he has been carrying the weight and wounds of riding this storm himself for 13+ years. I also love the shots of the faces of the people who love him realizing it's him and how he lets go of the sleeve, letting go of himself as this person they love because he has to concentrate on other things.
Thank you! It took me a while to figure out how to handle the opening "Undo this storm" section. At first I thought I should try to convey what the storm was, but there aren't actually a lot of scenes showing the chaos and (sound and pacing) the ones that exist are completely wrong for the music. Then I realized that, since it is in many ways a show about aftermath, I could show MCS feeling the weight of the tragedy rather than experiencing the events. All of the clips are ones where he looks pensive, of course, but also contextually they occur in the aftermath of him remembering or dreaming about the death of his father and the Chiyan Army.
no subject
no subject
no subject
no subject
Off I go to rewatch Thousand Eyes. AGAIN. *g*
no subject
I think that's a super smart choice for vidders, because it forces some kind of structure or thesis beyond, "I like these characters, and will now show you many clips of them."
It's true! Although I am also really fascinated by vidders who are able to use songs with the standard pop ballad structure, and create the narrative movement or variation entirely through the visual storytelling.
no subject
Same here re: sound tending to be a major factor! If there’s a song whose lyrics really fit, but the sound doesn’t I usually end up looking up covers of it.
no subject
Covers seem to be the answer for a lot of people! It never occurred to me, possibly because a lot of the songs I like but can't use are too obscure to get covers.
no subject
When I am deliberately looking for a song, I'll put together a playlist of songs or albums with a certain kind of sound and then listen for the right lyrics; sometimes I'll search for songs by keyword, but that's not as successful.
I agree, the sound of music is very important to me when watching a vid, partly because I don't parse lyrics very well on first exposure (and I can't be sitting there reading the lyrics, there's a vid to watch! *g*), so I'll often just get the feel and a few key lines - but then I rewatch vids like crazy, and usually the ones that grabbed me on the first watch reward me even more when I get familiar with the lyrics, because the vidder understands how to storytell. And vice versa - if there's too much of a tone/genre mismatch, it can kick me out of the experience no matter how apt the lyrics are! (Like, I'm more willing to accept a song that doesn't feel tonally right for the canon when it's comedy, etc. Or when my sister wanted to make a Mirrormask vid to 'It's Only A Paper Moon', I pushed her to find a modern cover, which was faster and therefore harder to edit but way more suited to the quirky feel.)
One of the things I love about Thousand Eyes is how the song/vid reflects on a meta level the buildup of his quest - we only see the last two years on the show, but the slow first section really conveys the weight and patience and grief of those thirteen years, and then the building rhythm gives a sense of the work that went in, and the climax really drives home how this slow quiet man caused a whole country to change. Amazing stuff.
no subject
In a presentation this year, absolutedestiny suggested an app or website that let you overlay source/trailer footage on YouTube with the audio of a separate YouTube music video, to get an idea if the song would work.
I can usually tell if a song sounds like the source just by listening to it, but I use a similar method to this one to get a feel for cutting and pacing: I play the song and jump around on the video timeline to see if the visuals can support the speeds I think I’m going to be working at. This can also be useful if I have multiple song ideas and need to choose between them, particularly because it forces me to pay attention to the lyrics, which I tend to tune out when I’m listening to music recreationally.
… which probably plays a large part in why I’ve had the right song just fall into my lap only a handful of times. More often, I’ve gone to Spotify and listened to an artist or a soundtrack that had the right mood and sound; sometimes the hunt ended there, sometimes it dragged on for days as I bounced from artist to artist or wound up switching genres entirely. Or sometimes, if I was very lucky, I found one song that did exactly half of what I wanted and then I got to repeat this process to find the matching half.
I had been thinking for a while about cutting more slowly, giving people more time to absorb expressions or feel reactions.
I appreciate comments like this, because they remind me that we’re not all coming from the same place. My ability to absorb expressions is not good and my ability to feel reactions is even worse; I often feel that vids are cut too slowly or too focused on faces in place of metaphor/connective tissue, and I tend to think this independently of whether I think the vid works or not. (This isn’t a comment on Thousand Eyes, which I haven’t watched for fear of spoilers.)
Two games I find myself playing non-stop are ‘how much visual information is in this clip and does that justify it being on the screen for this long’ and ‘how long does a familiar viewer need to parse this clip as recognisable and significant vs. how long does a casual viewer need to parse this clip to get a surface-level understanding of it’. If anyone ever figures these things out, I hope they let me know.
I think the standard pace of cuts has accelerated over the time that I've been vidding, irrespective of the pace of songs,
This is interesting, because I got just about the opposite impression from watching last year’s Fanworks premieres, this year’s Festivids and some of the stuff that’s coming out of the Youtube schools. Then again, my memory is bad and my perspective is skewed and I’m one of the worst offenders, so I always like to know how other vidders see these things.
As an exclusively small-fandom vidder, one thing I’ve noticed in myself is the awareness that I have to work harder to grab and hold a viewer’s already tenuous attention. You might think this would have resulted in more accessible and easily digestible recruiter vids, but it seems to have resulted in some gonzo editing choices instead, because I’m interested in experimenting with the form and seeing what I can do with it.
The other thing I’ve noticed is that I get bored of my own work very quickly and wind up staring blankly at the preview window while fine-tuning, and it’s surprisingly difficult to avoid doing this even though I know it’s stupid. I’m always analysing facial expressions or following the motion of whatever’s onscreen or thinking about the significance of what I’m seeing when watching someone’s else vid, no matter how familiar I am with it; I’m never just sitting there like a gymnastics judge, waiting to throw up a 9.2 or whatever at every clip transition. I think I need to learn to take longer breaks.
and I know I personally have a weakness for crowding in too much information into too short a time, even when the vid would be better served with more time to breathe.
Sometimes a vidder has a certain amount of information to convey and the audio just doesn’t respect it; it’s annoying to me that the laziest parts of my vids are often the ones with the best timing, simply because those parts have the least amount of work to do and the least load to bear.
For what it’s worth, I don’t see evidence of overcrowding in your vids, and I know what it looks like because it’s one of my biggest failings. Something to keep in mind is that we all rewatch vids we like quite often, or at least I do, and the cutting appears to slow down (sometimes by magnitudes) every time I do so, because that’s how AV processing often works.
Also, I’m a big fan of vids that aren’t obvious or comprehensible on the first viewing; I like puzzles and and it’s always nice to have something new jump out at me on repeated viewings.
And, as with sound, the key to pacing is variation: even very fast songs have slower bits and even very slow songs have faster bits; I tend to drift from vids as much when they're cut too fast for the music as when they're cut too slow.
Well, that's just like, your opinion, man. ;)
I think vidding or illness or some combination of both has permanently altered my ability to process music at its intended rate. Sure, it’s increased my sensitivity to variegation within songs, but songs which I used to think were slow and simple now sound fast and busy to me — particularly stripped-down, acoustic songs where the mechanics of all the instruments are amplified — and vice versa.
I think of cutting and pacing in general as the writing style of a vid; they have as much work to do as the audio and the imagery, and they have to match the concept the vid is trying to sell. Sometimes that means going against the grain of the music, but only when I think the concept benefits from the tension that causes. I thought about variation in pacing constantly when I was working on my most recent crop of vids; I was already ambivalent about splicing two songs together for Diamond and I thought that the seesawing that was happening even within verses might be off-putting or difficult to adjust to, but the changes just felt inexorable. I seem to have pulled it off though, so that’s all right.