Collage Redux: A Discussion
This is going to be much less of a tutorial and more of a discussion of how my mind is working these days with relation to composing your standard "images blended together" episode/scene collage. You may feel free to totally disagree with my aesthetic and subjective judgements...or (preferred! heh, heh) to be swayed what will surely be some brilliant insights. ahem. Since I won't be getting into much technical detail, this should be useful for users of any graphic editing program, including PSP, etc or at least worth a perusal for anyone who's interested in reading about the sort of things I consider when throwing images together. Or in pointing and laughing at some really hideous early work of mine. Oh, and just a caveat: my last bit of formal training was a really basic drawing class in high school, so my terminology is probably all wrong. Hopefully I won't embarrass myself too badly (well, other than with the old art, of course!)... Well, let's get to it, shall we?
9.1 THE HORROR! THE HORROR!
The starting point for this discussion is a collage I would like to redo (season 3 episode, Dead Man's Party). I'd like to say it was one of my very first, but sadly, no, I'd actually been collaging for a good 3 months by that point. And still, I produced the following abomination:
Lo, it sucketh mightily... I'd very much like to completely disown it. But in the interest of growing as a person, facing my inner demons, spanking those pesky inner moppets and what not, it occurred to me that this is my golden opportunity to give a brutal critique without worrying about hurting anyone's feelings... also to expound on some of the things I've learned in the 1 1/2 years since it was made, in the hopes that I can speed the trip from point A to point B for someone else, and that writing things out will in fact crystallize the lessons learned for myself. 'cause I'll be the first to admit that I still have a long way to go and a lot to learn!
You might want to click on the image above to open it up in a new window so you can keep it visible (painful as that may be) while you're reading the comments below.
Before I tear into it, let's take a moment to talk about the pretty much the only thing this piece does right: it actually has a concept. Here's what I wrote in my original notes:
When Buffy and Joyce discuss how to handle Dawn in "Blood Ties" and Buffy says "I wouldn't want my mother and my sister coming at me from all sides" I was reminded of this Season 3 episode, where she dealt with that sort of thing first hand. And so for this collage, I thought I'd put Buffy in the middle and have the whole gang surrounding her, all focused on her, and not in a happy way.
Having a concept is to my thinking pretty much essential for anything but an experimental piece you're practicing a new technique on or a fluffy desktop wallpaper to show off your favorite promo pic (and I'd argue that the latter would be vastly improved by having one.) What do I mean by concept? I mean some sort of idea that governs what images you select, where you put them, what colors/effects you use, etc. An image's concept is the creator's answer to questions like: what are you trying to show/illustrate? what's your point? why should I be interested in these people? A successful image makes the answers to these questions obvious to its viewers without off-stage prompting from its creator.
Of course, it doesn't have to be a particularly deep concept. "I like this scene" is perfectly acceptable. However "this scene makes me feel [insert emotion here]" is better, just like "I think this character is cool because [reason here]" is better than just generic "I think this character is cool", and "this is the kind of relationship this couple has/should have" is stronger than the perfectly acceptable "I support this couple". The more specific you can be about what you feel about the images you're using, the better. There's a certain essay-like-quality to a good image concept; like an essay, it should make a convincing argument about what the creator feels to be true or wants to be true. Of course, since we're working with images everything's much more subjective, and different people may take away a different message from your work, just like they may take away a different message from watching the show itself than you do.
Okay, so back to the original DMP collage. Concept -- Buffy's return does not go smoothly; everyone comes at her from all sides -- not bad. Matches the main story of the episode, which is what you're going for in an episode collage. Let's get right to all the myriad things that went wrong...
CUT OFF FOREHEADS
Working with caps with cut-off foreheads is difficult. There are some ways to get around the limitations. For beginners, the main one is to place the image along a well defined edge, either of the top of a canvas, or a floating box or line, so it looks like you *meant* to cut it off there. As you get better, you may be able to get around a cut-off forehead by blending diagonally/ovally and juxtaposing with other images/background so that again, it looks like you intended to have that part hidden. But I was nowhere near that level at the time, so I should have just not used them. That straight line above Buffy, Willow, Joyce & Gile's hairline? Hideous; glaringly hideous. And the little attempt at blending it in (soft edged, rather than straight edges like the cap) helps not a jot. It's still painfully obvious that the reason you don't see the tops of the heads is that my image didn't contain more information, and that's the secret that you have to closely guard when you're using cut-forehead caps.
REPEATING FACES
Why are there 2 closeups of Xander? 2 Willows? 2 Giles? "Because I like both of the caps of each" is not a sufficient reason. There *are* good reasons for including multiple images of a single character, but lets go back to our concept. What about the concept suggests we should have multiple exemplars of a single person? nothing. If anything, it suggests their being a crowd of *different* people, each ready to lay into Buffy. Okay, so that's out as a good reason. Aesthetically, is there anything? If I'd been using floating boxes or any other sort of artificial divisions, that gives an easy aesthetic reason to include multiple single-character images, snap-shot style. But no, this is just a straight blend of images, so there's no aesthetic motivation for it either.
Now lets look at the specific ones I've chosen. The "Challenging Xander flanked by Cordelia" does fit into the idea that Buffy is being challenged from all directions. But the profile shot? included because I thought Xander looked cool in it. Xander looking cool is not part of my concept, so that one should have been out.
Now, the two Willows. Those two images are way too similar to both be in the same general blending sort of piece like this one. They convey more or less the same feeling of Willow-upset, they're both from similar angles; they each compete with one another for attention, thus diluting the impact of either one. One or the other should have been selected, preferably the one with a forehead if only because it *does* have a forehead.
The two Giles is a little less of a clear-cut case, because the two images are very different from each other, and could both, aesthetically, work in a single image. I must admit, I really like the Giles-takes-a-private-moment-to-deal-with-relief-at-Buffy's-return one. But again, does this fit into my concept? Alas not really. Then again, the more stern Giles suggests a level of stridency that I don't recall Giles exhibiting towards Buffy. He was just happy she was back, and I suspect the guilt engendered by that little dressing down Joyce gave him in the previous episode kept him from getting wrapped up in being upset with Buffy. The cap fits the concept, but doesn't really express what went on in the episode. So I really shouldn't have included it either. A separate "Giles/Buffy" collage, expressing his concealed but strong feelings of relief/guilt would have been more appropriate.
FOCAL POINT ISSUES
I did make an attempt at a focal point here -- Buffy -- but it is it enough? Focal point, like balance, is one of those things that are much easier to get a feel for if you zoom way out. So looking at the small version below, where are you eyes instantly drawn when you first view it? Buffy, yes? okay, not bad... but then where? Step 1 of a good focal point is to grab your attention before any other element of the collage, but it also has to nudge you as to where to go next, and that's where this focal point falls apart. There are simply too many disembodied faces bobbing around like haphazard balloons.
If you're having trouble with focal points, or rhythm/balance, try shrinking the image way down. Sometimes this can really help problems jump out at you at a point where you can still fix them. For example, here's the original collage, way shrunk down:
It should be much easier to spot the lack of coherent game plan in the arrangement of the faces than in the larger version up top, since your gaze is not likely to get tangled up in small details, and you can really take in the whole thing with a single glance. Stepping way away from the monitor will also help. Sometimes I just put a work in progress on my desktop, so that as I'm entering and leaving the room for the next however long it takes, I'll catch glances of it, and this can help me spot problems and come up with solutions.
SKIN TONE ISSUES
Contributing to the "where do we go from here" focal point issue is the lack of consistent skin tone. Willow is so orange that the larger of the two Willow images threatens to overwhelm the central Buffy image. Joyce is so dark that it takes a moment to recognize her, especially right next to the blinding paleness of Giles. The other Giles is both over-dark and over-orange. All the competing colors add to the haphazard, thrown together, and let's face it "rank amateur" look.
Obviously not all people have the same skin tones. But there's a certain "different skin tones under the same light" look that I should have been striving for here, but which I ignored. That's why it's easiest to work with caps from scenes that were lit in the same way. For example, the outdoor shot of Buffy mixes poorly with the indoor shots of the others. Selecting a shot of Buffy that was *also* indoors, in the same medium room light as the other characters, would have meant not having to worry about trying to match her skin tones to the others.
GAP FILLING/DENSITY/VARIETY ISSUES
This thing is way too dense... in fact, I think I remember looking at that spot under Buffy's chin and saying, oh, I need something there and cramming that little shot from Gile's door in there. Which makes *no* sense logically speaking (doesn't fit in with my concept) and looks ridiculous. Gaps shouldn't be filled by cramming more faces into them. They need either full-body/distant shots or bits of texture.
And in general, all faces? sucks. I can think of counter-examples, but I would go so far as to say the general rule is: don't do it. Certainly don't do it like I did it here, with random faces bobbing around in nothingness, with just enough space between them to make them stand out from one another, but not enough to create any sort of interesting rhythm or relationship between them.
This piece would have been *vastly* improved by varying distant, mid and close-range shots, instead of having all these close-ups competing with each other. Think of it as a mathematical equation. A viewer cannot devote more than a 100% of their attention to your piece. So the more times you divide up that attention into separate elements, the smaller a piece of the pie each individual element will get. Tons of random caps packed densely together will dilute the impact of any individual one.
Of course, some people do some very wonderful work with intricate images that give you tons to look at and inspire many minutes of intense gazing. As a beginner, I should have dealt with the fact that that wasn't going to be the case with *my* chief-d'oeuvre-in-the-making, and increased my chances of making an interesting arrangement by use of 2-5 caps instead of 10.
COLORING ISSUES
Coloring tweaks should enhance the mood, not emphasize the flaws. Of course, as you may have noticed, I'm fond of bright highlights and mid-high contrast... but in this case, I should have toned it down some, because the cap quality did not support that sort of treatment, and brightening things up only emphasized the mottled "this is a crappy cap" look. Which brings us to:
CAP QUALITY ISSUES
The caps I used sucked. Analogue via-attenna broadcast, capped from VHS. There's only so much that can be done with that much noise/grain, so in most cases, no point in fighting a losing battle. I should have gone and found some better caps. Barring that, shrinking down to a really small size (like the small version above) and not taking the colors too far from where they originally were at will help disguise the craptasticness of the caps. Or going the opposite route, going with the flow, and introducing even more noise and grunge in a "I meant to do it this way" sort of approach.
9.2 SELF CHALLENGE
Okay, for further discussion/exploration, I've set myself a challenge: "using the same images (in their original poor quality) try to remake this collage along the same lines, but better, using no skills you didn't have then." Tough one. 'cause really, my first instinct is to scrap all those close-ups and add some variety. But let's see what can be done with the same raw materials, the same beginner photoshop skills, but a new perspective.
First, I think the Buffy/Joyce confrontation is probably the most important of the confrontations that takes place during the "everyone jump on Buffy" attack. So I start with those caps:
Note that I put Buffy up so that her head is along the upper edge. She's going to be the focal point, so it's more important that her lack of forehead be disguised than Joyce's. Meanwhile, I've scaled Joyce down a bit from the original, so that her face is approximately the same size as Buffy's, and lightened her a bit to match Buffy's in the hopes that they'll be point-counter-point. The brightness of the Buffy-background also does a bit to distract the eye from noting that the Joyce cap cuts off prematurely along the top edge.
Now, I go in filling in the other characters, this time limiting myself to *one* of each. I'm including Giles despite my reservations about it since he was part of the original...
Still quite flawed, but quite a bit better I think, and a reasonable effort given the self-imposed constraints... Here's the original layout, before coloring, for comparison:
Notice how in the new version, I've desaturated Willow and broken the Xander/Cordy cap into two pieces so that I could align them separately. Notice also that when your eye hits Buffy, it can slide down diagonally to the lower left corner, then loop up and around to settle on Giles and slip back across to the Willow/Oz corner. Here it is with a little color retouching... not as much as I'd like, but the cap flaws where too noticeable at brighter adjustments:
Better, yes? Just for fun, here's a retake on the issue, only this time, I allowed myself full range to use any of the caps in the original collection I'd taken when prepping for the collage way back when, and to exclude any that weren't working.
eh, not so good, since that collection exclusively contained more of the same problematic close-ups. Which bring me to:
Conclusion: you can seriously screw yourself in the "picking caps" stage. Mayhap you'll find some that you love individually, but that just don't work together. Or you have problems convincing yourself that your piece doesn't have to have *every* cap you love in it. Or maybe the ones you want have cut foreheads. Or maybe you're attracted to closeups (like I am) and you let that blind you to the need for variety (like I did for *way* too long, cringe, cringe, cringe). Or maybe the quality just isn't there to do what you want to do.
9.3 ANOTHER SELF CHALLENGE
Okay, this time I give myself free reign to use whatever shiny new caps I want to, but I have to stick to illustrating the original "everybody coming at Buffy" concept.
Again, I start with Buffy/Joyce since that little 3-ring circus yelling match ("I've had schnapps!" heh, heh) is pretty much the core of the conflict.
Note that this time I've gone with half-body shots of each, rather than close-ups, and that I've been careful to select caps that include the top of heads. Buffy is more important, so I make sure she's larger, but I position the two caps so that they're clearly in relation to each other, with Buffy and Joyce facing it off as it were. Since I'm going for conflict, I make sure that Joyce looks combative (rather than just unhappy and mildly ticked off, like my original choice) and that Buffy looks besieged (rather than just sorta sad and confused, like the original). And I leave bits and pieces of the background in to make the blending go more smoothly later on.
I put the duo over on the right side of the canvas because Buffy, as the primary character, is facing left, and I want her field of vision to encompass the entire piece. Because this is conflict, she needs to come face-to-face with the other players in the story. If I was showing Buffy-in-Denial, or Sad-Buffy, I'd want her to face *away* from everyone, avoiding eye-contact, retreating into herself, etc.
Next, I fill in the rest of my primary players... note that I'm pointedly excluding Giles. He's not in the scene (since he's off puttering desperately in his little old car, cracking me up with his "Do you like my mask? isn't it pretty? It raises the dead. Americans!") and he didn't take a confrontational attitude with Buffy vis-a-vis her return, so he has no place here.
So my Willow/Xander heads are back. But this time they don't just float near each other, they're in relationship with each other. Xander is positioned almost as if he's standing behind Willow's shoulder backing her up; and they're both paired up (ie ganging up) in relation to Buffy, who stands alone in the opposite corner. To accomplish this, I had to flop Willow over, as she was originally looking to the left. I've very cautious about doing this with closeups, and refuse point blank if the image is a central one. Most people just don't look quite the same in the mirror as they do face-on... it's subtle, but I find it distracting and avoid it as much as possible. But I couldn't find any even mildly suitable Willow cap from this scene looking right, so I gritted my teeth and did it.
Then by adding the distant shot of Xander and Cordy (which is from the background of the Joyce cap, I just separated it out so I could move it independently) I now have a continuous curve starting at Willow's head and looping sinuously over to Buffy. The mix of close-ups and more distant shots emphasizes the curve, which makes it seem more planned and much less haphazard than the layout of the first version, in which size was governed entirely by how much space I had for a particular head. (incidently, it wasn't planned, I just moved the caps around until I saw something that *looked* like it might have been planned.) And since Willow & Xander aren't as important as Buffy, but their heads are bigger, I reduce the opacity on each of them so that they don't compete quite as strongly with her.
Okay, getting there. If I was struck by the mood for a borderless collage, I could stop there, add a bit of texture to ease the transition from cap-to-black, maybe add some text down in the lower left and call it day. But all my other episode collages for the season are packed to the edges, so I'm going to keep that trend going here.
So I fill in a nice distant illustration shot that shows Joyce & Willow confronting Buffy-on-the-verge-of-leaving-again. And some zombie hijinks over the upper right corner. And a bit of wallpaper behind Willow's head to help smooth the transition between the lighter background behind her hair.
Now I've picked up the habit of cementing the blend of adjacent caps by putting mostly-translucent bits of other caps, particularly ones with lots of set detail, over the joins and masking out bits that annoy me. It's basically the same technique as I describe here using with stock images rather than caps... Here's the result of doing that here:
Finally, I tinker with colors, levels, filters, etc to add the zing.
Not brilliant by any means, and who knows, maybe in another year and a half I'll be cringing over *this* one, but still I think it's pretty safe to say it's substantially more polished -- not to mention more evocative of the confrontation that's the core of Dead Man's Party -- than the original.
Click to the left to view the full sized version of the collage.
9.4 COMPARISON
Now that we've got 2 images here, let's compare them for a final bit of discussion...
DIRECTION OF GAZE
In the first, I made an attempt to get all the faces facing toward Buffy. It's a step in the right direction, but just facing a head in a certain direction doesn't automatically set up a relationship between it and the focal point. Note for example that Buffy herself is looking straight out of the canvas at the viewer, not at any of the people who are confronting her. "Challenging Xander" is really facing more towards "Profile Xander" than towards Buffy. Joyce's gaze misses Buffy completely and hits "Profile Xander". The larger Willow's gaze is not bad, but the smaller one is off in la-la-land. The open-eyed Giles is looking out at the viewer, and of course "Profile Giles" isn't looking at anything. All of this contributes to a scattered effect and falls short of the goal of having everyone focussed on Buffy.
In the second direction of gaze is a lot better, though of course not perfect. It's pretty hard to get perfect, these days I settle for pretty good. Buffy is looking towards Joyce, but her gaze encompasses the distant Willow & Xander as well. Willow & Xander are both focussed on her. And of course the gazes within the bedroom scene between Buffy, Joyce & Willow are accurate since they are all in the same cap. Note how much more pleasing to the eye it is to have the gazes sorted out...
SHAPES
When you start piecing different elements together, your eye likes to make simple shapes out of them. The fact that it's difficult to do that with the first one is a key indication that there's something big-time wrong with the arrangement. I've tried to sketch some out in the center thumbnail above, but they're not strongly suggested shapes.
In the second however, we get some nice triangles (my favorite shape!)
FACE PLOTTING
If we just replace the faces in the first one with dots sized according to the size of the face, not what a random arrangement we get. No focal point, and not even remotely pleasing or interesting to the eye. About the only good thing we can say about it is that at least I didn't have them stacked in vertical columns so that people's foreheads are growing necks/chins of the images above them (ick, ick, ick, don't do it!), but rather staggered them.
Meanwhile, in the 2nd, note how we get that interesting little wave from Willow to Buffy. Looking at the plot points for the bedroom cap, I'm starting to think maybe I should have flipped that horizontally; that would fit better in with the top curve, creating a little inverted curve that locks in with it. I could go back and do that now, but I think I'll leave it in as an example...
9.5 MORE QUICK REDUXES
I couldn't very well redo the Dead Man's Party collage and leave the Anne, the first one of the season, in its original, uh, glory with all the rest of the season shiny and new. And so I figured I might as well throw in a little light discussion of that here as well. Here's the original, also from October 2001.
At first glance, not quite as bad as the DMP collage, mainly because the similar hues and faces create less of a scattered effect, and the cut-off foreheads are a little less glaring. Also, the coloring plays to the caps strengths (ie, dingy and gray, much like the caps originally were) rather than accentuating the flaws. Concept on this one however -- which runs something along the lines of "there's this girl. she's blond. and also Buffy. who is also blond. they're both in this episode. and they feel stuff. about stuff." -- is pretty damn weak. From the caps I've selected, all it's possible to tell about Lily is that she's kinda sad. Buffy meanwhile, is a little strident. But about what? And do these two characters actually interact? There's no indication of it here. I've made a really lame attempt at an artificial focal point by brightening the central Buffy face in relation to all the others, in the misguided hope that that would distract from the nearly identical Buffy face above and to the right. None of the juxtopositions make any sense whatsoever. I just found images I liked and tried to work as many of them as I could in as tightly as I could, jigsaw style. To sum up, lame beyond belief.
Now maybe the new version is lame too, and I'm just in denial. But I think it's an improvement:
I've taken a new shiny copy of the semi-profile image of Buffy from the left bottom corner of the original, and made it my starter image. It's from the scene were she runs into Ken, and she gazes off into the distance and we get the "it sucks to be a poor kid living on the streets" montage. The expression seems evocative of the underlying sadness -- rather than grumpyness evoked by the majority of the Buffy images in the original -- that I feel is the heart of Buffy's emotional state in this episode. She's looking at hard facts of life that she'd like to be able to change, but which are just beyond her power, and she's trying rather unsuccessfully not to care anymore, to move past that part of her life where "you help people" was who she was. I moved it up so that the lost forehead was buried in the canvas edge, and kept it in its original direction, placing it facing away from the rest of the collage to suggest Buffy's avoid-y-ness.
To suggest her relationship with Lily, I used the shot where she's resolutely trying to walk away from Needy Lily, which is of course also a reflection on what she's doing in the Big City waiting tables in the first place. Downcast Lily anchors the left side of the arrangement and suggests a certain kindredness with downcast Buffy (above, a little hard to see, I must admit) both changing their names and not wanting to talk about home. I've filled in the edges with various other caps of the scenery and texture of the episode, and overlaid both the "anne" name tag and Ken's "come home to Family Home" flyer. The coloring is a black-and-cyan-green-sparked-with-gold meant to evoke the city street at night, in keeping with the importance of the setting to the episode.
Click to the left to view the full sized version of the collage.
Our next re-do is from September 2001, for the season 1 episode Angel.
Once again, we've got face over-kill to the extreme. There doesn't even seem to be any rhyme or reason as to why sometimes the faces alternate Angel-Buffy-Angel-Buffy, and sometimes we get two shots of the same character side-by-side. Clearlreconciliationwas about Angel and about Buffy. Beyond that, based on the collage, is anyone's guess. Another key element to the muddle is that I try to cover both the tension between them and the eventual reconcilliation in the gamut of emotions evident on the faces. But I don't give you enough information to even suggest a progression from one to the other, let alone how they got there, so that whole point is lost. About the only positive thing I can find to say about this one is that at least I had the faces in the corners looking inwards, trying to connect. But in a totally, lack-of-focal point sort of way.
I'm not entirely satisfied with the rewrite, and I'm not sure whether hammering away at it more than I already have would help, or whether I just need to get myself a DVD-rom drive and make my own caps to get just the right images to work with, but it's definitely an improvement:
I've now pared this down to the essentials, leaving plenty of breathing room between and around the figures, without unduly isolating them (they may not all physically touch or overlap, but they don't feel unconnected) By placing Angel and Vamp!Angel right next to each other, I hope to convey that (a) Angel is a vampire and (b) this is the episode where that secret comes out. Darla (key figure entirely ommitted from the original), meanwhile, smiles wickedly at Angel's apparant discomfort, suggesting her role as a malicious catalyst in the drama of the episode. My Buffy image is stand-offish, non-commital, wary, a world away from Angel in her own corner of the canvas. But hovering behind her is the memory of their kiss. By this I hope to suggest the central question of the episode: what happens now that Buffy knows Angel is a vampire?
Decoration wise, I've used bits and pieces of blinds and other similar backgrounds from various caps to build a texture which will hopefully be reminiscent of the initial "Buffy-n-Angel's first kiss" scene, both because it's visually striking and because it of course the point where we reach the meat of the episode.
Click to the left to view the full sized version of the collage.
And finally, the Gift. First a little history. An episode collage for the Gift was actually the first piece of fanart I ever did. The summer following its airing, I bought my new computer and I had just figured out how to do captures, and it occurred to me that I should do something with those captures. I had such a good time that the rest is history. Here's what I made on that fateful day in July 2001:
As "first evers" go, it's not actually that bad. I actually seemed to go downhill after that for quite a while before I figured some stuff out. 6 months passed, and I finished a collage for each s5 ep, so I thought it would be fun to try to remake my first piece ever using the skills I'd learned in the intervening half year. And the result was this:
At the time, of course, I thought it was a vast improvement. Now I'm not so sure... Though in the original, the isolation of the individual caps, the way they fail to blend fully into their neighbors, is definitely a flaw. However, my solution (to make the images larger and cram them in closer together) is not a clear improvement. Yes, it eliminates the floatingness, but it produces a cluttered chaotic effect, which is really not the sort of frenetic mood one wants from a collage depicting death and grief. The substitution of the alternate Dawn image is I think an improvement, as her gaze drifts more accurately down towards the fallen Buffy than the version in the original. I also approve the of omission of the ghostly Buffy along the right-hand edge, as that image was (A) from a scene far and removed from the scene being depicted and (B) yet another Buffy face. The substitution for the kiss-image over the sunrise was definitely preferably to the original (which was very odd looking duplicating the bottom right corner Buffy as it did) but is itself a little odd when you start looking at it closely. I do like the improved contrast on the new one, but the accompanying overdone sharpness is not attractive, and the softness of the original fits the mood a lot better.
Other than the above mentioned differences, they both fall prey to the same flaws of excessive floaty-headness and especially duplicate-of-single-character syndrome. There's really no call for the 2 Spikes, the 2 Dawns, or the 2 (or more in the first) Buffy close-ups. The sunrise is however a nice touch on both, adding a sense of place and a little relief from all the faces, and an idea I wish it'd occurred to me to incorporate in my subsequent experimentations, wherein I quickly abandoned settings for excessive face-concentration. The use of the full-screen shot showing Buffy's lifeless body in the foreground and the scoobies coming forward in stunned grief was an excellent idea, but got a bit buried and obscured by the looming faces, esp. in the 2nd one. Adding a golden orange coloring was a decent choice, reminding us of the rising sun and the end of the long night (and Buffy's battle against Glory).
Which brings us to version 2003, now with 2 years of experience playing with photoshop:
To be honest, if I was doing this episode collage now, I probably wouldn't be concentrating on that very final scene on the ground. I'd probably go for Buffy's last few minutes of realization, her dive, her talking with Dawn, or perhaps do one centered around the fight against Glory itself. But in the interest of repeating the exercise of recreating something that is recognizably the same collage, but tweaked, I went with this... I've tried to introduce a bit more tranquility by cutting down on the images clamoring for attention. The observers now create an arc over Buffy's body, starting with Dawn (an image taken from after she descends and sees Buffy's body, just like all the other images, rather than from up on the tower like in the other versions, seemed more appropriate) and ending with a single Spike (we don't need to see his face to know who he is, so no duplicates are required, and the crying shot packs a lot more emotional punch). Giles is already so prominent in the full-scene shot that including his head shot was superfluous and clunky, so I cut it, as I did all the other Buffy shots (the point is she's dead. therefore going back in time to see her alive distracts from what the piece is illustrating). I substituted in a stock image of the sunrise because I didn't think the one I had from the show was working and I like the colors better on the stock. I also used multiple copies and sizes of the gravestone to introduce texture, as well as some of my own custom brushes (both of these shows up better on the full sized version of course)
So, huh, I wonder what I'll have learned by next year? Stay tuned...
Click to the left to view the full sized version of the collage.
And here we are, 2004, and the fourth major revision:
You know this just keeps getting harder and harder. Which is I suppose a good thing for a self challenge. Still, if I don't have to try to get that Buffy-lying-dead shot to play nice with other caps for another year, it'll be... um, well, it'll probably be okay by then, I hope! Because this whole annual revision thing is starting to get interesting...
This year, I decided that a all those close-ups had to go. Dawn's grief became my focal point, a symbol for the others'. But I couldn't leave them out without straying too far from the challenge of recreating the original, so I did my best to work them in. With hopefully a nice bit of atmosphere, and the text integrated this time...
Click to the left to view the full sized version of the collage.
Sep 2004: And proving that sometimes it takes an awfully long time for a simple truth to sink in, here's one from May 13th 2002 that I recently felt the overwhelming urge to redo. Thought I might as well add this one to the redux discussion as well.
Again with the lump-o-faces! sigh. Now yes, there are a lot of great close-ups of Connor to play with in this episode, it's true. But to cram them all into one piece? When you could make a half dozen cool and completey different pieces? Last I checked, Connor wasn't the Amazing Six Headed Boy. At least I did try to work in a few full-body shots here and there, that's a step in the right direction, as in the use of stock pictures to answer the question "Connor's Back where exactly?" But they're fighting to breathe among all those faces, and totally over-powered in scale and emphasis.
For the redo, I definitely wanted to stick with the "Connor's Back" message. I decided to veer away from hurt/confused Connor and stick with sneering/vengeful, to simplify and strengthen the message. So Connor's back, and it's not some sort of sentimental or introspective homecoming, it's gonna be messy. He's back with a purpose. So I selected an image with his weapon in it, to show what exactly that purpose is. And the whole reason we care that Connor's back is because Angel cares, so you really can't have a "Connor's Back" collage without him in it. He's our audience stand-in, so a gobsmacked image works fine. And as a reminder why it's all so surprising, over and above a son showing up wanting to kill his father, I put Angel mourning over his empty crib (smaller, back turned, so he doesn't compete with present-day Angel) and a shot of the portal crackling with energy. I also made sure to give all these elements some breathing room. They touch, they relate to each other, but they don't crowd like the original does. There's a sort of a C-shape your eye can follow from small Angel figure at the top, to stunned Angel on the left, to great big menacing Connor in the foreground. The city's still there, giving a sense of place, and the texture's roughish to emphasize the turmoil. The warm red/orange/brown colors suggest the hellishness of the place Connor's just come from, with the greens and blues over in the upper corner to represent the "normal" world. Angel's still getting a little baked by a draft from Quortoth, though... The text is no longer necessary because the images now tell the story visually, and words -- those words anyway, others might do, but I couldn't think of any -- would only detract from the emotion. As for the change in aspect ratio, no clear cut reason other than I've found it's just easier and more visually pleasing when you're designing for a computer monitor to have something wider than it is tall. Someday maybe I'll figure out a way around that...
Click to the left to view the full sized version of the collage.
Jan 2005: I'm hoping to eventually redo all the collages I did for season 5, back in my first phase of photoshop addiction. starting with 5.01, done Dec 2001:
typical newbie faults, yadda, yadda, see above.
For the redo, I wanted to focus on the main drama of the episode, Buffy vs Thrall. Is she under Dracula's spell? Is she truly helpless in the face of Tall, Dark and Undead? Pared it down to two images and incorporates as much gothic dark texture as I could around the edges.
Click to the left to view the full sized version of the collage.
again, with the typical newbie faults
I'm not thrilled with the redo -- this was a hard episode to try to sum up or tease, so much going on! -- but it's still a sight better than the original. We've got the Buffy-vs-Dawn eye-rolling fest illustrated, and the Harmony and her gang subplot represented. And it's all eminating from the writing-Dawn, just like the episode is.
Click to the left to view the full sized version of the collage.
from Sep 2001
from Feb 2002. Oh nice, a redo that's basically the same thing all over again. 5 months of practice, you would have thought...
from Feb 2005.
Click to the left to view the full sized version of the collage.
from Dec 2001
from Feb 2005
Click to the left to view the full sized version of the collage.
from Dec 2001. Ugg, face-orama. 'nuff said.
from Mar 2005. This is one of my favorite redo's. I think it illustrates well the power of using various distances (foreground Buffy, midground Dawn and distant Glory) to create depth and interest.
Click to the left to view the full sized version of the collage.
from Dec 2001 (Ee gad, I actually said in the notes I was really happy with this one and thought it had a nice flow. uh, hello? The White Cliffs of Mount Crying-Tara-more does not equal "nice flow")
from Mar 2005. The truly ironic thing is how much less effort went into the new one, and how much better it looks.
Click to the left to view the full sized version of the collage.
from Dec 2001. At the time, I wrote: "Obviously I've divided it up into the 4 'stages of Spike' covered in the episode, but perhaps a little less obviously I've got a woman in each, moving diagonally up from lower left to upper right." okay, so there was a plan, that's good. Kind of odd to include one Past Slayer and not the other, though, as far as the theoretical goes. And as far as the execution goes, less said the better. My eyes, my eyes, why does it burn? Oh right, bad fanart.
From Mar 2005. Still more busy and less stream-lined than I like, but I really wanted to fit those 5 women in there. The corner line brings us Spike from start (bottom left) to present (top right) with the middle reserved for the women who shaped that transition.
Click to the left to view the full sized version of the collage.
from Dec 2001. Well, could have been worse. Still ridiculous to have all those Buffy heads (plus the random Dawn and the extra Joyce) crammed in there as if one more slightly different expression is going to make things more meaningful. But I do still like the coloring.
from July 2005. For the redo, I decided to try to reflect more of the episode by covering the three major storylines: Joyce's illness and Buffy's dealing with it, Riley's issues & Glory's snake-plot.
Click to the left to view the full sized version of the collage.
9.6 FLY, BE FREE, COLLAGE!
Here's some things I *try* to consider when working on a new collage, and which I humbly suggest you might want to consider as well:
What's my angle? What am I trying to say? Do I have a point? What is it? Is there a certain emotion or atmosphere I'm going for?
Given my point, what can I do to emphasize it? Is there a certain color scheme (vibrant vs subdued, cool vs warm, etc) that fits in with the message/mood I'm trying to convey? is there a certain texture (rough/harsh/grungy for a bitter angry message, soft/glowy/highlighted for a sad wistful message, bright/energetic for a happy exuberant message, etc). Simple or complex? Bright or dark? To emphasize my message, should my cast of characters be avoiding eye contact? seeking eye contact? staring at the viewer or at each other? what expression should they have on their faces?
For each cap/image/element: does this belong here? Am I including this just because I like it, or does it have a bonafied reason for being here? Is there another element doing this same job, but better? Do I already have enough to convey the message? Does this new addition strengthen or dilute the message? If I already have an image of this person here, what makes this one special/different enough to justify its inclusion?
How's my balance/weight? If I zoom out, is the flow instantly obvious? Are there instantly obvious triangles, lines, curves? Is the arrangement "too heavy" on one side or the other or at the top or bottom?
Good focal point? Does my focal point instantly draw the eye? Is it then easy to get from the focal point to the rest of the image, ie, does my focal point draw me in to the image at a point that makes it easy for my eye to slide around and soak up everything else??
Even lighting? Do my images have convincing lighting? Do they look right next to each other, or is light coming from a different source (color & direction) on each individual one and make it obvious that they don't belong in the same image?
Smooth transitions? Are the backgrounds of adjacent images similar enough to each other and/or to the background that they're blending together well? Are there over-quick changes between images? Any stray edges of the original cap still showing? Are the transitions either gradual enough to look natural, or sharp enough along face edges, shoulders, etc to look good?
How's the image quality? is it good enough at this size? or should I shrink this down at the end to conceal the flaws? Or apply some sort of grungyness to make the artifacting look intentional? or get better images?
Of course, it's very difficult to think about all this at once, and I'm sure I wouldn't want all these questions applied mercilessly to even my latest stuff. I'm just saying, these are the sort of things that I feel like thinking about helps me improve my own work, and so I'm offering it up as some ideas for how to work on your own path... Of course you're never going to develop your own style if you're worried about breaking the "rules", strict adherence to everything I've said is not what I'm advocating. Hell, *I* don't adhere to everything I've said, certainly not in everything I do... but in addition to working on your technical "how-to" knowledge of your graphic program, it helps if from time to time you start making conscious decisions rather than just doing what I was doing initially, which was just throwing images and filters I liked haphazardly at the screen ...
And one more idea: aggressively seek out new things by other people to look at. And then don't just look at it and say "oh, pretty" (well, do that, but that's stage 1!) take some time to really analyse what it is about it that you like. Atmosphere? Composition? Texture? Message? Colors? Then the hard part, figure out what technical details contribute to those general things you like about it, and try to file that away in your head for future reference. And not just on-line fanart... DVD covers, movie posters, advertisements? All over the place, and all bursting to the brim with professional design experience for you to learn from... go forth and absorb!
And I think that brings us to the end of this rambling. Hopefully somewhere in all this mess I've said something useful!