#spreadsheet talk

—

because I’m transcribing for myself first and foremost, I don’t actually follow all of the guidelines set by organisations dedicated to web accessibility; if your goal is more towards the accessibility side of things, you should probably look at those guidelines instead. accessible captions will often remove filler words or stuttering from natural extemporaneous speech; I keep those, because I’m interested in the reality of how people talk, not what’s easiest to read. also, some people will edit out their filler words or grammatical mistakes in post, which is always very interesting to me.**

since beginning the spreadsheet project (a year ago now!), one thing I’ve learned is that most amateur transcription is very bad. there are a handful of blogs out there ostensibly dedicated to transcribing various mcyt clips, and most of them do a pretty bad job at making the transcription readable! common mistakes include overdescribing tone of voice or adding unnecessary detail, such as noting how many seconds a speaker paused before continuing (e.g. “[eight seconds of silence, then reading chat]”) or writing multiple words to describe inflection (e.g. “[he starts laughing, then speaks again]” instead of just “[laughing]”).

personally I differentiate between someone laughing in between sentences with “[laughs]” and someone laughing as they speak with “[laughing]” but that’s not an official standard anywhere, that’s just me. I also tend to keep any descriptions of tone of voice in the “notes” column off to the side, like this:

^I also put timestamps and body language*** in the “notes” column, so they don’t clutter the actual dialogue.

one piece of advice I’ll give is don’t do what I did and completely rework your entire process after months of effort, try to have a decent idea of your preferred style of transcription before you get too deep in it. for instance, one thing that I do is distinguish between lengths of pauses by punctuation, i.e., comma (“right, okay”) -> double-em dash (“right—okay”) -> full stop (“right. okay”) -> ellipses (“right… okay”) -> additional word (“right. [pause] okay”). I recommend deciding early on what you want each form of punctuation to indicate.

also I have personal standards for how I transcribe filler sounds (whether it’s “um” or “uh” or “er” or “erm” or “ah” or “eh” et al.) but that’s just because I’m insane and you probably don’t have to do that, you should just use whichever variation is most common to your local language, that’d be easier.

one benefit of transcribing dialogue from a youtube video is that youtube has an option to view the autogenerated captions for the entire video at once (if you open the dropdown menu and scroll to the bottom) and it’ll look something like this:

^don’t worry about the context for this, that’s just the latest video I was watching. anyway, keep in mind that THESE TRANSCRIPTIONS ARE AUTOGENERATED AND THUS INACCURATE. you can’t just copy them down without actually listening to the audio and double-checking that you have everything. youtube’s subtitles have difficulty with crosstalk and also any accent that isn’t generic american (transatlantic?); don’t rely on them to be accurate, but they can be a useful aid. I will generally write down as much dialogue as possible in the first “round” of transcribing, then go back over it and edit as needed. often I have to listen to a clip from multiple povs (although I do this anyway for the selective editing stuff lol) just to get a reasonably accurate idea of what each person is saying.

note also that the automatic captions on videos older than about a year or so will be markedly worse. you can still use youtube’s transcription as a tool, but it’s extremely bad most of the time. for comparison, here’s a bit from the first episode of third life:

^you can sort of get the gist of what’s happening? this is actually a comparatively good example; a lot of the time the automatic transcription just gives up entirely and there are long stretches of time without captions at all.

for extended vowels (such as the way grian talks when he’s in a mood), I will sometimes just duplicate them, e.g. “nooo” or “whaaat are you doing?” or even when, say, someone’s laughing through a word (“nohohoho”) or adding syllables (“sto-hop”), but that varies by instance and I’ve found it’s usually best to indicate that sort of thing in a note to the side rather than attempt to pinpoint exactly how it sounds. unless you want to use the IPA. I have on occasion used the IPA. for example (from xlife):

^for the first example of a particular pronunciation, I indicated it in brackets, then further examples were off to the side for readability. you absolutely do not have to do this sort of thing every time people pronounce a word differently, though; the main reason I did was because the point of transcribing this moment was to look at the argument over pronunciation.

sadly I do not (yet!) have a consistent way to indicate when someone is Doing A Voice. for the most part I’ve just inserted a tag in brackets, e.g. “[accent]” when ren’s doing his red king voice and then “[normal]” when he returns to his regular voice. I will also often keep track (in the notes column) of when someone is lying or being suspicious or using a particular tone of voice; if I have a name for it (like “dog voice” or “lying voice”) then I’ll usually just add that.

hopefully all of this has been helpful or at least interesting! if you have any further questions, feel free to ask; and, as always, feel free to tag me in any selective editing spreadsheets you end up making. I always love to see more of them.

—

*the standard I’ve heard is that you’ll end up spending a minimum of 3–4 times the length of time of the actual clip on transcribing, e.g. it takes 3–4 hours to transcribe a single hour of content. that’s for radio and television, however, and I’ve found that it takes me longer than that on average, in part because I’m not the best at audio processing (ironic quality to have) and in part because, in improv mcyt, everyone talks over each other very quickly.

**scott does this. obviously.

***two further examples (xlife and third life, respectively):

the relevant part of martyn’s post is this:

my thoughts are like. okay so on one hand you have to look at this from a realistic perspective: this is his job and he wants to make money. understandable. however, this also feels very weird to me! I will be honest, it feels very weird to make certain content inaccessible like that, especially coming from someone who makes such a big deal about lore and storytelling and the implications thereof. to me (<- layperson) it seems more reasonable to do member-exclusive commentary videos instead of making part of your actual storyline inaccessible to people who can’t or don’t want to pay you real human money. (I am not going to give martyn any money.)

objectively, from a narrative perspective however, making certain memories a DLC is Really Funny Actually.

—

further thoughts:

ultimately, the reason this kind of selective editing is interesting to me is because it means the content is different in different perspectives. if everyone went the bdubs route and posted 12h superfan videos, I would be way less invested.* part of the fun in watching different povs is seeing what was memorable to one character that another character would dismiss entirely. I was recently watching joel’s last life, and there’s a scene where he visited jimmy and said some nice things to him; this scene isn’t in joel’s pov, but it’s in jimmy’s. the implication here imo would be that this wasn’t particularly memorable to joel—he does this kind of thing all the time—but it was memorable to jimmy, because it meant something to him that joel would take the time to cheer him up like that. this is interesting character work! you can do some really fun analysis with this stuff! but that would all go away if everything became the same.

for martyn specifically, his editing tends to omit scenes where he acts more vulnerable or empathetic. he thinks of himself as cold-hearted and vicious, when in reality he isn’t at all. I love when characters lie to themselves like that. scott said pearl picks and chooses what she wants to remember, but guess whose pov kept that scene, and whose didn’t? certain characters (bigb for example) have eidetic memory, while others (grian for example) are much more selective: how do these characters interact?

where would the fun be, if we could see everything?

—

I’ve occasionally bemoaned the fact that we don’t get to see everything, but that’s not genuine; I’m just whining to whine. I don’t actually want, say, scott to keep everything in his videos, because then he’d be boring. half the intrigue is in what isn’t being shown.

in double life there’s a blue cornflower planted outside of scott’s house—the only flower there. in his pov it’s never mentioned or acknowledged. but if you watch cleo’s pov, you’ll see that cleo gave it to him specifically because he was jealous that cleo had given martyn a flower (which martyn didn’t even notice). in limited life there’s a conversation between martyn and scott where you genuinely have to watch both perspectives in order to get the full scope of what they’re talking about, because each of them edits out about half of what’s said. in limited life joel omits a comment scott made about his voice sounding high-pitched; from a character perspective this could be interpreted as joel’s own insecurities coming into play.

this kind of digging around to find answers is fun. it’s a way to interact and engage with the content beyond just passively absorbing what’s being shown on screen!

—

just to be clear, I have no issue with rpf as a concept as long as it doesn’t cross the three age-old lines (1. don’t bother the actual people involved; 2. understand that it’s just a fun game; 3. leave actual irl partners and family alone). it’s not usually my thing, but I’ve spent a decade or so in bandom-adjacent spaces so I’ve seen my friends in the trenches, basically. rpf is fine. I don’t really go in for it, but I also don’t think it’s any worse than seeing a picture of a celebrity you find attractive and thinking you’d want to kiss them. it’s human nature really.

the editing stuff does kind of push up against a weird c/cc boundary line* though, I think, because obviously the characters themselves are not doing the editing. scott-the-character is not filming storytime vlogs about abusing his husband, or whatever. it would be really fucking funny if he did, by the way, and I am now distracted by imagining this, but that’s not what’s happening. I don’t even think scott-the-real-person is doing this; I choose to assume that they have all talked about it and no one is being forced into anything. you can see in some povs (but not all. which is another interesting example of selective editing... martyn tends to cut out references to the mid-session break, but ren doesn’t, for example) that they do check-ins before, during, and after each recording session. these people are all friends outside of minecraft, in various combinations. I think grian (real) is probably pretty good at making sure everyone feels comfortable. and multiple of the real people have specifically said they create storylines and lore for their characters (off the top of my head: martyn obviously is notorious for this, but grian also, and pearl, and even the likes of etho, et al.). I choose to believe that, if anyone crossed a line, they would talk about it. because the alternative is imo kind of rude to assume that these grown adults can’t have a conversation about what’s okay to roleplay and what’s not. et cetera.

in that other post I said that memory is fallible and editing is memory, which is a tidy summary, I think! each character’s pov is representative of what (and how) they remember events and interactions. ergo: spreadsheet comparing multiple perspectives.

apologies in advance that most of these involve scott somehow. I have my biases and have never intimated otherwise.

so, for example, in limited life martyn and scott were A Thing. what kind of Thing is up to interpretation; I know martyn (real) has said they were a queerplatonic relationship, and you could definitely come to that conclusion if you only watched martyn (fake)’s pov! but in scott’s pov they were definitely fucking. and martyn cut out a lot of references to that:

in my editing-is-memory theory, this represents martyn specifically blocking out the memory of him and scott hooking up because he doesn’t want to think about it. fair enough lol.

(it’s not that he’s uncomfortable with the idea though. evidently.)

or, another example, scott choosing not to remember ren’s terrible accent in 3L (big same, buddy. I also wish to forget):

or, y’know, jimmy choosing not to think about his own discomfort with tango’s anger issues:

but I also think of this sort of memory-editing as a purposeful thing, sometimes. a great example would be that whole ‘scar, protect me’ scene in 3L where scar cut out half a second of grian saying that:

this is in the middle of a conversation and makes for a frankly quite awkward edit. I do not think that scar genuinely ‘forgot’ that grian said this. I think scar made a conscious decision to focus on the rest of the conversation instead of acknowledging this, because he is protecting grian, by not acknowledging it. this is very maddening and sexy of scar imo. and also grian is just way more complicated.

another classic example that I think really feeds into the idea that this kind of editing can be interpreted as misremembering/not remembering what actually happened is, of course, lizzie specifically redubbing a line to say she was proud of jimmy, when in the original scene she did not say that:

anyway! if you buy into the whole ‘watchers’ lore thing, specifically the idea that they feed on emotions, a lot of this also makes sense. personally I am mildly disinterested in a lot of ‘watcher’ headcanons, but I don’t care about evo and martyn’s not my fave, so that’s probably why! I don’t begrudge anyone their fun.

in conclusion, I will end with one of my personal favourite examples. have you ever purposefully tried not to think about an ex because you’re not quite over them and it feels weird? yeah, scott has also:

also I like to think he cut this because he forgot what kind of fish jimmy was (not a salmon).

TL;DR do you remember the exact colour of your first crush’s eyes? do you remember what you had for lunch a year ago today? do you remember every time you said something kind of mean to one of your friends? well, neither do these guys.

—

and I love parallels. for instance

(scott’s 3L ep 1 at 03:59 / scott’s PL ep 2 at 17:47)

insane narrative framing ftr! their relationship literally commencing in the very first episode ever of the very first season ever with scott in a position of power and authority, literally above jimmy, looking down on him. the kind of foreshadowing that wouldn’t be out of place in professional cinema—it’s very straightforward playing with levels; our pov character is physically higher on set than their partner, so this conveys something about the dynamic between them—but it’s entirely coincidental. because this is unscripted minecraft roleplay.

and then of course the echo of it, all these years on, when scott kills jimmy. the return to jimmy’s 3L skin, so he looks the way he did when scott saw him for the first time, the way he did when they were married; jimmy being literally trapped in the second scene but emotionally (unintentional presaging, ofc) trapped in the first scene, although neither of them knows it yet. in the second scene they both know it though. and that’s why jimmy cut out scott giving him a flower in episode 1 of past life.