#it combines things based on algorithmic thinking

[garecc]

If I see one more bad take about ai art i think I will lose my mind

[adventures-in-a-world-of-fiction]

I've seen a couple of comments from someone around paying Tumblr for stuff that I want to address. I'm not going to mention the person who made these comments because I'm not trying to pick a fight, but I think they're worth talking about. The comments in question are: "you think user money is anything compared to advertisers" and in a pinned post they tell people to not give money to Tumblr.

The thing is, user money can definitely be something compared to advertisers. There are multiple ways that an online company (in general, not just Tumblr) can make money, but let's break them down into three categories:

A. From the users - selling merchandise, subscriptions, premium packages, asking for donations, etc.

B. From advertisers - selling views and space on the platform to companies that use it to try and sell stuff to the users

C. From data - selling information about the user base to other companies that might use it in a whole bunch of dodgy and malicious ways, or just try to find better ways to sell stuff to us

All three of these are viable ways for a company to make money, and many companies use some combination of the above. What matters is what the company sees as their PRIMARY method of making money, because that is what drives their corporate decisions.

If none of the methods are making money, the company will shut down, and I don't want Tumblr to shut down - I like this hellsite. If option B is what makes them the most money, then they will make business decisions that make the platform look better to advertisers and this is likely to drive everything in a more algorithm-centric direction and give users fewer options to curate their own experience. If option C is what makes them the most money, then they will focus on features that enable privacy invasion and data harvesting. If option A is what makes them the most money, then they have to think about how to keep the users spending that money. Now, option A doesn't always lead to good outcomes - in mobile/online games it can end up as loot box gambling add-ins and pay-to-win options, but thankfully Tumblr isn't the sort of site where loot box mechanics would make a lot of sense. Which makes it more likely they'll go the other option: delivering the features that users want to keep them coming back and paying for subscriptions. 

I would much rather Tumblr goes for option A than options B or C because it means that Tumblr is more likely to put the user base first when making decisions instead of advertisers. We just need to show them that it's a viable option.

Tumblr is trying what online games have done for years - crabs and checkmarks are the equivalent of horse armour DLCs and cosmetics. They're trying to make the business work through microtransactions. If enough people spend a small amount, it can add up to a large amount of money. The point of crab day is to send a message to Tumblr that option A is viable so that they make the choice to focus on that. If everyone goes, "No, don't spend money on Tumblr, you're nothing compared to advertisers," then it becomes a self-fulfilling prophecy and Tumblr will have to go with options B or C if they want to keep making money.

I'm not giving Tumblr money out of naivety or because I think they're somehow deserving - I'm giving them my money because I would much rather they make money directly from me and give them an incentive to provide features I like, than by making the site worse so that they can exploit me.

[cleolinda]

My nephew reviews perfume

I'm not sure how my sister heard about Noteworthy Scents, but the concept is, you take a quiz, and they'll send you four Scientifically Chosen samples ($25 USD) based on your answers.

Noteworthy is a new, exciting way to discover your signature scent. Rather than relying on flashy celebrity branding or overblown marketing promises, we want our customers to be in control of deciding which Noteworthy fragrance is right for them. We’re excited that you’re joining us on this journey - we can’t wait to help you find your signature scent.

I have looked at all the fragrances on offer, and I can tell you, they don't happen to have the notes I would want in my One Perfume to Rule Them All (if there's no peach or ylang ylang, I’m not forsaking the rest of my collection). But more to the point, I don't want just one. I've been writing up the things I've been trying for a few months now, and I enjoy the experimenting. But I appreciate what Noteworthy's trying to do, especially for wearers who want something straightforward, and I was perfectly happy to let my sister be the one to try them.

(I did take the quiz, though, and after telling them I don’t like aquatic scents, they said they would send me one that smells like the beach. I closed the tab. My sister told the quiz that she does not like florals and she does not like amber. You’ll never guess what happened next!)

So her Discovery Kit arrived yesterday. She announced this by texting me,

I’m gonna give you feedback from [Nephew] smelling the perfume

Me: Yeah?

As you may recall from one of my music posts, my nephew is six.

(My sister gave me permission to post this.)

n,841

Lemon, tarragon, cedar wood. A powerful, understated blend of citrus and woods. Drawing from poetry’s ability to spark joy and inspiration, this bright lemon and bergamot blend leaves an instant impression before mellowing out into a rich amber and leather base.

My sister: [Nephew] said

smells like raspberries or being in a dungeon like down in a well

Me: ……..I do not see any of those things on the card

n,551

Lily of the valley, amber, sandalwood. A sensual, comforting, woody scent that cocoons the wearer in notes of warm, glowing sandalwood and cozy, soft cashmere—freshened up with Lily of the Valley.

smells like lime juice or being inside a pumpkin

n,307

Mandarin, basil, sandalwood. Inspired by nostalgic childhood trips to lemon groves, fresh flowers swirling together in the summer breeze, and the radiant colors of perfectly ripe citrus, this energizing fragrance stars juicy mandarin and soft sandalwood, plus light notes of basil.

green grape juice or being on top of an umbrella at the beach

Me: that’s very specific

n,057

Passionfruit, vanilla, tonka bean. Let n,057 become your signature: remarkable, evocative and distinctly yours. Keep your fragrance at home or take our travel spray with you. Fitting snugly in your purse, pocket, or suitcase, it's the perfect on-the-go perfume.

My sister: I think he’s losing his focus on this last one. He said

lemon juice at the lemon store

My perfume sample budget got used up on, like, ugh, actual necessities this month, or I'd give Noteworthy a try for science. On a second try, my quiz results were (very likely overlapping with my sister's):

A captivating blend of warm woods and zesty citrus, like strolling through a sun-drenched forest grove. An intriguing fusion of exotic spices and earthy notes, evoking the vibrant energy of a bustling cityscape. A delightful combination of tropical fruits and delicate florals, reminiscent of a refreshing breeze in a lush garden. And for an unexpected wild card fragrance, a scent that defies expectations because science cannot always predict desire.

Currently, my sister says that she can tell that the Noteworthy fragrances are well-made, but they're not "her"; apparently she's in the 11% algorithm failure group—or maybe she just needs time for them to grow on her! Who knows! She'll bring over her samples for Sunday dinner, and I'll report back if they do, in fact, smell like being inside a pumpkin.

Perfume discussion masterpost

[scientia-rex]

Hello, I appreciate your medical posts very much and having seen a post the other day where you said migraine was in your areas of special interest, I'd love to ask a related question. You talk about bodies and medicine and patient experience etc in a way that makes a lot of sense to me and I'd trust your take.

I have chronic migraine. I'm currently at 100% pain days, with varying severity. Very hard to pin down what is prodrome, the main event, and postdrome as it's all blurred into one. My migraine team want me to reduce painkiller usage (currently dihydrocodeine and paracetamol daily, and ibuprofen maybe every other day on top) due to rebound headache. I want to cut down because they're fucking expensive and I'm scared for my liver and kidneys. But I literally can't cope with life without them. I went off them for four months a few years ago and the pain was so severe and so debilitating I was the most suicidal I've been in my life. Without painkillers I can't get to the toilet unaided, rarely leave bed, even more rare to leave the house. It's hell. And that's not even considering the effects on everyone around me who has to pick up to care for me.

So what do I do? The way I see things, I need something to help the pain improve before I can use less painkillers, but the longer I go on trying to find something that works and not getting there, the more I think maybe I'm wrong in that. I know a bit about how codeine based painkillers can reduce your pain tolerance / pain baseline. I don't think it's an addiction issue because I've been at the same (over the counter) dosages for 4 years now. I just want to do all that I can to be better, but I also need to be alive to be better. I am stuck.

TL;DR - If you have any thoughts on the relationship between chronic migraine, painkiller use, preserving quality of life while finding a treatment, and increasing the chances of a treatment working, and where on earth the balance between all that lies, I'd really like to hear them.

Again, I absolutely appreciate if you can't answer this, don't want to etc. Giving advice online is notoriously tricky and all that. But a big thank you for your time in reading, and all your weight and exercise posts especially which make me feel so much better about my body. Wishing you all good things! 💖

I won't speak to your case directly, since I'm not your doctor, but here is my personal algorithm for escalating treatments for migraine (note that "abortives" in this case means something you take after a migraine starts to try to end it, while "prophylactic" means a daily treatment you take to reduce likelihood of developing a migraine):

-OTC combination of magnesium, feverfew, and butterbur, taken daily

-Triptans (insurance will usually demand patients fail at least 3 to cover a more expensive treatment)

-High-dose NSAIDs (as abortive treatment given risk of rebound headaches if used daily)

-Daily topiramate (insurance will always demand this is either failed or there's a clear contraindication)

-Daily calcium channel blockers

-Daily beta blockers (higher dose than used for anxiety or low-grade arrhythmias)

-Daily anti-epileptic medications (such as Lamictal)

-Monthly anti-CGRP monoclonal antibody injections (Aimovig or Ajovy; expensive so insurance will demand you've failed some or all of the previous meds)

-Abortive anti-CGRP orals (Nurtec or Ubrelvy)

-Abortive ergotamine, usually Migranal, a nasal spray (very expensive and must be repeated 15 minutes after initial dose regardless of whether symptoms are improving or not)

-Prophylactic Botox (I believe this is every 3 months, must be done in the office of a trained and licensed professional, usually but not always a Neurology provider)

-Sphenopalatine ganglion blocks (done by dripping lidocaine far back into the sinuses to reach the sphenopalatine ganglion, again in the office of a trained and licensed professional)

-Cephaly (transcranial magnetic stimulation at-home device), expensive so insurance hates covering it

Now, one of my newer tools, and my current personal favorite, is a greater occipital nerve block--easy and fast, low risk, and I've had about 90% success with my patients in aborting current headaches. Effects seem to last 3-4 weeks in most cases and since it's straight lidocaine (you don't have to include steroids, though you can) you can do it as often as needed. I generally do this in my office, but I did train one patient's spouse to do it at home given how frequent their headaches. The pharmacy lost their fucking mind about letting an outpatient have lidocaine. I don't know why.

I currently manage my pretty awful chronic migraines with a combination of monthly Aimovig, as-needed Excedrin (the combination of caffeine, Tylenol aka paracetamol, and aspirin is effective for many people but is a real risk for causing medication overuse headaches, the more official term for bounce-back), as-needed Ubrelvy (I can sleep after taking Ubrelvy but not Excedrin so it's a good option), and roughly monthly greater occipital nerve blocks (I teach my trainees to do it using myself as a subject). I wouldn't mind trying the Botox but it's a PITA to get in to see our only local Neurology provider and since my migraines are relatively well-controlled (probably 1-2 headache days a week right now) I don't think it's worth the effort.

I also really got a lot out of this lecture, so give it a try.

[fanhackers]

Vidding’s Grandchildren? Edits, corecore, and other video feels

Thinking about the descendants of vidding, since I was quoted in this recent article on fan edits, “Why Do Fan-Made Trailers Rule the Internet?” by Cat Zhang. The edits of the article, like the fanvids of old, are scenes from television shows and movies set to music.  But while these edits are typically much shorter and more feels-focused than vids, they seem to me clearly a descendant of the form. In my book, Vidding: A History (2018), I talk about the ways in which YouTube and the algorithms of the internet were already affecting the aesthetics of vids back in the 2010s (spoiler alert: they’ve became shorter & more intense) and we can clearly see this trend in the 2020s now that fans are firmly on short-form platforms like Insta and Tiktok.  The edits in Zhang’s article are all about the feels, and a sub-class of edits, corecore (as explained in this Mashable article by Chance Townsend, “Explaining corecore: How TikTok’s newest trend may be a genuine Gen-Z art form”) is often used to express chaotic or overwhelmed feels.  Townsend says that what makes corecore so interesting is that “one’s feelings that couldn't be expressed through words are instead presented through images. Whether that emotion is happiness, a fear of the future, or the excitement of falling in love, corecore edits, through the use of multimedia, speak to our common experience.”  The idea of expressing emotion by the artistic act of combining disparate clips with music–well, it sounds like vidding, but at the same time it seems a long way away, too. That said, a work like this hip-hop based edit of The Bear, made by an artist at the X/Twitter account “black boy cinematic universe,” seems to be doing the kind of reparative fannish media work vis a vis race that older vids did for gender and sexuality. Zhang quotes the artist as saying: “There’s an energy to the show where it’s being carried by the people of color. So in my edit, I want to make sure there’s a song that represents that.”  That’s a very similar (and familiar) vibe: that urge to make the thing that will Get. It. Right.

–Francesca Coppa, Fanhackers volunteer

[lonelyroommp3]

u dont have to answer this but i was wondering if u could elaborate on ur opinions about rpf -- bc i kind of agree with you that it's like, it's fine and it's not morally wrong or anything but it def is a little weird? idk ive never been able to rly explain my feelings about it so im curious what u think

honestly my main view on rpf (whether of the y/n self insert variety or the gay shipping variety) is it's literally fine as long as you're not putting it anywhere the people involved are going to see it. like when you boil rpf down to its bare essentials it is just fantasising about a famous person (or multiple famous people, in varying combinations), which is something human beings have been doing in their brains since people first started becoming famous. and i think expressing those fantasies through creative outlets like fic, art, 2013 polyvore outfit boards about going on a yacht trip with harry styles, etc is an entirely natural progression of that impulse

and in that way i sort of view it as similar to any other fantasy you might have about another human being: it is totally normal and essentially morally neutral, i hope you'd agree, to daydream about a person you fancy, whether those daydreams are PG-13 or heinously X-rated, whether they're a friend or acquaintance or stranger or celebrity. where it would become weird is if you went up to said person completely uninvited and said "hey, would you like to hear in elaborate detail about the things i thought about you doing to me while i was masturbating last night?" - and so i think the same applies to rpf. like obviously don't send your rpf to the person in question, don't show up with BLINK TWICE IF THE BABY IS FAKE signs to a louis tomlinson concert, archive lock your rpf on ao3, don't proudly post about it on websites that the people in question are active on especially in this age of algorithm-based social media where unless you lock your account you have no real surefire way of ensuring they'll never see it, etc etc. every time i see people talk shipping in the comments of an official f1 post a part of my soul dies

i think another big thing about rpf to me is that all celebrities are essentially playing a fictionalised version of themselves. no matter how authentic they seem to their fanbase, when you combine things like media training + PR/marketing obligations + building a Persona + the pressure of competing in a sport (for athletes) + the level of code switching inherent to being in what is essentially a public facing job, we as fans & observers are not seeing the real harry styles or taylor swift or charles leclerc or whoever else. we are seeing the version of that person that they want us to see (or, in the case of historical rpf, you're writing about a ghost reconstructed from fragments distorted through the lens of missing evidence, potentially biased historians, potentially even more biased first and third party accounts, etc), and so i don't really view rpf as inherently invasive because you're not really writing about the real person, moreso the constructed image of Celebrity Of Choice.

of course, that is not to say that rpf can never become invasive: the clearest example would be ship truthers harassing the people involved due to what they perceive as "evidence" of some great forbidden love story, but i'll be real there is a lot of F1 rpf specifically that i side eye from a distance because of how it will use drivers' intensely personal real life trauma as a plot point, which is a line i'm not really comfortable crossing with real living people, especially not when it comes to dangerous sports lmao. but i think that is an issue with some rpf enjoyers' boundaries and not an inescapable root problem with the very idea of rpf imo

[yuyu-finale]

Hi, sorry if this is an annoying question, but I would like to ask for advice regarding posting art on social media (you’re one of my favorite artists and I know you’re multi-platform). My friend wants to start posting art on social media, but they’re unsure which platform is the best. Would it be possible to ask for your experience regarding different platforms? I hope you have a wonderful day.

Hi anon! So excited for your friend to start posting their art on social media! It’s a lot of fun and a great source of inspiration and motivation to keep drawing and improve. It’s a fun way to make friends too!

I may not be the best person to ask about this because I don’t know much about things like. Engagement and algorithms and stuff, so this is all based on my experience.

So I have three art accounts: Twitter (main), Tumblr, and IG. I can’t really speak for other platforms but I think these three or like a combination of two of them would be pretty good for artists.

You might know I’m most active on Twitter. Twitter is great for starting out, making friends, and building/joining communities and fandoms. Once you have a starting point, growth can be quite quick as well depending on certain factors. The thing with Twitter is that the lifespan of a post is just like a couple of days at the most, and making all your art accessible in one place isn’t very straightforward. I do recommend making a Twitter, as the quick sharability of posts makes it a great platform for fanart to be seen by more people. Just make a thread of threads and retweet your own art every now and then to make it more accessible!

Tumblr is fantastic if you’re in specific fandoms or communities because the tagging system makes it easy for new artists to have their work seen. Reblogging is a slower process than retweeting so you might get lower numbers, but the lifespan of a post here can last months or even years, so all in all it’s still a great place to post art. You can use it as an archive too, as well as a place to keep a portfolio of sorts if you make a designated art tag.

Instagram is okay. It’s a great platform to see all your art in one place but I find discoverability on there isn’t that great for smaller artists, so it’s best to have IG and at least one other platform. I think it’s better for artists that already have a bit of a following, or as a second/third account. Also, the image cropping can be annoying. IG reels are great for posting process vids though! I don’t do this, but I have friends who do.

TLDR i recommend twitter (although twitter is more of a hot mess now than before so this might change) + tumblr and/or IG. My experience on all three has been positive so I think it would work out well for your friend!

[paddockpatrol]

The Female Drive on IG

Firstly, if you reblog this, please don't tag it with hate towards any drivers or their fan bases; that would be the complete opposite of what the original post is trying to do. It is never an entire fan base that acts out of line, but rather individuals within them.

Secondly, this isn't even really meant as a post about Max Verstappen, though he is the subject. Rather it's a post aiming to highlight how someone's feelings towards a (sports) celebrity can escalate. I doubt any of the authors of the quoted tweets started out with that level of vehemence towards Max, yet here we are. Rather I suspect their feelings escalated over time.

I decided to make this post because Drive to Survive 5 is about to drop and a new f1 season about to begin and passions/emotions will rise, mine included. In combination with has been happening with the tumblr polls over the past week, I am fearful of an even more toxic year in fandom than the past two. Moreover, I've seen too many people be dismissive of criticism of the hate polls towards Max and others. And while likely the creators of those hate polls in their minds were only having a bit of fun and would never think that they would ever write anything like the tweets above, I do wonder.... is that not where it starts?

To be clear, I am not saying the people behind those specific blogs ever would write something like the vile tweets above, but like a YouTube algorithm these things tend to spiral and we are all susceptible to the passionate craze sports can elicit.

So, ultimately this is a post asking all fans -and also very much myself- to consider that there are real people behind blogs (well maybe not the pornbots 😉) and that all the drivers definitely are very much human. There's nothing wrong with writing a heated post, but equally there is nothing wrong with keeping that post in the drafts until you've calmed down and considered what it would be like if you were the subject of your own writing.

Let's have a highly competitive but non-toxic year all. Can't wait for cars out in Bahrain.

[jewishbarbies]

I find the whole Swifties for Palestine thing kind of fascinating as an outside observer. Before I quit TikTok for good, the algorithm kept sending me videos of Swifties whining about how Taylor wasn’t “speaking up about the genocide” and how disappointed they were. I just kept thinking to myself, if they were so upset, why were they making TikTok videos about it hoping to get her attention? Why didn’t they just stop following her and buying her music?

From what I can gather from being somewhat Swiftie adjacent, this group of fans has been pretty vocal in their dissatisfaction. But the thing is, Taylor Swift’s album is still at the top of the charts and she’s still playing sold out shows. So none of this “activism” has made any difference.

This leads me to believe that 1) this group of fans is very loud, but ultimately a very small segment of her fan base and them not consuming her products doesn’t make a difference or 2) these people are claiming they are fed up with her and boycotting her online but they actually aren’t and are still supporting her. Probably some combination of both. I don’t know, I think it’s just really telling about how much of this online activism doesn’t translate into real action or real life.

reading this like it’s my morning paper

[detailtilted]

hi! I'm sorry if you've talked about this before, but I've been wondering how you enhance the spn con videos. like, how does it work? cause your videos look amazing, especially in comparison to the previous footage!

Thank you so much for being interested enough to ask!

At a very high level, the visual enhancements to the original videos themselves are made by running an upscaling program and then manually adjusting color-related settings in my video editing software. Sometimes I also fix aspect ratios if the original videos are squished.

Below are way more details than you probably ever wanted. 😅 I’ll put my very long answer after a break...

Upscaling

The very first thing I do when I start a new video is to take all the video sources I can find, pick out the ones I think I'm most likely to want to use, and see if I can get them to upscale. This is actually one of the easiest parts of the whole process, at least in terms of the actual time that I myself have to put into it. The computer does most of the work.

The upscaling is done through a software program. I don't actually have any idea, but I imagine it might be a dumbed down version of the type of software that movie studios use when they create upscaled Blu-ray releases of older films. I know the obvious question is, “What’s the software program?” I’m a little scared to say it without some explanation first, because it has “AI” in the title. I know and understand (and agree with) why AI is such a hot button today. However, this is not generative AI where it generates a new work out of nothing, and it’s performing a function that a human could not reasonably do.

The way I understand it, it runs an algorithm to analyze all the frames in the video and figure out how it can remove some of the visual “noise” and also add missing details to some frames based on what it finds in other frames. It then updates every single frame in the video. The videos I’m currently working with have 30 frames per second. A 30-minute video would therefore have 54,000 frames. For a human to compare 54,000 frames and make all the tiny changes required to each one would be… impossible I think. So anyway, after that long preamble, the name of the software I’m using is “Topaz Video AI”.

There are a variety of settings you can tweak, so I focus on upscaling one video source at a time because the same settings that work with one source may not work for a different source recorded with different equipment. I try different combinations of settings to small sections of the video until I find settings that seems to return noticeably improved results without adding any weird distortions. Sometimes I have to give up and proclaim the task impossible. If the video has too much noise and/or not enough good detail already there for it to work with, then it won’t upscale well. I get much, much better results with original video files than I do with files from YouTube. I think the videos were compressed when they were uploaded to YouTube and lost some of their quality.

If I find a setting that works, I tell the software to apply those settings to all of the videos from the source I'm currently working on and then leave my computer to crunch through them. That can take hours, and it makes my computer sound like it's preparing for liftoff, but it doesn’t require my involvement. I’ll often let it run overnight, then check it the next day to see if it came out ok. Sometimes I have to go back to the drawing board because I find parts of it didn’t come out well at all, but usually if the sample looked good then the full result also looks good.

Then I take my next video source (another set of videos recorded by a different person on different equipment) and repeat the process.

Color Corrections

This is all done in my video editing software. I’ve been using Adobe Premiere Pro. There’s free software out there that might be a better choice. I tried one early on and didn’t do very well with it, but I had no idea what I was doing. Video editing is new to me. So for me, the advantage of using a very popular software program was that there are tons of answers out there for just about every question I’ve ever thought to ask. Now that I have more experience, maybe I’ll give some of the free options another try. I have to decide whether or not to renew my current license at the end of the year and the price is pretty crazy, especially considering it's only for a limited period of time.

So… back to color corrections. My original attempts at this did not start out well. My first set of videos were the CHICON 2007 videos and those colors looked fine to me, so it never even occurred to me to adjust colors. When I moved on to Comic-Con 2008, I was confronted by a bunch of badly red-tinted and green-tinted videos and I thought, “Hmm, I wonder if this video editing software can do anything about that.”

And I tried. I drove myself absolutely mad. I’m not good with colors in general, I’d never worked with color adjustment settings before, and my eyes were going crazy trying to figure out if I was making things better or if I was just creating a new variety of bad. I put out a plea for opinions/advice here on Tumblr with some screen shots showing what I’d done so far, and sensitivehandsomeactionman came to my rescue. They took one of my screen shots and adjusted the colors in their own software to show me what could be achieved. And it looked amazing. Having that example to reference was a huge help. It was still a challenge for me, but I eventually got pretty close to what they did.

I’m not as good at it on my own, but I learned a lot through that process that I’ve been able to apply to my subsequent videos. I think every video I’ve worked with since then has had at least slightly wonky colors, so I've had a lot more practice. There are a bunch of complicated color-related settings in Adobe, but eventually I found the “Lumetri Color” panel which has a simpler set of set of controls consisting of a few sliders that let you adjust between two opposites until you find the right balance.

One slider is for Exposure and it basically changes the lighting from darker to lighter. Then there are two color-related sliders. One shifts between blue and orange and the other shifts between green and purple. I play around with those sliders until I think the colors look more natural, focusing mostly on their skin tone. Then if needed, I’ll adjust the Saturation slider a little to tone things down. For the most part, playing with those 4 sliders usually gets me something that I think looks decent.

Some videos are more difficult than others though, and I still consider this to be very much outside my wheelhouse. When I announce a new video on Tumblr, I usually display a comparison image of the original video versus the enhanced. Those help me see that the end result does in fact look at least a little better than the original. Sometimes after staring at the video for hours on end, I start to doubt.

If there are any wonky aspect ratios making the video look squished, that’s a really easy fix by just changing the scale percentages for that video in my video editing software.

All The Rest of the Work

All those things I just babbled about probably take up only 25% of the time I spend on each video. Less if the videos don't give me much trouble with upscaling or color adjusting. I could write another book about the rest of the process, but what I've already typed is probably overkill as it is so I'll try to be brief since this stuff isn't as much about the visuals anyway.

This is when I move on to the process of choosing which videos to display at which times and editing everything together into one (hopefully) cohesive video. I also try to choose the clearest audio that covers most of the panel. Audio changes are very noticeable, so I try to minimize them and use a single audio throughout the whole video as much as possible. Often the audio you’re hearing was from a different video than the one you’re seeing.

Once I have all the footage edited together, that’s when I start adding the subtitles and the extra content. These things are by far the most time-consuming part of the process, especially the subtitles.

I hope this answered your question. Thanks again for asking! If I didn't answer your question and if you want to risk having another book typed to you, let me know. 😅

[smute]

i always get the craziest ads on pinterest so ive decided to collect a few of my favorites.

i basickly only use it to look at architecture and design stuff these days so some of it makes cents based on my interests i guess? like a 7000€ oven for example. or a grill for the same price. or how about some premium front panels fitted to your ikea kitchen that cost more than the entire rest of your belongings combined?

like sure. yeah. algo sees me algo knows me. but it's so fascinating! like i always wonder why these guys even advertise online. and if whatever algorithm pinterest has really doesn't know that i'm just a broke student looking for inspiration for their sims builds. i mean by far the most common thing i see is designer furniture that's so exclusive you cant find prices for it anywhere. why would they advertise on pinterest AT ALL? dont you have a flagship pop-up store in st moritz or something? it's weird to think that there are people on pinterest in that sort of income bracket – the same pinterest where i collect my budget meal ideas and diy hacks 🤨

it gets weirder though. 1100€ whiteboard (324€ without the stand (serious architect's office vibes here)). 200€ doormat. (A DOORMAT!) and that last one is a 185€ brass light switch... 🫡

christmassy penhaligon's ads and 50€ underpance seem downright reasonable in comparison. except the pance guys also sell 15€ laundry the turgid in 240 ml bottles. that's 8 freedom ounces btw

[bright-and-burning]

okay yeah i didn't know that there's the different ai stuff and that this is generative ai BUT GOD AM I HAPPY to see that you know the difference between ai and ml like !!! this is not ai !!! it's such a trend now and it's so wrong and so frustrating beyond the, in this case, misogynistic aspects of it. like this is not ai !! but that sounds catchy and URGH so frustrating. okay that's all i love ur additional tags thank u xoxo

yeah!! i did my undergrad degree focused on this kind of stuff (and dropped out of doing a phd in it at the last minute) so it's Literally my roman empire. like. i took so many courses in this and THEN it blew up right at the end of my degree and the misinformation........ chatgpt getting massive my senior year of college made me soooo mad i legit started to resent my field. like this particular use is particularly egregious (why hire a woman in an area severely lacking in women when we can rampantly misuse important computational and REAL LIFE resources to create a fake woman to "talk" abt SUSTAINABILITY!!!) but in general i am like constantly infuriated by AI bullshit from a lot of different sides bc Everyone overestimates/glorifies it/turns it into something it isn't, essentially. (rant abt AI from an AI perspective under the cut that idk might be informative or interesting LOL. i tried not to get technical but i did get mad)

generative ai drives me BONKERS bc it's literally not artificial intelligence. chatgpt is NOT AI it's fucking glorified predictive chat and all the dumbass tech bros on linkedin and twitter who hail it as like world changing infuriate me lol. like chatgpt literally works by calculating the word with the highest probability to come next in the sentence/in response to that prompt based on the data it trained on. is that super impressive ignoring the real world stuff going on? yeah!! it is! it's doing really well and it's fascinating in an academic sense. but then you put it in the real world context, where dumbass tech bros and business leaders worship it as god and where SEO morons use it to turn the internet to sludge and it's like oh god WHY.

and the ai art bullshit oh my god. ok i took a machine vision class right. and like literally one of my projects was to write code that could take in a photo and output it in the "style" of another photo. like as the cs version of a creative exercise, basically (they give u a Lot of projects that are basically write ur own version of an algorithm that's already been written more efficiently by someone else, bc that's how you can kind of pick it apart and really understand it, it's like reverse engineering) to show us how that works. (photos from that project are below; i took the cactus photo and then i "combined" it with a monet. i am STAUNCHLY anti-ai "art" btw this was Lich rally an assignment lol). but do you know what i learned (and what my whole class learned)? it's not fucking magic. it's LINEAR ALGEBRA. it's linear algebra on a truly batshit crazy level, combined with some crazy optical physics equations, but it's literally math. technically speaking, if you had infinite time, and were really insanely good at math (and really really perfect abt not making mistakes), you could do it by hand lol.

ai art from prompts? that's just turning words into numbers (not hard!! i did a project that did that in my second year of undergrad in literally 6hrs!!!) and THEN doing some crazy math and turning numbers into PIXELS! the computer does not know what is going on. it is a FANCY CALCULATOR. WHEN U THINK ABT IT LIKE THAT ASSIGNING HUMAN TRAITS TO MATH IS INSANE.

like tech bros who are like bro chatgpt is aliveeee bc they talk to it is like. if i made my high school graphing calculator draw a smiley face and then was like omg it has emotions......

anyways. like do you know how many projects ive worked on that use AI/ML for GOOD??? like. same kind of techniques that ppl use to create generative ai "art" algorithms? used to do things like detect cancer in scans before human doctors can do it with any confidence. isn't that so fucking cool???? i interviewed w a professor who used machine vision (so literally the field that is now seen as being abt ai art) to figure out what nutritional/vitamin deficiencies ppl in remote villages in madagascar are likely to have based on SATELLITE IMAGERY. so that the overworked underfunded public health ministry could more easily meet their needs without necessarily having to do expensive testing on everyone! i mean, shit, i've worked on really cool sports analytics projects using machine vision. that's not exactly saving lives lol but like. just goes to show how many positive applications there are!

in terms of chatgpt vibes like. i've worked on natural language processing! it has so much more potential than spitting out misinformation!! silly projects for classes, like classifying what political party a politician belonged to based on their tweets, but also more serious stuff in the research i did, like analyzing international public opinion by demographic and country on various conflicts based on individuals' social media posts! analyzing covid vaccine opinions based on demographics, and how to encourage vaccination rates based on that!!

idk it's just. infuriating. that ceos and dumbass business majors (sorry to business majors but i have met a Very Specific Type who like to butt into my field and i am Not A Fan) have completely twisted these really and truly interesting projects and applications. to continually make more money and to cut out/replace more and more people. and the way funding has mirrored this kind of interest, in part bc it makes money (the remote villages nutrition thing is never gonna make money. working on chatgpt on the other hand...) and in part bc of like. basically fear-mongering? you get so much more attention for research in these really dramatized fields (and people lie ALL THE TIMMEEEEEEE to get more interest) and it's all CAP. it's BULLSHIT. and it's just to create buzz for big tech companies' bottom lines. they don't care that it's fear based, or whatever. like people being terrified abt the impact of generative ai bc of bullshit headlines and disinformation HELPS THEM make MORE MONEY. and it's so FUCKING STUPID!!!! it makes me SICK!!!!!!

anyways. it's all math, at the end of the day. and i found a job where i get to use it to help people but i still am like deeply grossed out by what comes out of my field and gets popular bc . i got into this bc i loooove data i love information i love finding things out. and i love using those things to help people... and there are ppl using the things i love to hurt people......... it makes me so sad. and mad.

[yeah-yeah-beebiss-1]

Not to be the decider of what art is but I think something created by a living thing or the innate beauty of the world is radically different from something created by a learning algorithm. The things you mentioned that are problems with AI art aren't separate from its makeup. One reflects something about being a living thing in the world and one is the sum total of all relevant art being treated as a data point to create something accessible and meaningless. It's automated capitalism which I could see being used to say something by another artist but by itself is nightmarish and hollow

Thing is, AI-generated art doesn’t spawn from the ether any more than a block of clay could become a sculpture on its own. If it did, every museum would be plagued by ghostly clusterfuck paintings spawning from thin air in the middle of the night. These algorithms are programmed by humans, the combination of prompts used to guide the algorithm are provided by an individual, and the data being input is art created by humans (at least for now, it’ll likely become more incestuous as AI output starts to enter the pool of data). If you chopped up a bunch of magazines at random, stuck the shreds in a leafblower, and blew them onto a glue-coated canvas, the outcome would still be considered a (bizarre and ugly) collage even if it was technically the leafblower and air currents that arranged them and the input was magazine pages made by others.

I won’t deny that the ethical problems inherent to current AI-generated art algorithms and how businesses will inevitably utilize them are nightmarish. But the “it’s not real art” argument rankles me because we don’t just get to arbitrarily say what is or isn’t art based on the context it’s used in. Vector illustrations are still art even if we all loathe those insipid soulless Google shape-people images. Screenwriting is still an art even though bland focus-tested Marvel scripts make bank. How much value these things have is up for debate, sure, but on their most basic level, they’re still works of art (in the literal sense).

[fipindustries]

first course completed!

that was arc 1 of introduction to magic.

and this is the first drawing i made of the main characters all the way back in 2019

what is this

now that the first arc is done i would like to take a chance to actually stop and properly self promote a little here. it was about time.

i am currently writing a novel called Introduction to magic, an examination of a magic system i came up with by way of following the lessons of magical apprentice Katerina Dolcevita under the aprenticeship of Maria Bellanova. the best way i have to summarize the feel and tone of it all is Fantasy AU-R63! Rick and Morty. imagine if rick and morty were both women and they were exploring magic rather than sci fi.

the first arc is intended to be a prologue of sorts, to get you up to speed with the nature of this world and the nature of the two main characters.

where this came from

i came up with the story and the concept around this world proper in 2021, while reading pale. i was looking at all the extra material that wildbow had written for its magical world and i found myself really wishing that we didnt have to just see excerpts from the magical books like famulus or 100 years lost. and it came to me that if i really want a book about magic to exist then i can just write one myself.

the magic system in this story follows a bunch of my own deeply held beliefs about life and magic. mainly that is kind of silly how we insists magic doesnt exist when we have things like computers and planes and psychodelics. it occured that if we were to live in a world where magic was real we wouldnt call it magic either, we would just think of it as the normal state of affairs. there is a post i read here, which i cant find right now, which said something about how weird conciousness is, how strange the fact that conciousness arises from the specific configuration of a brain. about how conciousness is the last, mysterious, seemingly ineffable property of reality. they concluded that we live in a fantasy world where our magic system has only one spell "summon daemon". obviously a lot of it was merely poetic and rethorical devices to see with fresh eyes of wonder something we take for granted.

and a lot of this book stems from a similar wish to want to see the mundane with eyes of wonder once again, which is why i insist that low level magic in this world is things as basic as writing and lighting a fire. is also the reason why a lot of the titles that i use for the magic specializations are normal every day professions.

the second, stronger impulse was to try to come up with a system of magic that could be broad enough to grasp all possible forms of magic humans have come up with and yet simple enough that could be understood in a few pages. the classification system i use on this book is based on all the broadest, most basic forms of magic practisce that i have seen in history and fiction. manipulating signs as drawings and writing, manipulating sound as voice and song, using tools and props, moving the body, combining and refining substances and materials, handling living creatures.

what's next

as we move forward on this story the format will switch to a more traditional narrative where we will properly follow the adventures of these two ladies. ocassionally i will dip back into textbook-like sections where i infodump about another interesting concept i thought of, but even in the narrative sections most of the chapters will be an excuse to explore some concept or idea. i gather by sheer quantity 80% of the content of this story is going to be maria explaining things to Katerina.

so yeah, if that sounds at all interesting please do read this, and please do leave a comment, i really want to see what people think of this work. i dont need money or donations or to be engaged in any algorithm, all i need to stay motivated and energized and thus continuing the story is to know that people are invested in it.

thank you so much for reading.

[thekingofgear]

We've made a new pedal!

Here's the first teaser for our new Pitch Magpie pedal, a pitch shifter that aims to capture the best of the weirdest pitch shifters from the 90s...

Background

For the past couple years, we've been trying to make a delay-based pitch shifter that sounds good. It hasn't been easy. There's a reason that most pitch shifters have switched to FFT-based algorithms: latency. There's a reason the monophonic Whammy is more popular than the delay-based PS3. Usually, to make a delay-based shifter sound decent, you need to use long delay times, which creates some serious latency. That's fine for organ type effects, but it can be pretty limiting. In contrast, reduce that latency too much and you've got a metallic ring mod, especially on the lowest string of the guitar. But with some tweaking and some crossfade tricks, we think we've found the perfect middle ground. And although there's still a tiny bit of lag, you probably won't care once you add the filters or vibrato haha!

Why filters and vibrato? Well, when I have used those old delay-based shifters, I always thought they sounded best when combined with other pedals — especially for creating special effects! So I though: what if those effects were combined with the shifter? Not just slapped on after the pitch shifting circuit, but actually integrated into the circuit, so that turning up the feedback ("overtone") effects the filter or the vibrato too. The results were even better than we hoped! Sure, it's nice that the pedal can double as a random chorus or a modulated highpass with the pitch shifter section disabled. But turn on the pitch shifter, and the resonance of the filter and the overtones of the pitch shifter start to interact as you turn up the feedback. And if you set the shifter near minimum with high feedback, you get all sorts of comb filter effects. With just a few knobs, the possibilities are really surprising!

Features

It's hard to express how many ideas and variables we packed into the Pitch Magpie. But the three modes are a start. Each is based around the same core pitch shifter, but each offers a unique effect:

Glis Mode (“glissando”) generates an auto-slide, adding a pitch slide when the input signal goes over a certain threshold (think: Hounds). The slide time is controlled by the param knob, which also selects between two slide shapes (basic or U-shaped).

Res Mode applies a resonant filter, selectable between lowpass and highpass with random modulation options. There’s always a bit of resonance, but turning up the overtone knob adds even more, with feedback and oscillation at extreme high settings.

Vib Mode applies vibrato to the the wet signal. Setting the pitch knob to center bypasses the pitch shifter, allowing you to create chorus effects by blending the dry signal. It can even work as a random vibrato (tape warble) with the param knob at max.

With all of our pedals, we try to keep the controls as concise and performance-friendly as possible. We could easily make this kind of pedal with a ton of knobs. But we wanted to find sweet spots and make it quick to move between them. Though we also made sure to give enough adjustment for things to get weird!

The Pitch Magpie's Param knob on takes this idea to the extreme. The knob is most dramatic on the Res mode: left of center, it controls the frequency of a lowpass filter; right of center, it controls the frequency of a highpass filter. But set the knob to minimum and it randomly modulates the frequency of the lowpass. As maximum, it does the same for the highpass.

The Pitch knob is quantized to the semitone, offering 31 distinct settings. For continuous pitch control (for whammy-style effects or microtonal adjustments), the expression input can be used.

As on the Oxford Drive and EQ-201, the footswitch is configured for “flexible” bypass. Tap the switch briefly to turn the pedal on and off. But hold down the footswitch, and it was act as a momentary switch, keeping the pedal active until you release it. This seems particularly useful with this pedal, especially on more chaotic settings. You can also alternate the default bypass state of the pedal by holding down the footswitch while plugging in the power jack.

[fandom-wordsmith]

General Fandom Terminology

Do you want to be more involved in fandom, but you don’t know all the crazy words fandoms use? Below is a collection of common fandom terms and their definitions. 

Fandom - The collective fans of a piece of media (including TV shows, movies, books, comics, video games, etc.).

Fanfiction/fanfic/fic - a fictional work created for free by fans of an existing media about the characters &/or world of that media.

AO3 (An Archive Of Our Own) - a website that acts as an archive for fanfiction, run by the Organization for Transformative Works (OTW), a nonprofit organization dedicated to preserving the history of fanworks and fan culture. As an archive, AO3 is not a social media platform, and it does not have an algorithm. Instead, it has a complex search engine that allows users to filter fics based on fandoms, characters, relationships, ratings, warnings, and other tags.

Zine - an amateur-created magazine. In fandom spaces, zines contain fan content like fanfiction or fanart that is made by and for fans of a particular fandom. 

Canon - a term adopted by fandom to mean the events and elements of the original media or text, exactly as they were published. 

Headcanon - an idea that a fan imagines to be true despite not being supported by evidence from the original media. 

Fanon - ideas introduced by the fandom that are not part of the original media but that are canon-compliant and widely treated as canon by members of the fandom community. An idea becomes fanon based on its longevity in the fandom. 

Blorbo (originally “Blorbo from my shows”) - your favorite character. The term comes from a tumblr post complaining about people constantly tagging their favorite characters on unrelated content. 

Poor little meow meow - a miserable, pathetic, problematic character, used as a term of endearment. 

Cinnamon Roll - a sweet, kindhearted character. The term comes from a 2014 article title from The Onion: “Beautiful Cinnamon Roll Too Good For This World, Too Pure.”

Shipping - a variation on the word “relationship,” this is the act of thinking that two characters would be good together in a romantic relationship and wanting them to get together.  

Ship - a desired relationship between two characters. 

Ship name - a portmanteau combination of the names of two characters in a ship. This could be first or last names, aliases, or occasionally a descriptive word/phrase. Examples of ship names are “Steddie” (Steve Harrington/Eddie Munson from “Stranger Things”), “Sleeping Warrior” (Aurora/Mulan in “Once Upon A Time”), and “Ineffable Husbands” (Aziraphale/Crowley from “Good Omens”).

Slash - a fanfiction in which characters of the same sex are in a romantic or sexual relationship. Slash often refers to two men, while femslash refers to two women. The term comes from the backslash symbol used to tag two characters as being in a romantic or sexual relationship, as in [Character 1]/[Character 2].

Anti - a person who is vocally against a particular ship and harasses people who create content for that ship. (Don’t do this.)

OTP (one true pairing) - your favorite ship. 

OT3 - your favorite polyamorous ship. If the polyamorous ship contains more than three characters, OT4, OT5, etc. are used. 

NOTP - pronounced “no-T-P,” this is a pairing or ship that you personally dislike. 

Plot armor - a character is said to have plot armor when they are too important to the story to die. 

DLDR (Don’t Like, Don’t Read) - a general piece of fandom advice that if you don’t like the content described in the tags, don’t read the fic, and especially don’t complain to the author that you don’t like the tagged material.

SALS (Ship And Let Ship) - A fandom rule that says not to harass people who like ships that you don’t like. 

YKINMKATO [Your Kink Is Not My Kink (And That’s Ok)] - A fandom rule that says not to dissuade or harass people who create kink content, or content more generally, that you don’t like. Sometimes referred to as “kinktomato” because of how the acronym looks. 

Beta reader - a proofreader or editor of a fanfic who is not the author of the fic.

If you want to learn more, fanlore.org is a great place to start!