Hacker News Reader

Making videos taking time away from actually making stuff and making the projects take 10X longer. So I stopped.

Check out Kenji Lopez-Alt [1]. He’s an award-winning chef who makes cooking videos by strapping a GoPro to his head and going to work in his home kitchen. He has basically none of the fancy production you see on cooking TV shows. Yet his videos are very popular because he’s a great chef and he tells you the why in addition to the what and how.

I’m pretty sure he’s made his setup just about as close to optimal as possible in terms of minimizing the time he spends on the video production part while still looking great. I think his one bit of fancy production is that he has a nice spot by the window to set a cooked dish for his thumbnail photograph. I think a bunch of his cooking videos also do double duty to supply photographs for his cookbooks, but that’s unnecessary for the vast majority of video creators.

[1] https://www.youtube.com/user/kenjialt

Kenji isn't really representative of the path most people can take.

His reputation is almost exclusively from his superb cooking articles (and recipes). The Youtube channel is a side dish, not the main course.

I said his videos are very popular because he’s a great chef. Nowhere did I imply that you can go from nothing to famous person by strapping a GoPro to your head. I mentioned Kenji because he’s an example of someone who makes great videos with absolute bare minimum production effort. The person I replied to talked about how video production had gotten in the way of the enjoyment of the craft. I think Kenji’s videos are a perfect example of how to avoid that problem. That doesn’t mean following this advice will lead to instant wealth and fame.

Ultimately, there is no formula for fame because audiences are fickle and trends are fleeting.

Do you reckon he uses the built-in microphone, or an external?

This channel is amazing. I learn best by simply watching others do, and this will surely help me advance my culinary art!

He definitely does not use an external mic. You can tell because of the reverb from his voice bouncing around the kitchen getting picked up on the camera’s internal mic. If he wore a lapel mic he wouldn’t have that problem!

But lapel mics cost money and they can be a pain to wear, especially when you’re cooking and need to keep your hands free without anything dangling into the food or the fire!

I've just watched over an hour of his videos, he's very good.

Apart from the good format I found myself thinking the audio is extremely good

Even with Kenji's minimal setup he still said recently he spends on average 6-8 hours per video on editing[0].

[0]: https://www.youtube.com/watch?v=NxOlI6QB0oI

I think he’s been trying to cut that down very recently. His latest video begins with him not even on camera! I think he’s testing the waters and he’ll find that even with the most rough, sloppy edit his videos will do really well.

I think DIY content is so hard to make. For me, it usually involves building stuff that I find interesting, but that is usually super niche and doesn’t appeal to a mainstream audience. Thus, less views which makes the whole time spent on making the video feel wasted compared to other projects I could be doing.

On the other hand, I could try to build projects that might be more mainstream for the better views, but then I’m spending time creating stuff I don’t want to make.

As a consoomer I find interesting that the production quality has bigger say over which DIY projects I watch than if the thing they are building aligns to my interest. For example JohnnyQ90 is filming himself creating very high quality RC cars from scratch, and despite me not having interest in machining/rc cars I still watch basically all his videos.

Your interest being niche may not matter as much as you may think.

That’s a good point! Though, I would still argue that things like RC cars are more mainstream then a video about building a RISC-V emulator or re-implementing an existing protocol.

You are right. Such concepts are truly abstract, such for illustrations are hard to make. The matches may not be perfect, but examples in this case come to mind:

[0]braintruffle - took the concept of computer fluid simulation as main topic over detailing specific implementation, few videos a year.

[1]Reducible - His explanation of GJK algo, even after reading papers on it, made me appreciate it.

Both utilise extensive animation skills, mastering meaning of video.

[0] https://www.youtube.com/watch?v=MXs_vkc8hpY

[1] https://www.youtube.com/watch?v=ajv46BSqcK4

People who are interested in DIY are interested in the process far more than the average person. They want to learn techniques on how to build the things they want.

You can't build a Youtube channel by highlighting what you've built. You have to explain how and why it's built the way it was. The end result is maybe 5% of the value.

> Making videos taking time away from actually making stuff and making the projects take 10X longer. So I stopped.

Isn't that why a lot of YouTubers (or content creators, whatever the name is), hire a videographer to do that work for them?

So a lot of YouTubers start primarily self editing up until they have enough of a following to hire someone’s services. It can be painful!

They do, but you need a large following and a decent amount of money for this to be affordable in the first place. I paid someone I know to edit some of my videos for a while, and even at a fairly low price, I would end up spending as much money on the editing as the video paid out to begin with. Same with thumbnails, a lot of big creators get someone else to make them, but the prices there are quite steep too. I used the same thumbnail designer as a few of my favourite 100K subs YouTubers, and it was like $200-300 a thumbnail.

So it's probably like, $1-2K per video if you want someone else to take care of the editing and thumbnail design aspects. Hence you'll likely have to do this stuff yourself for the few hundred thousand or so subs, at least if you don't have FAANG/trust fund money on the side.

If you have money laying around, sure. If not - and you aren't really making money - then no. And even then, lots of folks do their own editing.

It isn't all that different from other arts: If I paint, I might be lucky enough to pay a gallery, either through commission (often 40-60%) or through renting space (often, monthly payment plus a lower commission). You might be able to rent space. But most folks just have to do all the small stuff themselves. It isn't uncommon for artists to spend more time doing social media/advertising/mundane stuff than they do creating art, if you are trying to make money.

I'm not as intimate with the music scene, but it seems a lot of folks do similar stuff.

I wonder, what would come out, if all these editing things can also be automated and optimized by an AI. Leaving the video intact, but doing the cuts, translation, timestamps, colorizing, densing, audio gaining, lighting, so it all looks compared to things out there.

I really want to get in to video editing and posting on YouTube, but I just can’t come up with something to record as all I really know is programming which doesn’t make for great video content.

Can you educate me a bit here? If you do it in a single take, what do you do in your hour of editing?

import the video from your device. cut out/speed up the boring parts. add a voice over or music or corrections to dialog. add intro screens, cut screens, transitions, titles. tweak the audio, censor curse words if required. insert images or screenshots you reference. insert labels/graphics. export it. re-export it because you screwed something up. upload to youtube. fill out their required forms. write description, timestamps, link cards, end cards. create affiliate links if that's what you're doing. follow up on social media. etc. etc. etc.

you're probably thinking of streaming, which involves less work, but you have to do it in real time with OBS or whatever. maybe you're talented and can do it yourself, but sometimes that involves a second person at the controls.

sure you can just upload a raw video straight from your device. but this is exceedingly rare, because you've grown accustomed to at least the basic round of editing - you likely don't even realize what edits have happened. raw video is incredibly boring. will anyone watch it? if not, then what are you even doing that for?

you should try both sometime, it's not quite as easy or fast as it sounds.

I wonder how long it takes before someone makes an AI to do it.

Adobe Premiere, along with special purpose apps are already here. take the raw video, chop it up into pieces, grab the algorithimically determined "interesting" ones, stick a recent audio track behind it. volia! not as good as if you had a team helping you create content, but a whole helluva lot easier.

Thinking a bit more about it, I guess you could have an algorithm that detects the start of a fragment based on speech (e.g. you say "fragment start!") and you could rank the fragments at the end, e.g. "fragment stop score 8!" I suppose you could do that with open source speech-to-text tools. And use ffmpeg to cut and stitch everything together.

This exist, there's a software that will trim any non-talking parts of your video with AI as a smart video editor

Can you share more about these services please?

Editing is boring work because you have to go through hours of video and find the right fragments and stitch them together, then rewatch to see if the flow is right, etc.

Exactly, selecting what to put in the video is incredibly time consuming if you have a lot of footage. That's not evening mentioning editing. Every time I watch a section of video I would find things that can be improved. At a certain point, you need to stop polishing and just finish it.

Well, you can't build much in the ~10 minutes of a typical youtube video.

So presumably by "one take" hahamrfunnyguy means he:

* Stopped recording driving to pick up materials, placing orders where there's a wait for parts to come in, and suchlike.

* Stopped recording things speculatively, just in case he decides to make a video later.

* Stopped worrying about 'narratives' and 'thumbnails' and 'the algorithm' when working.

* Stopped cleaning off his hands and fussing with camera settings to get the perfect shot of every step on every machine.

* Stopped trying to make point-of-view videos and keeping an eye on the camera screen, so he didn't need the tripod between him and the workpiece.

* Stopped retaking any time he messed something up or flubbed his lines or the audio came out bad.

* Stopped worrying about set dressing like having a picturesque, tidy background for every shot

Then the editing process is just editing several hours of video down to a few minutes and maybe recording a voiceover.

Pretty good summary of all the stuff that goes into a DIY build videos. Now I just show the finished product and talk about some of the design details and problems solved along the way. I think there's a lot of value in this sort of content, but I haven't made enough of them to know if it's something people actually want to watch.

Can't speak for others, but I have several channels I follow where this is effectively the format... It's a bit harder at times where it's just a tripod in the corner, and might be more effective with a headband mounted gopro depending on your work. But other than that, just appreciate enough instruction and video to understand how things are being worked on, usually fast forwarding (in the edit) through the more laborious (time consuming) parts.

But I follow such a wide variety of topics from political, to technical, to PC gaming to food (keto, general, other) and home construction or even electronics DIY stuff. I think YouTube has trouble figuring me out... I tend to get fed too much of one thing over time, and have to go into my subscribe list (a few hundred) and start watching from the more recent, or specific channels depending on what I'm in the mood for.