Showreel ad V1

[Nate North]

Just looking for the standard feedback on this one, Too short, too long, Can't hear the voice over the music, Don't understand the product, Can't figure out how to use the site from this, etc.

This is my first attempt at an ad for https://showreel.video/

I wasn't too sure about adding on the high speed reel at the end, But I felt like just the site overview by itself wasn't quite flashy enough for an ad.

Anyone have any thoughts?

 

[Jack Lynt]

Wow, looks amazing. I hope SHOWREEL is a great success.

I think it's a pretty excellent ad. Music that's too loud, too loud to hear the voice, or the presence of music at all is often and very much a dislike of mine. YouTubers have been vexing me a lot with those very things. There was a brief moment when this mix crossed or almost crossed into that territory for me, but it was fleeting, and I think it has a fairly good balance overall. It helps that I happen to like the music you chose. I'm also listening with fairly decent headphones via my desktop, while I suppose many viewers will be on their smartphones, no doubt a significantly different experience.

I also think the high speed reel at the end is pretty great. I like it. Where I think I might like to see the reel move by slower, the clips to be longer, is before that, the stuff you have edited in with the tutorial, then we get the rapid fire reel at the end. But who knows?

The tutorial seems nice and well thought out. I bet you have more tutorials available on SHOWREEL, as well. Even being a low IQ and non-techy person, I got the impression this could be comfortably doable for me. Good stuff.

 

[Nate North]

Thanks Jack, I appreciate the review. It's genuinely helpful to get the raw impressions of different people. A lot of the issues people have hearing narration over music come down to each person having a slightly different EQ curve to their hearing, so it's one of those things where it's literally impossible to get it right without feedback from multiple people. Glad you could hear it ok, and that it didn't come off as overcomplicated. I made a real effort to keep it streamlined and simple.

 

[sfoster]

Just looking for the standard feedback on this one, Too short, too long, Can't hear the voice over the music, Don't understand the product, Can't figure out how to use the site from this, etc.

This is my first attempt at an ad for https://showreel.video/

I wasn't too sure about adding on the high speed reel at the end, But I felt like just the site overview by itself wasn't quite flashy enough for an ad.

Anyone have any thoughts?

I thought it was too long - it's like a mix of an advertisement and a tutorial.

 

[directorik]

I'm a filmmaker who will likely not use this.

From that perspective it's intriguing. It makes me want to give it a try.
The clips at the end confused me a little; confused might no the the
right word, but it made me focus on your built-in library either then
what I could do with this. Great library but anything in there I would
use in my own projects.

That's why I think this isn't for me. But I'm guessing there is a need for
this among other filmmakers.

It looks great and user friendly.

 

[Nate North]

My initial plan for this didn't include a library at all and I'm still wondering if it's a good idea or not. I began integrating the library when I tried to beta test it and ran into what I would call a shocking number of people that didn't have any footage to put into it. When I first announced it here, my mental picture was of people grabbing an old folder from one of their many shoots or projects, Quickly tossing it into the interface without much Thought and whacking the button to see what happened. A few minutes later they have a fun music video With a 100 different random moments from a film library they haven't looked at in years. Something like that.

I talked to different random people to see if I could get a little bit of help beta testing it, People that had called me up the previous year and the year before that to explain how important it was to buy a new 1500 dollar telephone that had the 8K camera instead of the utterly obsolete 7.5 K camera. Not one single person I talked to in two weeks seem to have a folder full of old video clips, or new ones. It's a little weird to me because I think I have maybe 70,000 video clips, plus an immense library of movies and television that I've purchased over the decades.

So in an effort to just try to get this thing pushed through the testing phase and make sure it had enough beta testing to risk public release, I made it so incredibly easy to use that literally anyone could use it In 60 seconds flat with absolutely no requirements. That was how the built in libraries came to be.

I'd say there are a very large number Of what I would call casual filmmakers, specifically on Youtube, that would have a use for this particular tool, But it's probably not for the same group of people that buy a set of Cooke S4 Prime lenses. It's a great lightweight solution for rapid iteration Of showreels, montages, background footage For narration driven videos, documentary sections, and what will probably become a big market section, people making AI video art that don't have a lot of editing experience. In my experience there has literally never been a faster and easier way to handle this niche task.

Think of it like an unfunded company that's trying to build an aircraft carrier but can't get paid until the first one is finished. That company easily has the infrastructure to push out a line of 250 CC dirt bikes. A lot more people use 250CC dirt bikes than use aircraft carriers and the launch cost for that line is obviously dramatically lower. We're both professional film makers, But I'd say we're easily outnumbered 10,000 to 1 by people just looking to make an easy YouTube video for some purpose. It's vastly underpowered for what I need to do but for a lot of the people in the crowd it's just perfect. No more spending 6 hours editing together 10 minutes of random video game footage To support a Voiceover review, That kind of thing. In short this little spinoff product is not for serious filmmaking, But really excels at a number of types of Lightweight use cases. The Gopro of video editors.

I still may remove the library section, But I'll need a little more feedback than 3 or 4 people trying it out. More like 5000. I'll run an ad campaign soon and see what happens.

 

[sfoster]

i tried to use the built-in library

I searched: Magic, Wizard, Mountain, Lake, Dragon, etc
I couldn't find anything that fit my mood

 

[Nate North]

i tried to use the built-in library

I searched: Magic, Wizard, Mountain, Lake, Dragon, etc
I couldn't find anything that fit my mood

There is a very simple reason for that. I still have to build the search labeler robot. It's been on the list for a week now and I am getting around to it, But right now The search functionality is very bad. I have a way to make all this work really well it's just going to take me a few days to get it implemented. I've spent the last 30 hours of work on your last request, Throw in any random footage in any format from any time. And make it all work perfectly. That's....... Taking a second but I've made a lot of progress. At the same time that I'm doing that I'm handling another needed goal which is Input clip audio synchronization. I'm sure you noticed in the test clips you made that the audio from the input films got out of sync pretty quickly. that turned out not to be one problem but a stack of maybe fifteen problems in the code, And I'm just getting that fixed right now.

Anyway I'll get the proper search titling automated and implemented very soon and then when you search for a dragon you will get a page of Dragons. Right now all the footage of Dragons and castles is probably named something like "alpha batch 53487 cine-widescreen gen-7 clip7449-r" Which is difficult to find in a search.

I'd also note that kind of a monumental task to cover all the bases in the world in a stock footage library. I have a lot of experience in this kind of thing going back about 25 years, And the way the math works, It's going to require a staggering number of library clips before a person can type in something like " Yellow motorcycle" And get enough clips to have some genuine creative freedom in making a 3 minute video.

Let's say one in every 100 stock clips is a motorcycle. Then one in every 10 of those motorcycles is yellow. Some are parked some are jumping But about half of the clips are just the motorcycle zooming down the road. In your case you just need the most common and available one where they are driving down the road. Let's say that accounts for half of the yellow motorcycle clips total. So now we have one in every 2000 clips that fits your needs.

You wanna make a fast paced 3 minute video Of yellow motorcycle driving footage. Maybe we need 200 clips. so for that to happen the stock library would need to have approximately 400,000 clips total.

So far I've been adding about 1000 a day. It will be difficult to sustain that rate. There are a few different scenarios where it's all possible though.

But let's add in one more multiplier. You want this to be your video so you don't just want it to be made of every clip of a yellow motorcycle driving, You want it to be the specific clips that you pick out to express yourself artistically. Enough latitude to make it feel like something you designed yourself. So at a minimum Let's say three times that many clips and now we're at 1.2 million as a clip library minimum. Lots of people have pulled this off and I probably will be able to as well, But not on the first week with zero income and zero contributors.

 

[sfoster]

There is a very simple reason for that. I still have to build the search labeler robot. It's been on the list for a week now and I am getting around to it, But right now The search functionality is very bad. I have a way to make all this work really well it's just going to take me a few days to get it implemented. I've spent the last 30 hours of work on your last request, Throw in any random footage in any format from any time. And make it all work perfectly. That's....... Taking a second but I've made a lot of progress. At the same time that I'm doing that I'm handling another needed goal which is Input clip audio synchronization. I'm sure you noticed in the test clips you made that the audio from the input films got out of sync pretty quickly. that turned out not to be one problem but a stack of maybe fifteen problems in the code, And I'm just getting that fixed right now.

Anyway I'll get the proper search titling automated and implemented very soon and then when you search for a dragon you will get a page of Dragons. Right now all the footage of Dragons and castles is probably named something like "alpha batch 53487 cine-widescreen gen-7 clip7449-r" Which is difficult to find in a search.

I'd also note that kind of a monumental task to cover all the bases in the world in a stock footage library. I have a lot of experience in this kind of thing going back about 25 years, And the way the math works, It's going to require a staggering number of library clips before a person can type in something like " Yellow motorcycle" And get enough clips to have some genuine creative freedom in making a 3 minute video.

Let's say one in every 100 stock clips is a motorcycle. Then one in every 10 of those motorcycles is yellow. Some are parked some are jumping But about half of the clips are just the motorcycle zooming down the road. In your case you just need the most common and available one where they are driving down the road. Let's say that accounts for half of the yellow motorcycle clips total. So now we have one in every 2000 clips that fits your needs.

You wanna make a fast paced 3 minute video Of yellow motorcycle driving footage. Maybe we need 200 clips. so for that to happen the stock library would need to have approximately 400,000 clips total.

So far I've been adding about 1000 a day. It will be difficult to sustain that rate. There are a few different scenarios where it's all possible though.

But let's add in one more multiplier. You want this to be your video so you don't just want it to be made of every clip of a yellow motorcycle driving, You want it to be the specific clips that you pick out to express yourself artistically. Enough latitude to make it feel like something you designed yourself. So at a minimum Let's say three times that many clips and now we're at 1.2 million as a clip library minimum. Lots of people have pulled this off and I probably will be able to as well, But not on the first week with zero income and zero contributors.

Yeah it's quite a lot!! Personally I'd rather use my own footage, looking forward to creating some action music videos once that's working.

 

[Nate North]

It's on the way i've been working on it a lot the last few days and the results are actually really cool and fun. I've solved a lot of problems already and the audio sync is really close to perfect But just gradually slides off the mark over about two or three minutes to where there's maybe a quarter second delay three minutes in. It's just taking a lot of refactoring to get things like built in upscaling and reformatting To work perfectly with the down to the millisecond audio matching. As of yesterday I had the error rate down so low That we're literally talking about one 1 thousandth of a second off, But the problem is that it stacks per operation, so 700 operations in and you're almost a full second off which is super noticeable when people are talking or a car crash happens etc.

 

[sfoster]

It's on the way i've been working on it a lot the last few days and the results are actually really cool and fun. I've solved a lot of problems already and the audio sync is really close to perfect But just gradually slides off the mark over about two or three minutes to where there's maybe a quarter second delay three minutes in. It's just taking a lot of refactoring to get things like built in upscaling and reformatting To work perfectly with the down to the millisecond audio matching. As of yesterday I had the error rate down so low That we're literally talking about one 1 thousandth of a second off, But the problem is that it stacks per operation, so 700 operations in and you're almost a full second off which is super noticeable when people are talking or a car crash happens etc.

ah yeah that could b esomething like 23.976 FPS being interpreted at 24FPS.
I dealt with something like that once in resolve

 

[Nate North]

I'm just going to leave this as a note to myself so I don't forget to do it months down the line when I've completely forgotten this discussion.

But regarding your question, Once I have time to do it I can simply keep a database on the site that tracks statistics on what people searched for and how often they were able to find it. Once I know what people are looking for and not finding, I can build a backend process that automates creation of that exact missing footage, prioritized by request amount, And then adds that to the library dynamically without me having to do much.

I could do it right now but..... I've got somewhere between 50 and 100 outstanding goals like that and any one of them could consume between a day and three months of effort, Making prioritization a very difficult game since I don't know how long each goal will take. Sometimes things that seem like they would take an extremely long time and a lot of work get done in a week, And sometimes I take what I think will be a two hour side trip Like getting the audio lined up and it ends up stopping all progress dead in its tracks for two weeks before it's finished.

 

[Jack Lynt]

I have no idea. Does SHOWREEL need a marketing budget, like a movie or a game? And I suppose the AI tool market is already crowded and getting more crowded. Maybe something like YouTube ads?

 

[Nate North]

I have no idea. Does SHOWREEL need a marketing budget, like a movie or a game? And I suppose the AI tool market is already crowded and getting more crowded. Maybe something like YouTube ads?

Yeah that's exactly right. Organic business growth on the modern Internet Lies somewhere between myth and legend. People will get lucky here and there for a number of reasons, But for the large majority of companies that luck is bought and paid for via marketing.

I'm already working on a Youtube ads campaign for this, And I have significant experience running similar campaigns for companies in the past, But I'm horrifically underfunded, So I need a good bit of luck for this to work.

If you keep a really tight rein on your ad spend and hypertarget, you can get some real impact out of a relatively small ad investment, so I'm going to combine that approach with whatever inventive guerrilla marketing techniques I can come up with. It's a product that would be a big hit with the right kind of customer, and word would get around once saturation hit a certain threshold. It's an incredible convenience for certain niche tasks and the kind of tool where you wouldn't want to go back once you'd been using it for a while. If I can get even a small customer base reliant on it it will be enough to drip feed marketing and grow the company. I used to do all this work manually, and I'd be up all night cutting together a three minute reel that flowed well with a music track. This thing will do that for a dollar in 10 minutes, And that's version 1.0.

Most significantly, I have already developed technology far beyond what's available here, So if I can get over the hump of introducing it to the market in its most basic form it can grow into something far more powerful. Right now my strategy is to keep the learning curve or barrier for entry as low as possible, Then slowly introduce more powerful features after people are already familiar with the basic concept and interface.

As far as competition in the AI marketplace, there's tons of it right now, and it's all moving very fast and unpredictably. I have a strong advantage at particular moment because there isn't a single fully automatic video editor out there, So I've got pole position if I can make the world's cheapest ad campaign work out. No small task in today's rather hostile ad market.