Hacker News Reader

Comment by olafura | Hacker News Reader

Refreshingly honest about the video quality you should expect in the main promo video. Typical problems like dynamic range, color profile, light bleed and tracking problems with the top of the head being cut off at one point. Pretty typical unless you spend more money on your drone or camera equipment. There are so many companies that fake the footage or pick much easier shooting locations.

View on HN

I think this is going to be one of those things where we learn that nobody actually cares about any of that outside these forums. MP3s were garbage, early camera phones were garbage, Snap picture quality generally is garbage, bluetooth headphones suck, all of these things were and are decried by the tech community, and it turns out that people just don't care, because even at garbage quality it's still cool to have a robot take your picture.

Having a robot taking your picture only sounds cool go hacker news people ?

The comment you are replying to claims the opposite of what you interpreted it as.

The cool thing isn't "having a robot take your picture" the cool thing is "having a picture taken of you".

Even without the robot angle "cool" aspect of a drone, it's a step along a progression that's shown clear demand: first selfies, then selfie sticks, then no-stick-needed non-selfie selfies.

Drone photography predates selfie sticks.

No, it doesn’t.

https://www.theguardian.com/uk-news/2014/dec/23/photograph-1...

If you're allowed to use that as an example of a selfie stick, then I should be allowed to use James Wallace Black’s 1860 aerial photography. :-)

https://time.com/longform/aerial-photography-drones-history/

Or maybe the pigeoncam from 1907.

https://www.newyorker.com/culture/photo-booth/the-turn-of-th...

It seems much more likely that the stick in this picture was used to press the shutter release button of a camera on a tripod than to suspend the camera itself.

In terms of invention, sure, but definitely not in terms of wide consumer availability/adoption.

Except that mp3s sucked until the iPod, and camera phones sucked until the iPhone. The idea that product quality, ease of use, UX/UI, or polish doesn't matter is just wrong. This thing is absolutely going to bomb (reminds me of GoPro's Karma) and I'd bet money on it, but SNAP stock is already in dire straits, so their incompetence is probably priced in.

Camera phones sucked until iPhone 4 at least.

No, camera phones were very good since the Sony Ericsson days - ~2006.

I have a book from that era all about mobile camera photography, written by a professional photographer, and illustrated only with pictures shot on a Nokia phone. The pictures are of excellent printable quality with amazing colors.

~nobody used the Sony Ericsson devices though, everyone had stuff like the Motorola Razr with terrible little cameras in them.

I had SE k800i in 2006. It had a Xenon flash and the photos it took we not surpassed by smartphones for a surprisingly long time, especially at night.

It was extremely popular brand in Europe for many years. At one time almost everybody had a good Nokia or Sony Ericsson. I saw Motorola Razr like one time in the wild.

Well, I speak from an American perspective and the Razr was everywhere, but according to Wikipedia their peak was around 10% market share. Nokia was much much more popular.

America was way behind at that point.

IIRC a combination of weird regional roaming agreements, odd incompatible networks and the sheer size of the country leading to lots of coverages blackspots meant that the US just didn't get widespread adoption of decent mobile tech.

Yeah, that makes the iPhone hype much more sensible. I remember it as very much underwhelming, I had a phone like that 2 years before it came out, and it had apps.

Can I ask what the name of the book is?

https://www.megaknihy.cz/foto-video/22886-fotime-mobilem.htm...

Can't find it at home to provide some scans, sorry :-/

My Sony K750i from 5 years earlier was phenomenal. Only 2mp resolution but the lens was brilliant and I've still got a bunch of photos I printed from it that still stand up to scrutiny.

I think at that period the US was way behind the rest of the world on mobile tech. I remember people in a phone shop in New York being really impressed with it.

I mean who cares about the stock, they have 5B in annual revenue, 58% gross margin, they’re cash flow positive, and spent 1.5B in R&D last year. Snapchat and Discord are the messaging apps.

By basically every measure they’re extremely successful and following the Google strat of their main advertising business finding moonshots.

> Except that mp3s sucked until the iPod

Just being able to download music from the internet was already a game changer, it didn't suck! Before that if you wanted to hear a song from an album you didn't have at hand and you wanted it RIGHT NOW, you had to call into a radio station and ask nicely for another human being you've never met to play it for the entire city.

Waiting 60 minutes to download "Stairway to Heaven - LIVE rip" by "Guns N Roses and Bob Marley and U2 and Green Day" was a privilege.

I’m arguing the opposite of what you think I am. MP3s won because you could download and throw a gazillion of them on a portable device, cell phone cameras won because they were always on you. If this thing wins it’ll be because it’s cheap, you can swap the battery fast enough to make up for the shitty battery life, and the ux is simple enough that people who don’t care about drones can get drone shots. I’m completely on your side - the technical specs don’t matter if the UX is right.

> Except that mp3s sucked until the iPod, and camera phones sucked until the iPhone.

What is this, alternate history?

I do some extreme sports for hobbies, where I am often the only person with the sense of mind to take shots with my trusty olympus TG2 (yes it's old, but still works) of others with me, whether underwater or 6Km above. My professional photographer friend often berates me for my framing, perspective, lighting choices etc, but when I'm hanging almost upside down and only using one hand covered by a mitt, whilst nobody else has the energy to pose or even think of stopping, I think there's only so much you can do.

Anyway, the point is, my friends have lots of pictures of themselves doing these activities, versus non at all.

Brain is an associative data store. You have lot's of stuff stored there, but you can't recall things without suitable key. Even a technically lousy photo can work as the key that bringz back the memories.

The best camera is the one you have with you.

I have "terrible" pictures of all sorts of things but what's important are the memories attached to them.

I was very surprised to hear the rotor buzz. Most drone promo videos pretend that when the motors spin, it makes moody beats sound.

I was too busy being deafened by Frank Sinatra a split second after the rotor noise played to even remember hearing the rotor noise; rather than faking the silence they just over-wrote my memory of it in the advertisement with shock-and-awe style Sinatra -- I guess that's okay.

They have plenty of videos, here is one without any music: https://www.youtube.com/watch?v=qpz8Q2spioo

I was positively surprised, despite the music. Some time ago I was very annoyed by all these "drone projects" on kickstarter or concept artist sites about drones following you all day as personal assistants, for example. IMHO, drone noise is just as limiting factor as the battery life and products should have it on their promotion materials, even if it is for a split second.

And that's pretty much the idea - they expect you to put music over whatever you've captured anyway.

But while you're doing whatever it is that you're capturing, whether or not there is a loud humming noise the entire time can completely change the dynamic, especially if other people are around.

That is nice, but there is still intentionally misleading marketing fluff like this:

>Each rechargeable battery allows you to capture content for 5-8 flights, depending on the flight mode(s) selected. Rechargeable batteries can be swapped for easy use...

Why phrase it like that? That gives zero impression of what the actual battery life is because I have no idea what a "flight" means as a unit of measurement. Just tell us how much time it lasts.

I have no idea on how many minutes it takes to get a photo. I personally like their metric choice.

I'm not sure that I follow what you are saying. Are you suggesting the expectation is that each flight yields one photo? It takes milliseconds not minutes to take a photo. I would expect this thing to be taking photos nearly continuously during flight which would make flight time the most informative metric.

I don't think you're their target demographic.

Asking for clarification on battery life means I'm not their target demo? That is rather condescending to their customers. Battery life seems like a basic thing someone might want to know before spending a few hundred bucks on something like this.

>Asking for clarification on battery life means I'm not their target demo?

I agree that it's condescending, but it's also obvious to me that there is a growing market trend to sell to a class of people that simply buy the Next Big Thing without ever learning about it, using it, or questioning the motivations behind it.

In a way that's a great group to capitalize on, they may lack the education or desire to ask the tough questions, they don't actually use the product so you don't really need to support them well, and they tend to buy oriented strongly with advertisement.

When I see a technical product marketed this way -- bright colors, flashy every-person advertisement, zero-training-required, no real technical specs -- I always imagine that it's in that class of product. Sort of the opposite of 'prosumer' classed product.

A titch bit more useful than a Funko Pop doll until The Next Next Big Thing is released by BigCo , the firmware gets dated, the batteries die and remain unreplacable due to a lack of support for The Old Big Thing, and then BigCo drops software support due to trying to shove people into the new model, and then it becomes under-bed trash with a dangerous LiPo in it.

> Asking for clarification on battery life means I'm not their target demo?

Precisely

What is this supposed to mean?

"Flights" is a bad metric if they don't define it. Whatever your age, gender, income, it doesn't change that it's meaningless.

Each flight is a snapchat story appropriate sub-minute interval, like 15-40 seconds sort of thing. I really don't want to sound condescending, but if that isn't immediately obvious then you aren't the target audience.

I’m not a Snapchat user and I just looked through the videos on my phone and I have tons that are in the 15 to 25s range. In fact I have hardly any that are longer than 30s. Short segments of drone video from a durable device you can throw in a bag seem fun and useful.

Their target market measures the value of this thing in number of Snapchat Stories it can create so really the flights does make more sense. Their target market isn't going to want to divide 3 minutes by 20 seconds.

They're probably too ashamed of the actual value in minutes...

Reminds me of some internal corporate presentations: when the numbers are good they show percentage increase and when they're bad the show absolute values, and when they're really bad they "forget" to give the time range of that absolute value...

I fly FPV drones, I have one very similar to that size. running a 1s 300mah lihv I get about 3 to 5 minutes. 5 minutes damages the battery. but the tiny lihv never liv long enough anyways.

I'd expect it to get close to that time, but it has so much more plastic around it. but then again it's not transmitting video.

That matches with the estimate in another comment chain[1]. That is low enough that listing it will likely be off-putting to many potential customers. It is why I think it is misleading to try to hide that number with a more opaque metric.

[1] - https://news.ycombinator.com/item?id=31195956

Can you recommend a good starter FPV drone?

The standard advice is: Buy a transmitter and practise for like 5-10 hours in the simulator before pulling the trigger on a real one. Radiomaster Zorro, Jumper T-Pro, TBS Mambo are some good ones. Go directly for ExpressLRS/TBS Tracer long range protocols so you wont have to worry about your quad falling out of sky because of range issues.

If you want something that is more "ready to fly" (goggles, quadcopter, transmitter combo) You do have more choices these days:

1) On the more expensive end - You have DJI FPV combo [1]. Is supposedly very beginner friendly. https://store.dji.com/product/dji-fpv?vid=101601

2) On the low end - Something like this https://www.getfpv.com/emax-tinyhawk-ii-freestyle-micro-brus...

The problem with these ready to fly combos - especially on the low end (analog video) is that, you end up getting very toy like transmitter and goggles, that wont be of much use for you once you want to buy/build more quadcopters. So I highly recommend getting the transmitter and goggles separately.

Commenting I like the bicopter design has advanced features like follow user mode and long flight time

Thanks! My preference would be to start really cheap just to see if I like the hobby and want to stick with it. I don’t think I’d mind buying a better transmitter later.

So how does this drone stack against the competition? I know nothing about this field and I expected a price around a 100 mark. 250 sounds like a lot.

I don’t fault them for this specifically. IIRC the 30 minute airtime standard is some kind of FAA regulation; it’s not a differentiator one way or another between drone products.

This drone will definitely get nowhere near 30 minutes.

The default flight is 30 seconds.

If you're using this for Snap, one "flight" is actually a pretty reasonable time unit. Snapchat videos can be a maximum of one minute.

So far as I can tell there are no real spec numbers anywhere. No video size/framerate/resolultion specs I could find (admittedly I only spent 1 or 2 mins clicking around looking for them).

Which kinda indicates the target market they're gonna aim this at. This is for people who want to know "Can I upload the video to Snapchat?" not for people who are wondering "Is this 1080p? Or 720? Surely it can't be 4k?"

I prefer sample videos to specs, because specs are pretty much worthless. The 2021 Macbook Pro has a 1080p camera, but the actual quality is worse than the 720p camera in the 2015 Macbook Pro.

The website has some sample videos, and you can tell the quality is very poor compared to pro drones, but definitely good enough to share with friends and family.

There are if you follow "Pixie Support" link in the footer, here are the specs you were looking for:

Images Standard: 4000 x 3000 (12MP)

Video Default: 2.7k Frame rate: Up to 30 FPS

https://support.pixy.com/hc/en-us/articles/5039928089236-Pix...

wobbly:

https://videos.ctfassets.net/svn43w404u4n/4oG6RAF5QALMWEBdlX...

not just wobbly but what small drone enthusiasts call "jello", significantly lower video quality than like a $399 dji mini 2

it is cheap and non gimbal stabilized, so can't expect much more

Are there no better digital stabilization algorithms? This seems like it could be fixed in post processing on your phone/cloud.

there absolutely are, without a gimbal, by using a somewhat large sensor and movement/orientation/inertia sensor ICs to perform a software crop and stabilization as the video is recorded.

such as the hypersmooth 2.0 or hypersmooth 3.0 on a gopro hero9 black/hero10 etc. but those are considerably more expensive cameras.

search youtube for gopro hypersmooth 3.0 and you'll see some examples of the same gopro with and without hypersmooth turned on.

I've used Mercalli to fix this in post-processing but it's not a simple fix.

This is caused by a non-global shutter so you have to spend a lot of processing power to fix it.

A recent, but really good, open source one is https://gyroflow.xyz/ . There's also Reelsteady, Hyperflow etc..

I mean yeah, it's about half the price of a DJI Mini 2 ($449 v $230). You'd expect it to perform worse. It seems totally reasonable for people who just want to take some shots for their snapchat story but aren't really interested in having a real drone.

The Mini SE is readily available for $299 and I don't think even the Snap demographic is going to find that a camera with no gimbal is acceptable for the application.

I'm not predicting whether it will sell or not, I'm bad at that, but it's just not a lot of drone for the price.

Where is that video from?

sauce on this succubus?

Why isn't it trivial to cut out that noise?

is it 4 motors? especially with data about current RPMs of the motors, with a FFT and some adjustments, it should be pretty trivial to deafen the sound.

From my listening many motors are effectively pretty simple instruments with normal harmonics etc which are easy to model. I could be wrong though.

There's going to be a significant white noise component due to the turbulent airflow, which won't be as easy to cut out.

There's so much white noise outdoors that removing the dominant harmonics that make drones, well, drone,† would make a huge difference in the palatability of being around them.

†The noise quadcopters make has nothing to do with why they're called drones, but is the perfect description nonetheless for the noise they emit

If they combined it with audio from the phone they would have a much better chance of isolating noise.

The image stabilization is garbage

> video quality you should expect

The video that they try to show as if out of the drone is heavily edited to make it look similar to the other camera that they have used. It is very likely that they used non-stock camera software in the drone to capture raw and post-process for this ad, which will most likely not be available to the public.

There is probably a market for pocket drone that can carry iPhone and take selfie just like this.

I'd be pretty worried about an accident that damages my phone; I don't think I'd want something like that. But I wouldn't be surprised if some people do.

Isn't iPhone too heavy for drone to carry around photographing?

Not that I think it should be done, but since the drone is carrying the phone, you could have the phone save a GPS point of where you are standing, and then the phone guides the drone away from that point to photograph toward it.

Depending on GPS accuracy in the phone it could work well. GPS waypoint missions are already well established in the drone world.

an actual FPV flight hobbyist 5" prop size class quadcopter (such as you might fly with goggles and a real remote control) can typically carry a gopro hero8/9/10 but a gopro is also a lot more compact, brick shaped, rugged, and has very extensively researched video stabilization systems in gopro's proprietary "hypersmooth" which writes the stabilized video into the HEVC file it records.

putting something flat and large like an iphone on a 5" FPV quadcopter would be awkward and bad.

iPhone 12 mini weighs 135g, over half the 250g limit for mini-drones. It seems unfeasible for a pocket-size drone in the near future, unless the drone could pull power from the iPhone and not need its own battery.

Mavic Mini has a payload capacity around 180g, but it also has a nice camera already.

Drones need power FAST for short bursts (aka one flight). I don’t know if the iPhone battery supports that usage. It would be cool though if the phone could be a truly universal tool like that.

The pocket drone needs a way to know where you are in order to direct the camera at you and come back at the end. Right now they do so by pairing the drone with your phone (I guess using Bluetooth).

Tracking via Bluetooth would be much less trivial than "Ok, I'll fly toward that smiling face until I see a hand in bottom camera, then I'll just start descending until it looks close enough to shut down my motors".

It also makes it seem more real and fun

Reply on news.ycombinator.com