Hacker News Reader

Comment by OliverM | Hacker News Reader

I was pretty excited by this, so I went to pre-order. On the add-ons page, there are some nice additions like a leap motion controller. But there's also this:

    2D-to-3D Conversion: 1000 Credits
    Convert 2D images into 3D depth images for holographic display.
    +$150.00

    2D-to-3D Conversion: 100 Credits
    Convert 2D images into 3D depth images for holographic display.
    +$20.00

There's no mention of this type of per-interaction charging model on the product's splash page.

Instead of excited I'm now disillusioned and irritated.

Imagine Adobe tried to charge you every time you converted an image format in photoshop. Is the conversion process so complex that you need to upload a sample image to their servers or something? Why on earth can't I do that locally as many times as I like? It totally stymies experimentation.

View on HN

There's a bit more info about the process here: https://lookingglassfactory.com/3d-photos

It is a cloud-based service, Owl3D. They use AI to deduce depth info from photos. You can subscribe to Owl3D directly, for $8 a month, and get 300 uploads per day for that (so, 9,000 a month for $8, vs 20 for $100 Looking Glass). I don't know if that direct Owl3D subscription can take advantage of depth info though: https://www.owl3d.ai/faqs

I mean, it's a 3D display. You need 3D things to use it. This is just a convenient service for converting your own 2D images.

You can always bring your own 3D models/videos/etc into it, you don't need to pay them to do that. I personally don't see anything wrong with this.

There's no mention anywhere on their product page that you'll need to pay additional fees to convert 2d images to 3d to display them, or that you can perform this service yourself with a bit less convenience (which I'd be perfectly happy with).

Why would they mention things they don't sell on the product page? Do you expect Photoshop to mention "oh you could use something cheaper made by someone else"?

I don't expect them to mention that?

They make the types of things that can be displayed on it pretty clear on their “Use Cases” page, IMO. I would not expect a 2D image to inherently show up on a 3D display; otherwise, what would the point be?

Maybe your mindset is more indicative of the general population, but I do think most people understand that 3D holograms and 2D photos are different and would not expect a photo to suddenly have 3D properties on the display as if by magic (perhaps that is hoping too much though).

I don't have the expectation that putting 2D photos onto a 3D would magically work. I expect to need to convert them (it's a bit much to assume that I would think otherwise).

I don't expect to have to buy credits to convert them. There's no mention that I need to, until you get to the extras page when you go to purchase. That changes the entire set of assumptions I had (from "I can play with this and use it to explore whatever I find interesting", to "I have to pay to use it in a way they dictate").

Since then I've found that you probably don't need to do that. But that's not the impression that 'extra' gives. They don't mention anywhere that you don't need the credits to do the conversion.

You can make your own 2D->3D conversion they’re charging to use their service to do it.

I think our difference is that you expected to be able to use 2D photos on this device at all. There is no third-dimensional information in a photo and to me it is amazing that there is a way to make them 3D. I never had the same thought you did because if something was not already 3D I would not have thought I could use it with this anyway. Perhaps this is why I am not bothered by them charging for the feature.

I bet it's a NN that they've trained to make their image quilts based on photos and it may be hard to ship to the end user. That being said SaaS is a cancer upon humanity.

> it may be hard to ship to the end user

just upload it to some publicly accessible website? It doesn't look that hard.

Ship in this case doesn't mean "deliver an asset" but rather "deliver a user experience", nailing the latter consistently is non trivial. I do think people making SaaS products start by wanting to charge rent and then work backwards from there to a justification most of the time.

This is only for converting 2d photos to 3d. If you have photos shot using portrait mode on Iphone/Galaxy/Pixel, they have a local tool that extracts depth info. AFAICT that is free, although it requires at least a GTX1650 for some reason.

From the Minimum Requirements on the product page:

HoloPlay Studio works with PC and Mac.

To import, edit, process and convert depth photos, videos, and other holographic media, almost all PCs and Macs with HDMI and USB-C will work.

To run and create heavier-duty interactive applications with Unity or Unreal, we recommend a PC with a GTX 1650 or equivalent, CPU with Intel i5 or higher, and at least 8GB of RAM.

The iPhone captures depth when you take the photo. So the data just needs to be extracted, not generated from scratch.

Yeah that's why I can't understand why they need a dedicated GPU.

See jaywalk's reply to you from earlier: https://news.ycombinator.com/item?id=27982727

It sounds like the discrete GPU is only needed for specific things which seem like they could reasonably need a discrete GPU

(We're both talking about the "Minimum Requirements" section on https://lookingglassfactory.com/product/portrait, in the bottom right of the page below the big blue buy button)

Reply on news.ycombinator.com