Does anyone here own a Mosaic 51 camera?

I just saw a post on LinkedIn showing the Mosaic 51 camera. Here’s the website for it: https://www.mosaic51.com/

Seems like the project is by Jeffery Martin (360 Cities founder) and Sphericam 1 and Sphericam 2.

The specs look good. Has anyone used this camera? If so what did you think? Also, how much did it cost? I always get worried when the price is not shown (that usually means it’s very expensive!).

Via the jungle telegraph its price is around EUR 17,500

Oh, wow. OK much more than I thought. Out of my budget :frowning:

I really, really, really want this camera to live up to the expectation.

But I do have reservations. The photos they use for the promo on their site (via the Mapillary link) leave a lot to be desired (the lighting is terrible, quality seems average at best): https://www.mapillary.com/map/im/gZFAirbOONg6yOsdXUxV1w

The price point is a little worrying too. This is more towards the cinematic VR 360 cameras like the Titan, vs. Street View cameras like those from NCTech.

I’m not writing it off yet, but will wait to see some reviews and the camera in the hands of real users first.

1 Like

The camera was announced, last year, but never released, probably for some kind of issues.
Jeffrey Martins presents it as the revolution in the SV industry, but he has some marketing problems as published very bad pictures on the official website.
He says that there are customers with camera already delivered, but any example online.
I can’t understand how someone can accept to spend $15/$17/$20K. without any proof.

1 Like

There is an interesting discussion on Facebook too: https://www.facebook.com/groups/360.video.professionals/permalink/2730227060540863

I saw the price quoted at EUR 20k there @JimGayes360

1 Like

Another discussion here: https://www.facebook.com/groups/GoogleTIPS/permalink/3247721598613800/

From the founder of the companies responses (Jeffrey Martin), it looks like the camera is designed to go up against NCTech for Street View commercial captures.

The difference being although the camera is far more expensive than NCTech offerings, you don’t have to pay for the expensive NCTech upload and processing fees.

1 Like

Check out the low frame rate videos on YouTube of the Mosaic 51. I am not happy with the dynamic range, not to mention the outrages prices. Looks like there is a to get some money out of it before it is obsolete legacy.
Also, I wonder when this high resolution is useful while most viewers are on mobile device screens; what is that high-value niche? 8K seems the sweet spot to me.

I’d agree @OmniSynThesis360VR.

For computer vision, the kind of stuff we’re interested in here at Trek View, high res imagery is much more valuable. For example, being able to see the bark on the tree can help determine its type.

If your sole intention is to upload to Street View, the only one who benefits now from high-res imagery above 8k is Google. The viewer is not seeing the image quality you upload. Google compress the images, so you don’t get true 11k (or whatever) on Street View. At least not today.

Oh, and beware of YouTube video quality @OmniSynThesis360VR. That too is compressed by Google after upload. I can’t speak for the quality of the Mosaic cam, but what you see on YT won’t be the same as the output from the camera.

Hello,

I’m Jeffrey, founder / ceo of Mosaic. I know some of you on this thread either from online interactions or in person, such as David, Federico, Jim, and Joscelin. It’s nice to see you all.

We are a startup, building a camera that has been specifically designed for ground-level mapping and infrastructure inspection for B2B clients. We have designed the Mosaic51 camera specifically for that, and not for other stuff. In fact we have been very specific in our product design for a narrow range of applications, and have made an effort to not make a general-purpose 360 camera that is suitable for other stuff such as VR filmmaking or consumer applications for example. Designing products in general is very hard. Designing hardware products is even harder, and requires serious commitments that can’t easily be changed. So we have said “yes” to certain things, and “no” to lots of other things.

So as an industrial-grade tool for large-scale visual data collection by business clients, we have made a lot of very specific decisions in our product design, for example: waterproof military-grade screw-in power, ethernet, usb connectors (these are very expensive, but worth it for certain clients or applications; lenses that are protected by custom lens windows with specific anti-reflective characteristics, and again are waterproof via an o-ring and screw-in metal ring - these are not $5 lens filters you buy on Amazon, and it makes a big difference; a housing that is hermetically sealed with a valve on the bottom, and is passively cooled; GPU chips inside the camera which allow very substantial capacity for edge AI inference tasks and other stuff to be run on the camera in realtime; a 12-bit compressed RAW codec; and other stuff.

All of these things take time and money to design, prototype, test, and build. Additionally, the device itself is made of certain components that have a certain cost, and many of those parts can cost a lot more if you’re buying 10 or 100 at a time, rather than 1000 or 10,000 at at time.

Finally, some stuff in the world is priced according to the value it brings, and other things in the world are priced with a specific markup from the cost of the parts they’re made from (that’s what a commodity is).

So while I’d love for all of you guys to have our camera and use it - and I can assure you, it will overall do a great job and make your life easier even if (like anything else) it’s by no means perfect - we can’t make everyone happy all of the time, and maybe our focus is on bigger / more industrial clients at the moment rather than individuals, not only because of the price but due to other reasons as well.

Regarding the resolution - all things being equal, more is always better for some clients who need to either 1) read as much text as possible - can you read the opening hours and the menu on that restaurant window? and 2) any kind of 3D/photogrammetry benefits from as much resolution as possible, with no upper limit. Of course all things are not equal and there is a balance to be made.

Regarding the image samples, we will post additional samples (DNG files, h.264 files, source / stitched) very soon, thank you for that feedback.

Finally, let’s start a discussion about what YOU want to see - If any of you have serious pain points in your current data collection/photography/camera usage, and would like to have a serious conversation about what you would like to see in a future product, then I’d love to talk. Of course this is not the only product we’re ever going to build :slight_smile:

cheers.
Jeffrey

2 Likes

Always great to meet dedicated people, great that you are here too.

What I feel would be really great is the popularization of 360 photo mapping to have better maps and have the imagery too. If it facilitates 3D photogrammetry, then so the better. I feel it would help humanity to feel closer if we would have this. To get this within a reasonable time frame, as many people as possible should be able to afford a cam that is just good enough. As you know Jeff, I have a Labpano Pilot Era. I would appreciate the following.

  • A form factor size that is in between Pilot Era and Pilot One.
  • High-precision positioning. Refer U-Blox ZED F9K. (Most important for mapping.)
  • Better dynamic range.
  • Less parallax anomalies in stitching.
  • Capture once, share on many publication platforms with the easiest possible workflow.
    I consider 8K to be sufficient for the moment. As capture refresh should be at least bi-annual, higher resolution will be attainable over time, but is now last in my priorities.
1 Like

HQ | Blog | Spotted a Trekker? | Become a Trekker | Facebook | Instagram