Everyone is photo scanning assets and uploading for free - does that make artists redundant?

Whats the point in modeling an object in blender from scratch when any random person could take some photos with their iphone and churn out the whole model with textures with a few clicks and then upload it often for free on a big website like sketchfab??
With the tools getting better and with AI retopology probably just around the corner, im asking myself whats the point in Blender anymore unless i want to be an animator??

What are the best options to make a living in the field of 3D graphics going into the future??
Is it all worth it anymore?

1 Like

There is still the matter of creating custom assets if you want a unique scene.

Also some points to take home based on what you find here…
Models • Poly Haven

  1. Premade assets can save a lot of time in scene assembly, but chances are your art will then contain a number of objects that look identical to ones found in a thousand other pieces. If you don’t want people to say it looks like the one from this website or that artist, make your own models with your own materials.
  2. Choice, if you decide to stop modeling, then your vision becomes entirely dependent on what you can find online. If you want to create something truly special and can’t find anything that matches your vision, tough luck kid, wait a few months for more models to come in or shell out the cash for models at places like Turbosquid.
  3. Specifics, you will likely need to at least edit an asset if you are specific on how something like the rocks or the furniture should look. Otherwise, you will be forced to settle with ‘close enough’.
  4. If you give up modeling in favor of scanning, then keep in mind there are many things that would be nearly impossible to find using legal means. If you want a science-fiction scene with radioactive waste barrels, good luck driving to the nearest nuclear power plant and gaining entry. Then there’s obviously the potential issue that is going to every store for the purpose of photographing every item (in the hopes you find something with the texture you want). It gets even worse if you want to create a fantasy scene (for instance, good luck finding creatures that do not exist).
2 Likes

How are you going to photo-scan something that doesn’t exist?

Did artists that draw become redundant because cameras came out?

5 Likes

What are the best options to make a living in the field of 3D graphics going into the future??
Is it all worth it anymore?

Right now there is a demand for 3D modelling so just do it! Don’t think that something would disappear in near future, it will take years if it really would deprecated. Those years you can invest in your education towards being technical artist, procedural, animation etc. You’ve must always learn and adapt to the market demand.

Photogrammetry is not perfect, you’ve must clean mesh, create new topology, clean texture, create PBR texture - so there’s always hand work for people, even if workflow would be semi automated (auto topology creation, auto PBR texture creation etc.).

As people mentioned above, you can’t take photo of:

  • fantasy monsters
  • sci-fi stuff
  • low poly art
  • stylized art
  • 2D art like anime, comics etc.
  • everything realistic that dosen’t exist like person with specific features or objects hard to capture like animals, transparent objects, underwater stuff, inside of something stuff etc.

It will never replace traditional modeling until there will be an AI that would make an art exactly on person demand. If something like that would ever exist, don’t worry for your employment - probably AI and machines would replace most of people work :wink: (this would be the most wonderful and scarriest thing at once).

ok points taken, and i do know from a company director of a large manufacturing company that for example concept artists that did the initial dreaming up/drawings of a new product design were paid more than the engineers the CAD people the product visualization people etc…

So there is the whole area of creative thinking and designing new concepts that dont yet exist… I guess that does still make 3D artists relevant…

But still i have a depressing feeling about all this, i do think an AI is coming soon that will be able to dream up new designs and have them dreamed up with good topology and PBR textures…

On a side note, if you ask yourself how can all this photogrammetry and photo scans be utilized, you realize its already begun and that is the new Unreal Engine for games, with its unlimited geometry and consoles with huge disk bandwidth to stream in these huge high poly models for games that are hundreds of gigabytes in size…

So do we all need to be game designers now, just putting together game environments and levels using pre made assets from quixel and what not/?

On the contrary, it’s an iteration, optimization of production, engineering, design processes and there will be more demand… I had figured this some years ago. The creative spirit has just gave us tools to fix & evolve better stuff :wink:

From all the technologies available now and from what we can see, here are some points that dismantle the idea that 3D artists will no longer be needed in the future:

  • People almost always miss how AI technologies work and what are they going to replace. AI is used to create tools that help people in the field or make their jobs easier and more productive. End-users (customers for the 3D assets in this case) don’t have the knowledge to produce good quality 3D models using AI tools. These tools almost always contain a learning curve and a very high amount of options that an average user will not comprehend. For example, how would a person know if an AI tool retoplogized the model correctly if they don’t know what good topology is in the first place? same goes for UV unwrapping, baking…etc.
    All these 3D art concepts already became 100x times easier with new tools (AI or not), yet 3D jobs are still rising.
  • Many 3D art-related fields require the creation of models that are not real. This includes environments that don’t exist in real life (just as mentioned above), stylized 3D art assets, sci-fi or fictional assets, product visualizations of concepts…etc
  • As the creation of 3D models becomes easier with the introduction of new tools, the expectation of the models created becomes higher and higher. So although new tools might make it easier for someone to make 3D assets without the need of an artist, the quality of the produced assets will be less than what is expected in a good product.
  • AI tools are amazing, but they are overrated. Currently, there is not a single available AI tool that can actually retoplogize a model completely well. A professional will always be needed to complete what the AI had done, at least from what is currently available.

A situation analogous (yet very different) to what you say is web development. Despite many available tools that should make the creation of a website a piece of cake (such as WordPress, Squarespace, Zyro…etc), people still need web developers. In fact, people who use these simple services still hire someone to use them, because, with all the fancy customization options provided, it is just easier and more efficient to hire someone that already understands them instead of learning yourself.

I don’t share al that worries. Exchanging scans for hand-made modeling is just a step in the pipeline. All other hard work is still there. Moreover, scanning has a lot of hassles, models are full of bugs, cameras are getting in the order of billions polys res and tools are not still there…
AI is overmarketed. Still has to show any real everydary power. It’s just all proof of concept. And then… If they’re gonna do any AI auto unwrapping or similar I’ll be more than happy to use it…

Theoretically, you can turn those into scannable items by building them out of things like wood, stone, craft paper ect…, but it would not be an easy job and would require the development of many new skills.

More likely, you can print out the real world object to scan by using a 3D printer, all you need to do is hop into Blender with the 3D print addon active and…oh forget it.

1 Like

Photo scanning requires a TON of cleanup - An artist’s work

Photo scanning requires proper art direction of the objects being scanned if it’s a prop or costume piece - An artist’s work

Photo scanned assets will need to be contextualized. They won’t be used in a vacuum, they’ll be arranged with other assets. - An artist’s work.

The non-abstract naturalist (what some naively call ‘realistic’) aesthetic is not a singular goal that everyone works towards. Hell, most who even think they’re working towards it, actually aren’t.

Everything that is representational is filtered through a subjective mind. In the case of the games and vfx industries, they’re filtered through MANY subjective minds before you see the final product.

The workflow for Battlefield V-VI and Star Wars Battlefront (their first one) were entirely photogrammetry based. Do you think they fired a single goddamn one of their artists? No, man. They NEEDED all hands on deck. Not just for processing, but for flourishes. This is like worrying if mocap will replace animators. Raw mocap is terrible, no matter how many filters you throw at it. Look at any Borderlands game, all those super stylized humanoid characters with exaggerated proportions flopping around like mascots and hitting weak ass key poses. 3d is iterative. You’re GOING to need to touch up every data driven piece.

2 Likes

Everybody is 3d scanning and uploading for free ? That’s disgusting. Where ?

1 Like

If you have a photorealistic scene with thousands of objects to fill up the mid / background, then cleaned photogrammetry is certainly a time saver. But nothing beats modelling for the personal touches that are absolutely key making a unique experience. This goes for the creator as for the consumer. The 3d devil is in the detail, and that devil is definitely human and not machine.

Sktechfab. A lot is for free, then there’s the different cc licences where you have to give credit to the authors, and there are paid models as well…
It’s not all photoscanned stuff but a lot of it is and some of it is very well done without unnecessary high poly count im sure it took quite a bit of producing but the way they put it out there for free feels like a bit of a kick in the teeth for artists that spend time modeling stuff themselves… CC0 - A 3D model collection by Thomas Flynn (@nebulousflynn) - Sketchfab

I downloaded this model:

just for the sake of trying. A simple check with 3D-print addon shows:

A lot of cleanings to do. :wink:

Recently a lot of museum institutions started scanning their collections and published them. This is one of those. Better for the people, if they like it. Not a real problem for artist I guess, some more references…

Hi Thesonofhendrix

The simple answer to your question as raised in your posts header is … no

Photogrammetry is simply another tool and a great one at that which has opened up some incredible possibility’s. But to create any high quality useable 3D asset from raw scan data is still a huge amount of work.

If you point a phone at an object you can get a scan of some kind. But you will not get a good quality 3D asset.

I’ve been working in-depth with photogrammetry scanning for several years now. Both on fine art gallery sculpture projects and also historical archive and conservation projects. There is a large skill set demanded to work with this at an accurate high quality level. The initial scan is only the start of a long creator led process and detailed large scale subjects will often involve combining and working with the results from multiple different scanning sessions.

Hopefully this is seen as a positive reply. Take advantage of scanning to enhance your work and open up new ideas and potentials. Combine scan data with original hand built designs to make something new. Make new amazing environments with your own and pre existing scanned assets. It’s an amazing new tool for artists and creators. Not a threat to them.

All the best

5 Likes