What do designers and archviz people want in terms of technology?

[Christopher Nichols]

So I told Jeff that I would start to pay more attention to this forum as I have lost touch for years. I am curious as to what people are actually looking for in terms of technology these days. I even did a podcast with Jeff about it....

 

https://labs.chaosgroup.com/index.php/rendering-rd/cg-garage-podcast-15-jeff-mottle/

 

So I am going to start reading this forum a lot more but if anyone wants to point me to a few interesting and key posts, I would love to see them.

 

Chris

[Corey Beaulieu]

I'm by no means the guy to consult on the subject, but if I may put my thoughts out there I'd say this.....

 

There exists a constituent of artists and studios that can do photo-real to an incredible degree. There is a growing number of organizations creating Real-time scenes and apps/websites that are blowing people away. There are also a great number of artists telling architectural stories in non-photo-real ways that are extremely compelling.

 

What there is not, in my world, is solutions to the everyday problems that we face.

 

Max is great, but it is crazy limited for the extremes of arch-viz and concessions are made regularly. Detail is also becoming king in the render world. Clients are expecting magic more and more as they themselves become accustom to our high-level accomplishments as an industry.

 

I think that the place for the technology of now (and I say now meaning the most immediate need) are the advancements that solve our daily grind. Asset libraries on the level of Ikea's custom tools that they have been presenting at various forums and the tools tat allow us to model extreme designs and the environment tools to create elaborate scenes are major hurdles in my book.... Forest Pack is a great start, but there can be so much more....

 

The landscape of Arch-Viz is forever growing/changing, but the place for Tech in my world right now is to catch the software up to the talent. We have all compromised detail for speed and used Photoshop to cover our tracks for too long. Vray 3.0 is a great step and tools are emerging regularly beyond that, but our core toolset of Max is not yet multi-threaded and craps out on us in our times of greatest need. In order for the bigger minds of our industry to invent the way forward, I think we need to get these road blocks out of the way and make software explorations a regular event. It is a shame to me that people must create custom tools to invent. smaller studios and individuals cannot do this easily and these people might just be the gate keepers to our way forward.

 

I will add that I believe the people in this industry will find a way regardless, but bogging everyone down with the need to learn more and more software packages is only slowing our progress.

 

This isn't particularly well written, but it is my off-the-cuff opinion.

[RyderSK]

After 3 times re-reading the original post, I am not 100perc. sure if I should click on one-hour long podcast primarily, find interesting posts, write what I want in Vray from ChaosGroup or follow the original question in topic.

 

So I will instead continue same as Corey based purely on the topic.

 

I want two things:

 

'Instarealism®' - Stuff that looks photoreal out of the box without additional effort on top of building the scene. Shaders based on existing world (more PBR/Disney like, not glossy and glossy intensity,etc..) Tonemapping that resembles existing real world camera sensors (and not disgustingly weird Reinhard). What I see in framebuffer, is what I see in my NikonD800, not something that requires curves hocus-focus to even get near real without anything to really based it on.

 

'TheSims®' - Workflow. Something as Corey touched on. The workflow in software is still cumbersome. We are finally getting some decent asset managers that make 3dsMax bit more enjoyable beyond the classical 'click merge/wait/fix paths+whatever else' like the DesignConnected's Connecter and SiggerVMPPC for materials, but that could be integrated further, more together into simple streamlined set of tools that let me build whole scene in matter of minutes without ever going outside of native tool (3dsMax).

 

It could be summed into 'One click magic button®'. Something everyone dreams (or fears !) will come soon, but all I see is even more robustness more flexibility bullshit to accommodate everysingle lost souls' needs.

[Scott Schroeder]

Smart integration and interoperability between apps. I have never understood how if you go from Revit (Autodesk product) via an FBX export (the end all world savior of Autodesk inter-op format) and import that into Max (Autodesk product) that your curved surface is now a chunky mess of 1.21 gigawatts of verts. Yet, the same curved surface exported via DWG (Autdoesk product but has the stigma of the what era are you from? The 1980's?) format and your curved surface is perfect as you would expect a curved surface to be.

 

Add to that, actual testing of some real world uses. The interoperability videos that you see are always of boxy rooms. Well of course it will work then, but what about what the rest of the world does?

 

It's becoming rarer that we live in one software package, so let's start to get these things all to play nicely together. Whatever technology is implemented, I would like the dev team to take a step back and breathe. Release it correctly and make sure it functions in a real world workplace and not some isolated Autodesk lab. "Do it right the first time." as Mike Holmes always says.

 

We are well to creating hyper real imagery, but I always feel I have to push it through the software with my foot to get anything done.

[Nils Norgren]

So I told Jeff that I would start to pay more attention to this forum as I have lost touch for years. I am curious as to what people are actually looking for in terms of technology these days. I even did a podcast with Jeff about it....

 

Hi Chris and welcome back, I think it might make sense for everyone here to get a quick CV so they can direct some of their questions to your frame of influence, (many are fairly new to this forum) There are still plenty of old-timers like me, although I admit that I don't post very often, and tend to focus on a narrow subset of the site.

 

Again, welcome back and tell Lon to take it easy with the dancing at siggraph!

 

-N

[Christopher Nichols]

Hi Everyone:

 

Nice to chat again Nils. We should get together. I doubt anyone really needs to see my CV.

 

Thanks for your responses. Here is what I have heard so far. There is little drive for new technology. There is a lot of drive to fix and make the current technology you actually have easier and simpler to use. Such as:

 

Make integration between different applications easier.

 

Just so you know, this has been a big goal in another field called VFX, which I have been doing it for a fairly long time as well. What happened a few years ago is that Sony Imageworks and Industrial Light and Magic came up with a new "universal" format that goes well beyond the outdate OBJ and FBX formats. It is called Alembic and has the extension .ABC Alemic files support way more, including point cache, hair, UDIM textures, just to list a few. For that reason Max, Maya, Houdini, etc... (all VFX programs) now support Alembic. In fact, in V-Ray, you can read an ABC file directly instead of an VRMesh.

 

I am not sure that ABC files would work for the Architecture and Design field as is, but there is no reason why it can't. Alembic is Open Source so it may be interesting to find out the needs of the Architecture, Design, and ArchViz community and make Alembic work as part of their workflow.

 

Quick and Easy Asset Libraries

To be honest, the architecture and design field is already way way ahead of the curve on this. There is a TON of assets that people can have access to. That are of extremely good quality. Back in my day it would cost me thousands of dollars for a single chair model. I have found that most people today tend to make their renders mainly about the great assets that they have purchased instead of the space itself.

 

New plug and play shader

This is a really good point. There is a lot of work being done here. Unfortunately it has been polluted by buzzwords and misinformation. PBR make a lot of sense for people coming from the real time word where the basics of energy preservation in a shader was never set. Same thing for the guys from the Renderman side. So Hyperion was very cool to that crowd. But if you already come from proper energy preservation, you tend to look around confused about what all the bizz is about. As an example, having a variable like "metalness" is what people call for when they want a PBR and it makes no real sense. Real world scientific values like Index of Refraction, makes more sense.

 

With that said... there is work being done that can generally move us away from standard Blinn BRDF. At the same time, there is a lot that can still be defined through these shaders so I don't see them going away any time soon. Having things wrapped up in a nice little package where you get more intuitive parameters makes sense to.

 

Anyway... just my initial thoughts on this. I would love to hear more from you guys. Sorry the podcast is too long.

 

Chris

[RyderSK]

Thank you for response Christopher ! Oh, despite watching your tutorials years ago, I had not put you together with that Vray's IRL video until you put up a photo :- )

 

I don't think Corey was alluding to much as assets themselves when he mentioned libraries but more the integration of them into workflow. IKEA was the one who boasted about having catalogue where you click on asset and then place it. We already have similar tools available (though only fairly recently), like the Connector, but still not simple enough.

I can imagine the simplicity similar to STEAM video games catalogue (and their mods installation). You browse assets online (for example Evermotion or DesignConnected), click to buy, and instead of it giving you zip, you could directly place it into 3dsMax and that would create a library, thumbnail, filepaths and you could directly place it in your scene. This is already possible through UnrealEngine4 and its Marketplace. One click and use it.

So almost every digital application on market (from different sectors) is more user-friendly and evolving in better way than 3ds Max. Obviously it could be argued it's more "pro" tool than Unreal Engine, but that doesn't mean it can't be streamlined in same fashion.

 

Regarding PBR vs current popular raytracer's (MentalRay/Vray/Corona) material. Of course, the same core is already there regarding energy preservation and other internal stuff. Same regarding new BRDFs like the pretty nice GGX. But it's indeed packaging/shell that really makes it that much better.

Metalnes is hardly gimmick in my view, the IOR difference is often visually indiscernible between varied insulators so whether my wood is 1.52 or 1.6, matters not (but the PBR just rescales that to [0-1] as well ), and no one knows the numbers anyway. But insulators and conductors/metals differ to point, where it makes logical sense to separate and automatize it for users to choose based on logic. The current materials (with exception of Maxwell), don't support complex IOR (ok, Vray introduced a map into reflective slot for complex IOR, but this just further destroys the shader's clarity ), so absolute physical plausibility isn't issue/argument anyway, since you have to do it with curves. But again, with minimal visual difference for most part.

 

Same goes for the other attributes introduced in the Disney paper like direct coatings inside single shader, etc.. all parameters re-attributed on [0-1] scale where each spectrum end is physically correct, but the choice is purely visual depending on artist.

Even though for example the UnrealEngine4, offers only simplified shader from this paper, working with it is so much more pleasant and you could teach it easier to non-technical types. I have spent years reading on technical papers but when I teach others how to create photorealistic plastic material and stainless steel I see blank stares. With artist-friendly approach like Disney's PBR, that could be taught in fraction of time.

 

I don't see the need to scratch anything, people love their legacy ways. But I wish renderers would at least introduce alternative shader, exactly like the one from Disney. Some harken to old days (Mental/Vray/corona), others created artificially too complicated ones (Maxwell) and others (like Redshift) something completely chaotic and illogical, a melange of all above approach.

 

 

All of the above written, could be done in few afternoons by skilled scripter in existing applications. But yet I don't see anyone putting much effort that way. Instead we get more buttons for this and that, each time catering to very niche audience.

Edited March 27, 2015 by RyderSK

[Christopher Nichols]

Thanks for your response. I know what you mean but I still have a hard time with the way things are said as it is very misleading. By calling things physically based rendering it is implying that other are not physically accurate, at least to the artists. Also by arbitrarily making values normalized with no real measurable value in the real world, such as metalness, all this really does is make it easier for people that have a hard time with names like Fresnel, IOR, etc...

 

We are in the process of looking at other ways to get real world measurable reflectance for people to use in their renders. I am also working closely with the guys at ICT to do accurate mapping of skin data where I can get a measured value from scans and know that it matches 1 to 1.

 

Nonetheless, once those are working it is a matter if putting a nice little package around it so that artists can use more familiar terms to process the data. I know the feeling, however, when using a fluid sim, I sometimes wish there were parameters like splashiness.

[RyderSK]

We are in the process of looking at other ways to get real world measurable reflectance for people to use in their renders. I am also working closely with the guys at ICT to do accurate mapping of skin data where I can get a measured value from scans and know that it matches 1 to 1.

 

Something you can disclose ? Even outside of skin ? Sounds very interesting. Same regarding the authoring of scanned data. How can I author materials scans to match 1:1 ? This is something Quixel guys say as well about Megascan but don't reveal much behind the processing of those scans.

Ease of use for both measured specular reflectance and albedo is something I would love to have at disposal. For metals, we can already semi-conveniently copy curves from refractive index (or use Vlado's OSL shader to map complex IOR) if that extra level of realism is necessary, but when it comes to Albedo, it's still just wild eye-balling and extrapolation from few charts. Making this easy for layman, would be fantastic.

 

 

Last thing regarding PBR.

I see how the buzzword could be offending to raytracers who worked in such way from beginning, it wasn't my intent, I don't see magic behind it. From my point of view the nomenclature&parameters from Disney are superior for artists compared to classic spec/gloss for we have now. And since these renderers/engines/apps using it are getting so much steam in public (Unreal will soon have more users than all raytracers together), sooner or later I think someone's standard will have to go, I can't see it living separate lives forever. What I would like right now, is to do perfect static off-line raytraced images (Using Vray or Corona) which I then translate directly to real-time app like Unreal without doing conversion to values and textures (like inverting glossiness so I could get roughness). Understanding both namings, I can do that myself easily, but why eventually not get it on same boat instead of translating/calibration. Simply offering alternative shader could be way to please such crowd.

[Tommy L]

Welcome back Chris Nichols, I have a vague memory of a road with an overpass VRay tutorial, circa 2005?

My frame of reference for 'what technology dev' I'd like to see has a narrow horizon. Although I have experience with a broad array of software/hardware, my daily 3d tools are VRay and Max.

For me the developments that need to be seen in our industry need to happen outside of the lab, but in regards to your question, whats the context? Idle curiosity? Or market research for commercial development?

[Christopher Nichols]

Thank you Tommy:

 

The Lab itself (at least the one I am setting up at labs.chaosgroup.com) is really setup to engage with the users. People may have some ideas of things they want to move forward and we engage them. It is an opportunity for you to help drive the development of technology. As an example, a lot of work has been done on GPU technology because we had real projects that we worked on to drive that development. Not only did this drive V-Ray, but we also got NIVDIA involved as well. In fact, NVIDIA was the company that help push it.

 

Try to look 1, 2, 5, even 10 years from now. What do you imagine happening? Where do you see your work going? What would you like to see?

 

Does that help answer your question?

[Mario De Achadinha]

Hi Chris, thanks for your feedback and everyone's thoughts and ideas, think it's a great thing you trying to do by asking the guys in the front lines what they need. If I would want one thing it's simple but probably complicated. Photo-real rendering (GPU) in 10secs in realtime and accurate, no proxy, memory & shader problems, also easy to use one panel for settings. 1 button fixes noise 1 button fixes GI blotches.

 

Chaos group should have a hardware spec made available for accurate hardware tested results, this spec should realistically be under $5000 complete build, anything over is pie in the sky, think the Titan is making that gap get closer. I know hardware isn't your guys core business and you can't control the market, but when you make the process a vertically Intergrated chain(Apple). As an example create a pc spec called "the vray beast", built and tested by chaos to get results in 10 secs, could be interesting, because what you test and guarantee we can get those same results if we also buy a "beast". Also no Mickey Mouse scenes, large production size projects.... Ok that was a bit off subject but you get what I mean ....consistency with software and hardware that equals results.

 

Deadlines with clients are always fast and furious. This is becoming more intense and competitive. We are still required to deliver high end results regardless of timelines, which means TIME is always the factor. Our value will come from quick feedback from the engine, in order to get it back to our clients and work through the revisions to make the end product resolved & high quality. Our process must be more focussed on design development, art and less about memory, rendering and techno jargon as the client couldn't care less why our shaders don't look realistic they want results, let us worry more about our clients requirements then technically issues with the engine.

 

Hope this gives you some ideas

 

 

Sent from my iPhone using Tapatalk

[Christopher Nichols]

Thanks Mario:

 

Many of these things are close to or already pretty much there, which is why I wonder why we haven't seen much interest in the ArchViz community in this. Do they just not realize it is there?

 

Super fast rendering scenes:

We are now looking at not just fast rendering scene in seconds, but in fractions of a second.

Yes, it requires a lot of hardware to do that, but read on, as I bring that up in another point.

 

No artifact GI super fast:

GPU rendering now supports Light Cache which makes interior renderings on the GPU extremely fast. Since GPU rendering was traditionally limited to only doing brute force GI, they slowed down a lot on interiors. This is no longer the case with the latest V-ray using Light Cache.

 

Hardware testing:

We do work with hardware companies a great deal. I have a mountain of GPUs in our office that we test all the time. NVIDIA even loaned us a VCA. Sure it costs $50k, but if you want to see how fast GPU rendering can be, this is what you need.

 

And yes, you are right that the cards are expensive. But if it was not for all the testing and development we did on things like the K6000s ($5,000), we would not have something as good as we do ready for the Titan X (a card that is even faster, has just as much ram, and $1,000).

 

You have to test on the latest because that will eventually be mainstream. Eventually 12 gis will be standard on most cards, and people will be complaining about the price on the brand new 64 gig video cards.

 

New technology, on the hardware side, is always expensive when it is cutting edge technology.

 

Beast Machine:

Your own personal beast machine will go away. It make no sense to buy a machine that does nothing most of the time. While cloud rendering may seem expensive now, that will change quickly, very quickly. When it comes to high performance rendering, you will pay by the cycle and not worry about hardware going out of date any more. What if you sent your large 5k render to the cloud, got it back a few mins later, and it only cost you a few bucks? Why pay $5,000 for a huge workstation? Just do it on your laptop.

 

TIME:

Between high end GPUs, Cloud, Flexible systems that can do GPU or CPU, Biased rendering that can dramatically reduce render times especially for interiors, you will not see rendering as process any more. Rendering will be something that just is. Settings as you know them will be going away as well.

[Mario De Achadinha]

That sounds all exciting Chris and exactly what we looking for, thanks for your insights into what's happening. It's just a case of what we experiencing with GPU now, when we got our latest Vray 3.0 we saw all the videos of GPU and speed then tested it out and was getting 2 hour GPU renders which made us revert back to our CPU rendering solution. Unfortunately when you have a team of visualisers to upgrade all the workstations to the latest specs and the cards are still too expensive so it's a slow process . For us we need to wait for the hardware prices to come down before GPU can be solid workflow in our production as we not getting renders in seconds for the moment, but I am sure it will come it's on our radar for R&D

 

 

Sent from my iPhone using Tapatalk

[Mario De Achadinha]

Very cool video

 

 

Sent from my iPhone using Tapatalk

[ryannelson]

Why even bother with GPU workstation renders at all? By the time GPU rendering tech catches up, it'll be edging on Cloud solutions being viable, don't you think?

[Christopher Nichols]

@Mario First off, I'd be very curious as to why your GPU rendering was taking 2 hours. Did you try and reach out to people at Chaos Group about that? Again, our software is only as good as the feedback we get.

 

@Ryan It is a valid point but I don't see those being separate issues. GPU rendering could end up on the cloud as well which mean you have to be ready for it. I also don't see GPU rendering being that far out of reach, under featured, or simply not ready for prime time.

 

We have been working hard with Kevin Margo to help him make his short. Back when he did this teaser he was getting 7 mins a frame on each workstation. On a 32 core CPU render, it would take 3.5 hours a frame. Granted it had $15,000 worth of GPU in each box at the time. Today, that would be $3,000.

 

I think the results speak for themselves.

[ryannelson]

Thanks for the reply, Christopher. That's more or less the assumption I had. GPU renders will be viable on workstations for us kind of folk here in the very near future, especially with cards like the TitanX being available now. I was more implying that by the time it becomes the standard approach, it will probably be fighting with cloud solutions (which may still be GPU based) - probably within 5 years? I think my wish would be for future tech would be a painless transition into cloud rendering. I don't intend to undermine the viability of GPU renders at all.

 

That short is unreal, it's beautiful! I want to see more!

[Mario De Achadinha]

Chris, Very cool video again. We didn't speak to chaos group as we felt we didn't have the power required to get quick results with GPU as seen in demos before, hence 2hours still with grain, so didn't want to waste anyone's time.

 

Granted that's our problem as we not working with you guys to get GPU working for our workflow, so we take responsibility for that. I can't speak for everyone on the forum, but if I had to talk about what normally happens in our studio. When we under pressure to get projects out we will try something new on a live project as a test. If it doesn't work such as GPU in this case, we will revert back to old workflow, because we know we will get the results to get the project done.

 

From that point we parked GPU, as we felt we were not ready for it as we don't have the hardware under our desks. I am happy in next few weeks to have a teamviewer session with chaos group to discuss hardware specs and work on the file to see if there is something we doing wrong with GPU, the key for us is we will only consider to introduce it into our workflow when we confident it's faster then our past techniques. Ultimately we will use GPU everyday if it gives us the quick render results we after. Question, has chaos group done a survey of the arch viz industry to find out how many studios are using GPU 100% in their workflow, could make interesting reading and educate us on what's happening in our industry?

I think GPU is on everyone's mind, just not sure who is using it on a daily basis, maybe other guys have better success stories of implementing GPU in their office?

 

Sorry to make this topic all about GPU. The original post was about technologies, just find it an interesting subject at the moment.

 

 

Sent from my iPhone using Tapatalk

[Francisco Penaloza]

Hello Chris I am big fun of your work, I have learned a lot from you.

 

Regarding your question I have to be honest with you, when I read it, I instantly though that this conversation will go in direction of GPU rendering. Sorry I am not a big fan of GPU rendering as a solution for me at least, my projects scale is way beyond what a GPU rendering can do now, and not even in the close future. Also I am rebels by default and it seems a little unfair to me that only one company has absolute control of that technology, and this concept point me to answer to your general question of "what archviz people want of technology"

 

I agree with Juraj, and other people that mentioned that over all rendering process should be more straight forward or at least not mysterious as it was for several years. I am not afraid to learn software, I worked several years in Mental Ray and you know how much you need to dig in to the render engine to get the results you want. But for instance when you see an render engine like Corona, so straight forward you think what the heck, why I have to worry about sampling. But I also understand that my priorities are different from instance an artist from ILM. They need all the flexibility that V Ray or Renderman give them, they need the technical stuff, we archviz we don't need that much, and an architect that does his own renderings, he need lots less. so make a software that fit all this need is very hard to implement.

 

So my first point would be simplicity and efficiency and maybe dedicated/focused technology.

 

My other point is sharing, with Juraj comparing Game industry with their "Real world shaders" yes that sound really cool but, game engines they are not that many compared to 3D softwares or traditional rendering engines, and we all know that we can not render a Maxwell shader in V Ray, I am really envying their cohesiveness or standardization (if we can call it that way) But compared to standard render engines have ready to pup assets is almost impossible unless you concentrate in one render engine and one provider, which really defeat the purpose of have many options.

The same thing with GPU rendering only NVidia card, can you imagine how much advance we would have now if Intel, AMD, IBM and many more could produce CUDA cards or parallel processing card? maybe they will also be writing new shaders or apps to work in that environment, we would have way more options now for sure.

I really hope this does not happens to Cloud applications, hope we won't end up with only Autodesk, Amazon and and Google controlling all the cloud market and all the other options will be so small that it won't be smart to invest resources on them.

 

We have BIM now, in theory would give you the possibility of have one 3D asset and transmit all the information from one side of a project to the other, and "Seamless" generate visualization in the same model. We all know that never work seamless and that happens under the same company software. You transfer a REVIT model to CATIA or ArchiCAD and you loose most of the info. Alembic works great in VFX, FBX work sometimes in the rest of the planet.

 

My last point is cost, you mentioned that cutting edge technology is expensive, I agree with that, but I also dressage. I think that if now we are smarter than 10 years a go, we should be able to do the same thing, better and faster, then it should be cheaper. But not always works that way we all are paying a premium for technology that it seems a luxury but in production environment it is a necessity really.

 

I don't want to sound pessimist, I am all about technology but I feel that several of the problems that I pointed are not technology related and that is what I feel is dragging us to not advance faster.

[artmaknev]

What I really want is a technologically advanced desk/chair combo unit made specifically for long periods of seating! I want that chair to change positions so your legs can go up and your monitor go up too, its like a system that changes your entire position, relief stress on your legs, neck and shoulders! I think we all need that, we are working the same way for the past like what 1000 years!!!! Seating on chairs and using regular desks, its not good to our health, we need better seating setup!

 

Better yet a complete working cocoon!!

Edited April 13, 2015 by artmaknev

[RyderSK]

Better yet a complete working cocoon!!

 

What's the issue with the one you got right now ?

 

[Francisco Penaloza]

oooh lets go crazy with this one, look how happy this dude looks like

http://www.formfunc.co.za/blog/wp-content/uploads/2012/01/man-.png

 

or this other one.

https://www.shinynewwant.com/wp-content/uploads/2012/02/emperor_625x3401.jpg

 

But actually is a good point, how long you think input devices such mice or pen tablet will stay as main control of our 3D software, hands, globes or magic wands will ever be efficient enough?

I don't think touch screen will work for us, seating holding your hands in to a screen hurt your back more than resting your hands in the keyboard all day long.

[heni30]

What I really want is a technologically advanced desk/chair combo unit made specifically for long periods of seating! I want that chair to change positions so your legs can go up and your monitor go up too, its like a system that changes your entire position, relief stress on your legs, neck and shoulders! I think we all need that, we are working the same way for the past like what 1000 years!!!! Seating on chairs and using regular desks, its not good to our health, we need better seating setup!

 

Better yet a complete working cocoon!!

 

Here's a thread I started a while ago:

 

http://forums.cgarchitect.com/75638-anyone-work-standing-up-occasionally.html

 

Also Wired magazine, for a long time, in their staff list credits, listed a woman's name who was the office masseuse. I guess she just walked around the office giving shoulder rubs. (I'm guessing that's as far as it went ;-) )

 

Brookstone has massage chairs in their showroom with moving parts that massage your body with different setting choices. To offset the initial cost you can make it coin operated.

Edited April 13, 2015 by heni30

[artmaknev]

hahahaha thats great!! I really really really want that happy dude chair thing!