Shapeways Pricing

I'm a little confused on how shapeways does Pricing for their items.

One model I uploaded came in at 8 inches height, and cost $550.
Another model, same scale size, at 8 inches, cost about $1,000's.

Either I am doing something totally wrong, or the price selection is a bit wonky.

I'd love a set price option. I'd buy alot more if I knew I could get everything printed at a 10- inch scale for $550.
8 inch scale, 450 and so on...
But when the price is seemingly random for various objects, it makes it a lot trickier to modify it so it's affordable at a decent scale.


service is awesome though. just wish its more user friendly for beginners
thanks

Pricing is primarily proportional to volume, not just to height. Something that is 8 inches tall, wide, and deep will be relatively expensive even if you use tricks like hollowing the model and leaving a 40mm round hole connecting the hollow interior to the outside. So think thin walls, hollow shapes, and a large 40mm diameter pass thru hole to minimize cost.

how would one do that for this figure?

https://scontent-a-lga.xx.fbcdn.net/hphotos-xpa1/v/l/t1.0-9/ 10922828_10152960014035211_4469312279531293352_n.jpg?oh=312c 7b3d8bdbaefaf410031b0658b723&oe=5536D45A


I uploaded it to shapeways and it came out at 4 inches and $550s. I'd love to get it to ten inches and $200s...

It has no 'inner wall' so to speak. IF you delete a polygon in the front, it will show right through to the other side.

packingprotons have you tried making the model hollow?

If you give it a decent escape hole (check the material guidelines for the specific material's requirements) and make the walls have thickness you might be able to bring the price down by a good amount.

This tutorial can give you a little more info https://www.shapeways.com/tutorials/creating-hollow-objects

Making the object hollow is perhaps not the biggest challenge as I assume any decent 3D application can do this. However making a nicely integrated escape hole is another story. I would suggest after you have hollowed your object that you think a little bit different about the escape hole. If I were you i would make the escape hole behind the belt buckle or tv screen or whatever it is in the front of the model in his belt height. Then make this piece as a 2nd object (a cap/lid) you can snap on. If you offer it for sale in e.g. strong flexible you may even have both objects in the same file meaning that the buyer would only have to buy 1 product (2in1) and snap the piece on themselves.

Cheers, Mike

Love the model...not sure who else knows the reference but it's fantastic! OP, it sounds like whatever program you're working with is treating the model as a surface instead of a solid. What software is it?

Hollow with an escape hole will definitely cut some costs but with the new pricing, it's hard to say by how much. If you're printing in WSF then you're only going to save $.07 per cm^3 when you trade material for machine space by hollowing - the inside space will still be unusable as far as the algorithm is concerned and you'll still pay for the machine space.

Still, it will cut some costs so give it a shot. I would say remove Krang from the belly and make two escape holes for the body suit - one within the area of surface where his head comes in contact with the inset 'ceiling', and a second hole within the area where the bottom surface of Krang comes in contact with the inset 'floor' - leaving an offset around so the hole is there, but the top and bottom of Krang can still make contact for gluing him back in. Additionally, make the same sized escape holes in Krang so they match up to the holes on the 'ceiling' and 'floor' of the inset. This way the complete model will be hollow, they'll be separate pieces which will make it easier to paint (if that's part of the plan) and once assembled, all escape holes would be covered up.

It is called SoftImage..

OK, I think I understand. I made it hollow and created an escape hole..
-edit well, that didn't work. it claims that this is the measurement...
10.106w x 13.419d x 4.95h

and the cost is 850 in white strong and flexible. something isn't adding up

You also need to thin the walls as much as possible but remain thick enough to retain structural integrity. Here is an example of a big shape item. I think I went with 0.1 inch thick walls, hollow interior, and 40 millimeter holes on both ends which results in a cost around $150. The old version jumped to around $300 when the new pricing rules hit the fan last year before I made these modifications.

http://shpws.me/qlyb

I don't think you'll do much better than $850 if your dimensions are in inches.

The overall dimensions will remain the same, but with the missing material inside, you should see a price drop. Is the figure that is $850 hollowed out the same as the one you mentioned in your original post that would cost $1000's at the same scale when solid? What was the price before the hollowing with escape hole?

The calculations that they use for price are very precise - assuming previous bugs people were seeing have been fixed. So knowing how they make those calculations is crucial to understanding the price to print. As I mentioned before, the price for material usage is only 1/3 more expensive than the usage of machine space (which is the empty space in and around your model that results in degraded powder on account of other objects not being able to fit inside) so the price drop for the removed material will not be proportionately equal to the subtraction of material from the model.

Out of curiosity, is there any place on the model that if you split it along where two forms meet (for example, the entire chest and back area between the 'suspenders' from the front where it meets the 'belt' to the same location on the back) that is over 40mm in diameter? Here's a picture of your Krang body suit model with markups to illustrate what I mean:



If you can split the hollow model into two parts (don't implement the previous escape holes or separation of Krang himself for this example) along the red line from front to back and the distance between suspenders at any point along that split from front to back is 40mm or more, then they will be able to fit other objects inside of your model and you won't be charged for the usable machine space inside of him - which is the majority of your costs right now with the hollow model. So you'll incur a mere $1.50 labor cost for the extra part but potentially save $100's on machine space, as the machine space usage will be more like this:



Rather than paying for ALL of the empty space inside. The red line is the shell of the model and the blue area represents where they can't place an object within - empty space being where they can and therefore won't charge you.

You could also make the split where either arm meets the 'suspenders' as long as the smallest distance between shells along the outside edge of the split is 40mm or more.

yes, I believe it was the same scale..maybe a little bigger. it did go down a bit, but not as much as I was hoping.

I am trying to keep the figure in one piece.. If I start selling these,trying to explain to buyers why it arrived separately and how to figure out the assembly would be a bit of a pain.

Material also counts towards machine space. So "air" is paid at 0.21 rate and material at 0.49, so the saving is 0.28. Upload a cube if you don't believe it, and remember you will pay around 1mm of "atmosphere" for such cube.

I was already wondering why machine space cost almost as much as the material itself but, if I understand correctly, they're including the space occupied by material as well in that price? How does that make any sense? Isn't the whole purpose of charging for empty machine space to recoup losses on material not used? And yet they're charging for material not used...where material is used!

So...

They're charging for material.
They're charging for machine space.
They're charging extra machine space for areas already being occupied by material.
They're charging an extra labor fee, per part.

And not only are they charging for machine space around the part, but they're doing this for every single part in a multi part design and therefore duplicating machine space they've already charged. All this based on the assumption that the group of parts needs to be split up and the designer hasn't already considered the best layout - which let's be honest, most of us have been doing for years.

What I find most insulting is that despite the fact that the designer may already have done that work for them to figure out the best layout - they still charge them as if they uploaded multiple files with no consideration, and are paying their engineer part of that labor to "fix" what isn't broken.

I was very understanding of the new price changes, but there is a lot less transparency than they led on. I'm really disappointed now.

Machine space costs are like the rent you pay for an apartment having some square footage. Material costs are like what you pay for utilities in that apartment. Your monthly rent doesn't change if you use more or less electricity that month. And the border of dead space around your model is there because it's a keep-out zone for other models. There's also a lot more dead zone between models because you can't pack things with 100% efficiency, and those costs are added to overhead that's embedded in everything else including part handling fees. The end prices are skewed more to machine operating and human touch time costs than they are for the materials.

If you read some of the background material, then especially on the WSF printer, the unused "support" material can only be recycled a small number of times before it degrades below the acceptable point for being used on new items. That is part of the theory for charging for machine space.

The analogy appears to make sense but it doesn't work perfectly. For the sake of simplicity, we'll leave labor out since it doesn't factor in until post print.

If what we're 'renting' in an 'empty tray scenario' is a portion of the cost to run the machine + degraded powder (let's compare this to renting an apartment but never using any utilities) then we're being charged for 'renting' that space because someone else can't use it, in essence. Using utilities in an apartment scenario costs extra money because additional resources not local to the apartment are being used as well.

But in the 3d printing scenario, any materials (utilities) being used are not additional; they occupy a portion of the space we're already renting, and the extra costs are due to the state of the powder once the printing is done. It's used powder vs. unused powder. But instead we're being charged for the unused powder twice. In other words, machine costs should cover cost to run the machine + degraded powder, not cost to run the machine + degraded powder + used powder - that third item is already covered by the material cost. Instead of swapping out degraded powder with used powder, they're simply tacking on the price of used powder and taking nothing away from the cost of degraded powder, which is what is actually happening during the print.

If we imagine a theoretical scenario where the unused powder is 100% reusable then we would incur only our share of the cost to run the machine, would we not? And that cost would be exactly proportionate to the space our model's physical 'boundaries' occupy. Add in the cost of powder that is degraded after the print due to the space our model occupies, since that's how it actually is, then we incur that cost. Wouldn't it make sense then to replace the volume of powder that, instead of being degraded, is used in the actual form of the model, charge us more since the powder isn't being reused, and call it a day? But instead, they simply tack on the cost of the material, keeping the cost of the original unused space.

And now that I think about it, labor makes more sense as a comparison to utilities than material cost since it actually IS an additional resource to the process. Those additional costs make perfect sense to me, but not the duplicating of machine costs.

Yeah, I understand that part completely. Of course they have to charge for unused powder since it's not the same as fresh powder after the print.

But consider this - if they couldn't reuse ANY powder, then the machine space cost would be an appropriate fraction of the cost to run the machine plus one set cost for the full space our model occupies (used or unused) since it's the same result either way. But since they CAN reuse the powder, we should be paying a fraction of the material cost for unused powder - and that's worked into the machine space costs. However, what we're actually paying for in the machine cost is: cost to run the machine + unused powder + used powder at unused powder price + used powder at a higher rate. That just doesn't compute!

I wouldn't have a problem with just charging for machine space. That would make things much more like buying into a portion of a wafer to have your own integrated circuits fabricated. In that world all customer costs are based on wafer area. It doesn't matter if you put 1 transistor on your chip or 1,000,000. You can put 5 bond pads on the chip or 10 bond pads, 10 interconnect traces or 1,000 traces. The price of equal sized chips are the same because the overhead costs of capital equipment, electricity, and labor mostly dwarf the material costs used in fabrication.

So in the simplest case the 3D printer machine space could just be a rectangular bounding box volume which you pay for and can use in any way you wish. Other people's designs would not be allowed to penetrate your box, nor would your designs be allowed to penetrate into other people's boxes. Or you can do something a bit more complex as in the present pricing method where your model is discounted when allowing access to internal useable voids with a proper sized hole.

We can quibble about how overhead costs and profit margins are distributed to every model produced. Just a machine space cost, just a material cost, or some combination of the two (which is what we have now). I think what we have now is reasonably fair. Costs of a model aren't just determined by one tray, or how much powder is recycled from any particular tray. The pricing algorithms are presumably based on the cost/revenue history of hundreds of trays in any particular material. The primary way for Shapeways to lower costs is to find more efficiencies in their processes which presumably would be passed along to customers in order to remain competitive.

I just did some test pricing at another service as a first stab to understand their algorithms, and they don't even explain their algorithms. There's no law against checking what the prices are for very very expensive designs at other services. You'd be cheating yourself if you didn't. The first thing I noticed was that the lowest price appeared to be $5.90 regardless of model size for small model. How about two of those models connected by a sprue? Also $5.90! I have as of yet not passed any test models through their system to get a better idea of what they are doing for large models with or without voids and varying sizes of holes to access the voids. I wouldn't be surprised if their algorithms are similar to what Shapeways is doing but with different thresholds and constants in their equations.

I think machine space 'rental' would be a fantastic option to have. I'd be curious to see if they charged based on the average density of full print trays, how much that would cost us and how likely it is that experienced designers could make it worth their while.

I'm okay with the fact that large models with the 40mm hole get a discount - but models under 40mm, which are the reason that 40mm hole rule makes sense for the larger models, don't get to share that discount even though they are doing their part to make it possible for that discount to exist. If you look at the big picture, more large models with 40mm holes and less models under 40mm means more wasted material and more people receiving discounts. Shapeways will either eat costs or have to raise prices over all, causing yet another stir. As the pricing structure is, I predict a lot less small scale models being printed in WSF.

Did they take projected changes in model types and sizes into consideration when crunching the numbers? They can't predict the future but they had to know that there would be less small-scale models as a result - even just looking at the number of models being printed that would end up costing more than their markup would give them a ballpark figure to work with. Because it seems to me that if enough small-scale modelers leave the scene, Shapeways is going to have a bunch of people designing large models with holes in them to receive discounts and not enough small models to fill them. We can assume or hope it never gets to this point, but those with small scale designs have been hit the hardest and are most likely to stop using the service entirely, print in other materials, or start designing larger models instead, changing the entire dynamic of the pricing structure once again.

It's not a hard and fast rule that models smaller than 40mm in size have gone up in price. Most of my models went down in price (although that surely doesn't apply to all classes of models and designers), and when I ran test cases of certain scaled geometries for various sizes many also seemed to now be cheaper for small sizes with even bigger drops when the 40mm aperture size benefits kicked in. Unfortunately they no longer show the "old" price so now reminiscing about the old days is a rather moot point.

It's likely that the distribution of model sizes has shifted somewhat. A big impact might be due to multiple element models connected with long sprues to lower cost. From what I've seen a long connecting sprue, which minimizes added machine/material space, will cost a fraction of the $1.50 for an extra part reducing that cost, and you can repeat that some reasonable number of times depending on the parts. Too many big fluffy models like that could actually prevent larger, denser, and more profitable models from fitting into a tray. But depending on how smartly they pack their trays all they need to do is to wait for enough models to be purchased until a tray hits a specific packing goal with a proper mix of model types and sizes. That goal could be based on packing efficiency, absolute revenue, or whatever. I have no idea about the specifics of their factory numbers but this issue is probably the least of their worries, IMHO.

I suppose we'll have to wait and see. It won't be the least of their worries if the combination of orders that start coming in costs them money ;-)

I definitely see more sprued and looped (when possible) models popping up in the future. So far the only way I've found to bring costs on my models down is looping a certain combination of parts together and printing the rest as is. I save some on labor, incur extra material cost for the loop, and some more cost is added for machine space between the pieces. With most combinations, this only raises the price. On one or two that I tried, it dropped a few dollars at best. These methods tend to work best on files with a lot of parts since you're adding material and machine space cost to cut back on labor, which is only going to help a file with enough parts that fit together closely enough to offset the added costs.

Overall, I'm not complaining. I've still managed to make my parts somewhat affordable, but on prop runs that I used to make a grand, I'm making closer to $200 or $400 now, after raising prices on the kits significantly. A kit I was able to sell for around $40 before would now be around the $60 mark. It's just lucky for me that my customers are all DIY prop builders and are willing to pay not just for a product but have immense appreciation and consideration for the amount of time it takes to design these things. And I'm sure with some combination of spruing and tight packing multiple of the same part together I can save money (since these are run based and I don't sell them through Shapeways).