Measuring the success of a 3D model?

Soldato
Joined
8 Oct 2007
Posts
2,844
Hi,

I'm doing some research for a university assignment which involves 3D modelling. I'm trying to find something that tells me how to measure the success of a 3D model, but can't find anything.

Any links or info on this would be great. Thanks
 
What exactly do you mean by "measuring success"?

A way of knowing that the model was a success. It's hard to define because obviously how good a 3D model looks is always going to be down to opinion to some degree.

For example to measure the success of a retail product you can have a sales count to tell you how successful it is (I guess)
 
Success in which way though?

Just because a product sells well, doesn't mean it is a successful design, it can be down to successful marketing.

You need to find a better definition of what you are measuring!
 
Success in which way though?

Just because a product sells well, doesn't mean it is a successful design, it can be down to successful marketing.

You need to find a better definition of what you are measuring!

I think you're right.

Let's say the aim is to make the model very high in detail and as close to photo realistic as possible.

The end product will not be released or sold in any way.

I still have no idea how to measure the success of that. (Or am I on the wrong track?)
 
The only metric I can think of for a model is its accuracy. Accuracy could be measured via how much of the model is true to the real thing, so say a model is 89% accurate. It deviates in some way from the real thing by 11%. Not a great metric though, so you could refine that.
 
Would a survey be up your alley? ie Here are three images, tell me which is the CG one. Survey 100 people. That kind of thing. Or does it need to be programmed (you didn't mention this)?
 
Ok I do 3d modelling and rendering for a living.

Basically you can't measure the success of a 3D model on it's own because in most cases it's used in conjunction with other materials, brochures, people talking about it that sort of thing.

You also have issues with the way it's rendered, you can have photorealistic versus 'sketch' style and some reports have shown that architecture gets better feedback from less realistic imagery than the photorealistic images. Your idea of a purely photorealistic image may actually give less success because it's the wrong finish for the intended task.

The idea of accuracy (from gord) is fine except when you take into account some initial ideas which are presented to your client don't have measurements to relate to as it's just to show a concept or an idea and so it's scale etc are purely perceptual. You also have characters/models which are distorted deliberately.

Now from a business perspective a successful 3D model is one the client is happy with and is within the current quote and allotted timeframe, and ideally comes back for more lol. If it was for my own scale of success I'd never be happy as I there will always be little things I would like to do differently if I had more time.
 
You first need to define your objectives. Why are you producing this 3D model? What is 'success' in this context?

Once you know your objective, you can define the criteria for whether you have met the objectives or not.

For example, if you are producing extremely realistic animations then you need ways of definition realistic, and this would probably be people's opinions. So you need ways of measuring people's opinions (ratings out of 10, etc).
 
Hi,

I'm doing some research for a university assignment which involves 3D modelling. I'm trying to find something that tells me how to measure the success of a 3D model, but can't find anything.

Any links or info on this would be great. Thanks

What you're researching is too general and does not make sense.

In the world of 3D modelling, the model is as successful as your requirements. There are too many objectives to consider but really you need to define exactly what you're modelling for and the limitations of your project.
 
Hi guys sorry for the late reply, Christmas etc :p

Thanks for the replies, some good info. I think I know what to write now :)

I'm going to use accuracy to measure the success as it's the only thing that is not based on opinion. I'll also be writing about how hard it is to measure the success of a 3d model. And I'll use some kind of survey as an evaluation method.
 
Just wanted to pop in here to say oL - that signature pic is awesome. one of my favourite albums! good man!

and feel free to use me in your survey if you want. i've been working with CG for many years now, so if I can be of some use go for it.
 
Just wanted to pop in here to say oL - that signature pic is awesome. one of my favourite albums! good man!

and feel free to use me in your survey if you want. i've been working with CG for many years now, so if I can be of some use go for it.

Hey, yeah I know it's an awesome album, one of my favourites too, so unique. Thanks for the offer for the survey, I'll contact you when the time comes. Probably won't be for a while, I have so many different assignments to work on right now >.<
 
Back
Top Bottom