model or ps

[Jonathan]

As no expert to the architectural visualisation field I wondered how more experienced users would tackle object repetition. For example, working on a wine shop interior rendering do I model every bottle of wine or should this be done in photoshop?

[erickdt]

I generally look at "Photoshopping" things in as a work around that I'll use only when there is no time to re-render. To me, it is much easier to map, say multiple different wine bottle labels, in MAX then it is to **** around with the image in Photoshop which could take literally hours depending on the complexity if the shot (and still not look quite right).

 

Just my 2c...

[Claudio Branch]

Use a distribution tool such as Scatter, Particle System, or Forest Pro 3 in tandem with proxies and you can tackle pretty much anything repetitive.

 

Here is a scene with 16+ million polys using Forest Pro 3 as the distribution method for trees, brush, riprap, and driftwood debris.

[alexg]

I'd say model it, then you'd have a bunch of wine bottle library to throw into your next projects.

 

If you decide to photoshop it, it would take the same amount of time to find the most appropriate one. And what happens if the client decide to 'change some of it' ?

[3D_IC]

Lighting is also a big issue with photoshop. If the light on the comped object isn't just right then it looks odd. Best to model the wine bottles and then they react to the scene lights and may also cast replections, shadows etc. The labels can be added in PS if they are not good enough.

[dean@pikcells]

You gotta think about time here. It may be quicker photoshopping 1 image, but if the client changes something such as the view, then it would have been quicker to model it properly first time. Plus modelling stuff generally gives far better, and more realistic results.