MikeA

You bought a dollar bill object? Isn't that... A rectangular plane?

Keep up the good work - I'll be trying this out! If you can add a the ability to have a user specified number divisions on the arrows - evn better : )

I think not.

So, did Mr. CEO say anything interesting? I missed it.

I think the critical questions you have to ask on a job like this: How close are you going to get? What will you be able to see in the shot? Then you can model to the appropriate level of detail.

Workplanes are a really powerful and important feature for me too. I'm surprised I don't see them in use more often. I did wonder if they were available in Blender...

A simple but very handy script... thank you Cairyn!

What would be the Python code for 'No scaling' ?

I've always used custom system builders in the past, but having recently stripped and rebuilt a couple of systems - power supply /motherboard / CPU change + all the bits and pieces - I was pondering building the next one myself. But I do agree: with the price of Threadripper parts - it's mightly expensive if you screw up - and the risk is right in your hands.

This is coming from someone who has never used Octane - or Cycles - so bear that in mind : ) I do use Redshift (Rs) as my main renderer. The difficulty with these sorts of threads is that any opinion is very much coloured by one's own use case. I can totally understand where Rectro is coming from because the areas that are especially import for him are somewhere way down the roadmap. Here are some random thoughts to consider: 1. All renders have limitations / weak points. What are your important needs? 2. The key advantage of Rs is speed - both in look development / IPR and final frame rendering. 3. Reliability / crashiness is difficult to quantify. In lookdev general usage I find it OK. In final rendering runs it's been rock solid for me - and it renders millions of polygons without breaking a sweat. 4. It's easy to get a decent level of quality in Rs, but it takes more work - tweaking settings etc - to get to a very high level of quality. 5. I think it's reasonable to say that the Rs team are not great at estimating timescales for development. Things frequently get pushed back, or take much longer than first suggested. 6. On the positive side Rs issue updates pretty frequently - maybe every month or so on average. As well as bug fixes most of these have some form of tweak or new feature. You can see a roadmap of planned development here: https://trello.com/b/QASr74yB/redshift 7. The downside of this constant development is that things get broken. On the positive side they are often fixed on the next update. 8. The documentation is a weak spot IMO. 9. As Redshift is now owned by MAXON it's reasonable to imagine that might deliver some advantages to the C4D user over other third party renderers in the future.

I was thinking of building a 3970 system - a first full build, and you guys are making me nervous

If I'm understanding correctly, I don't think this is possible in Rs. I'd recommend you post on the official Redshift forums for a definite answer. There has recently been a major re-write of the Rs shading core to enable various new rendering tech, and I think that the whole area around 'volume rendering' is one that Rs acknowledges as weak, so hopefully this might get attention sooner than later. - but I would definitely check on the Rs forums.

How are the rendertimes if you completely hide the glass spheres?

Fundamentally what's happening is that the displacement is distorting the texture mapping - so the texture is shifting / swimming on the surface slightly. To fix: Look in the Rs object tag on your sphere > Geometry > Reference section - right at the bottom. Set to Object and then drag your sphere into the object slot. I think that should fix it.

This is the ultimate question. Answer: set all your values to 42. : )