Maxwell Render | Official public forum

It might be great for studio work but for arch viz IBL just sucks at present! I have spent all day testing - test after test after test to get the levels of each channel correct and again just given up. With materials, one or two tests and you have it nailed, but with IBL you could probably keep going for days and never get there I guess unless using Modo or other with IBL realtime preview.

Again 5 years on battling to one day get a result from IBL I'm still no closer!

I so hoped after the mass of requests for this feature it may have been delivered in V2. As well as sharp shadows from IBL that seems not to work unless you use a black HDRI / MXI with a tiny tiny sun!!!! Though then you have a different map to the other channels so testing, testing, testing!

Richard, it's about the dynamic range of HDR you are using. Maxwell or other engine does not matter at this point. V2 is fully capable of rendering HDR illumination at its best. So what is the dynamic range all about? Simple... 90% of HDR images around the web have only a few stops and they are not giving good quality in lighting because, the difference between paper white and sun white is not high enough.

In this page, there are several free HDRI you can try:
http://www.openfootage.net/?cat=15

The test renders on the page clearly shows "sharp" vs "soft" shadows so you can understand which HDRI has higher range

What are you having most problems with? Getting the background right vs the illumination channel? I'd just render out an alpha and not worry about the background being blown out or too dark.

And like Tom said, most free HDRs you find on the net are crap (and paying ones too). 3 fstops is not really high dynamic range....

Here's a Dosch HDRI (radsky-017.hdr) coming with the installer:

The sharp shadow is faint but it's even visible on object itself.

Here's another Dosch HDRI (dh207.hdr) coming with the installer:

The sharp shadow coming off from the bulb in the room is quite clear.

Here's a very decent HDRI from Moofe (sample01)
http://www.moofe.com/news.php?n=1-000-s ... downloaded


Exporting a Maxwell Physical Sky as HDRI will easily prove it works, too.

I so hoped after the mass of requests for this feature it may have been delivered in V2.

This is already done and is working perfectly in v2

I tend to have troubles getting the images located correctly firstly and secondly as Mihai suggests setting levels correctly between each channel.

Just takes me forever to achieve both with no preview and in the absense of channel adjustment under multilight.

Listen I really heed what you are both suggesting but still it's a wish to see this function included, would make life so much easier!

I've just got to get back to sorting my method in SU to locate images correctly then I'll be at least part way toward some success!

tom wrote:

I so hoped after the mass of requests for this feature it may have been delivered in V2.

This is already done and is working perfectly in v2

Sorry mate I meant the request for multi channel support in ML!

Richard wrote:As well as sharp shadows from IBL that seems not to work unless you use a black HDRI / MXI with a tiny tiny sun!!!! Though then you have a different map to the other channels so testing, testing, testing!

Yes, I was talking about it instead. Because, there's no such multichannel approach in IBL. Why? IBL has only 1 illumination channel and the other 3 are Bkg (which is not a source of illumination), Reflection (this is not an illumination source, too), Refraction (again not an illumination channel). It means, none of these 3 have any tiny bit of contribution to scene lighting so that's why this request cannot be valid.

Hmmmm? Mate I have to disagree!

Bump the reflection channel up and see what it does to windows, metals are other speculars in any scene!

It's increasing the reflected image, so? But, can you illuminate a scene using only reflection channel? Multilight is there for only controlling the illumination, not for tweaking the amount of reflection and others. Makes sense?

Yeah sort of and look I do understand, but still say that the level of the reflection channel and the illumination channel must be set independantly in intensity to get a correct looking result! This needs to be set by testing as it can't be done in ML!

Even as Mihai suggested, don't worry about the background being blow out as it can be photochopped later, if there was ML control of all channels would that be necessary? Why then even have a background channel?

I hear but, see it's a matter of apples and oranges. You can only initially control them independently, not during the render or in the post.

Unbiased renderers can cause severe frustrations!
Don't shoot me, i'm just kidding

tom wrote:I hear but, see it's a matter of apples and oranges. You can only initially control them independently, not during the render or in the post.

That's why it's in the wishlist!!

Richard wrote:That's why it's in the wishlist!!

Yes, and I'm trying to save you wasting time in developing dreams because it's a wish that won't come true. I wish the planet not to be a sphere but it is.

tom wrote:I wish the planet not to be a sphere but it is.

That's what they want you to believe.

I understand the wish, but it really is a matter of post processing and not "Multilight".