All posts relating to Maxwell Render 1.x
User avatar
By aitraaz
#103808
sweet pic :)
User avatar
By psanitra
#103812
Hi Tom

I can see text on your image saying: Maxwell Render RS1+

Does it means we will have the Render Settings RS1/RS2/Preview like we had in early alpha?
Will we be able to save our own render settings?
(you can move this into official question topic if you like)

Thanks
User avatar
By mverta
#103817
psanitra wrote:Hi Tom. I can see text on your image saying: Maxwell Render RS1+ Does it means we will have the Render Settings RS1/RS2/Preview like we had in early alpha?
Will we be able to save our own render settings?
(you can move this into official question topic if you like)

Thanks
I'll answer this... it doesn't have to do with render settings. It's one of our internal code numbers for which engine we're using for that test. When Tom said it was done with an "early test version", he means one of the engine revisions after RC4 but before 5. So like, I think this image was done with what could be considered RC 4.2 or something. In fact, if we rendered this image today, it would actually be even sweeter.

_Mike
By JDHill
#103822
...not for long...
By ajlynn
#103823
Romans wrote: - There is no translator or support for Multi/Sub-object materials. This is really a serious problem if you want to render trees for example (like xfrog-plants). It is not a handy option to re-assign Maxwell materials to every part of imported 3rd party tree-object. Is there a solution for this?
I've used multi/sub with Maxwell. You need to have all the subs as Maxwell materials - either go in and change them all or instance them all into their own slots then run the covert script: http://www.maxwellrender.com/forum/viewtopic.php?t=5512

Since a multi is just a container for a bunch of materials you can't really have a Maxwell multi - just a multi full of Maxwells.

For the camera issue - just do what everybody else does and make your camera target at the same Z value as your camera. (This is the same thing you'd do with a regular camera...)
User avatar
By Jozvex
#103825
Ernest Burden wrote:Have you seen anyone making an app to create/edit IES files?
Just a few days ago on CGTalk we were talking about creating IES files, and eventually we found this (direct link):

http://www.design.leftshift.ru/ies_generator3.exe.rar

It's called IES Generator and it's great! Pretty simple program but it works, and it shows you interactive renders of the light as you make it. It's a very hard program to find because it's Russian and was only posted on some Russian forums by the guy who made it.

After you unzip the file from the RAR, you need to remove the .share (I think?) extension off the end of the name.

;)
User avatar
By lsega77
#103835
Jozvex,

that ies app is pretty cool, thanx for the link! :D
User avatar
By Brett Morgan
#103848
Great link jozvex, nice to see you in the maxwell forums again :)

Brett
By joie
#103870
Mverta;

Is there any chance you could post the rendertime of the "announcement threads" images?, specially the "cornell boxes" ones.

Thank´s
By Peter_K
#103885
Mike,

Could you try a simple "bounce check" scene in RC5? Like a labyrinth with one of the walls being a emitter, looking from above? I´ve looked at the cornell box you posted, and to me it looks uhm, not as "real" as the beta rendered boxes I´ve seen, and I feel the front of the smaller cube too dark. But maybe its because of low luminosity settings on the emitter?

/Regards Peter K
By Romans
#103894
Adam wrote: I think you're misunderstanding something; pixel dimension is the only relevant output setting for a render engine. dpi is only relevant for printing.

This is true too. But image-quality is also defined by resolution (i.e. 1600x 1200 by 300 dpi). You will lose quality if you have to resample the renderings (with 72 dpi?) for printing (generally 150 to 300 dpi or more) in Photoshop. So it would by very helpful to have an option to SAVE the renderings with a certain image-resolution. This has nothing to do with the settings for pixel dimension for the render engine (its a post setting).

Regards Roman
By pixelarq
#103903
Hello, Romans
Romans wrote: - Cameras: the Maxwell camera produces 3-point perspectives, that means, that the vertical edges (lines) are converging with height. This is not very useful for architectural images there vertical lines have to remain vertical (2-point perspectives). In 3ds max exists a so called camera correction modifier that resolves this issue (works fine with scanline and also with mental ray). But this helpful modifier cannot be applied to a Maxwell camera. Please inform me if there will be a solution in Maxwell.
I'm just another user and not NL crew but I think Maxwell will not work with 2-point perspectives nor orthogonal views.
It's just not possible as it's not possible with a camera on real world. That's the simple Maxwell's phillosophy: expect real-life realism AND limitations. 8)
You can make some hand-made correction with photoshop or similar software.
User avatar
By max3d
#103908
Romans wrote:Adam wrote: I think you're misunderstanding something; pixel dimension is the only relevant output setting for a render engine. dpi is only relevant for printing.

This is true too. But image-quality is also defined by resolution (i.e. 1600x 1200 by 300 dpi). You will lose quality if you have to resample the renderings (with 72 dpi?) for printing (generally 150 to 300 dpi or more) in Photoshop. So it would by very helpful to have an option to SAVE the renderings with a certain image-resolution. This has nothing to do with the settings for pixel dimension for the render engine (its a post setting).

Regards Roman
Roman, you really misunderstand this issue. The dpi setting is needed for some software and printjobs, but it's just a setting. There is no relation with the image itself. You could if you wished, change it in a postscript file with a text editor (I used to do that in the early days of DTP).
Just load the image in Photoshop and change the dpi setting. No resampling needed. You'll just see the dimensions in cm/inches change.
By pixelarq
#103909
Ernest Burden wrote: Sunlight + glass = architectural visualization

There more to it than that, of course, but without that, you're not getting off the launchpad. I acknowledge that you did make a definitive statement, but on the chance that you are right, and MWR1.0 does not properly handle light through glass--it will make Maxwell next to useless for the entire architectural market. I am qualified to make that statement. It must be in 1.0

I know everyone has their 'make or break' features, but this is a big one. I've managed almost my entire career without SSS and normal mapping. But without sunlight through glass? Can't do my job.
I vote for this. Please fix sun, sky, clipmap and sun-glass.
Without those any render engine is almost useless for archviz.
By pixelarq
#103911
Hi Ernest,
Ernest Burden wrote:
ajlynn wrote:For the camera issue - just do what everybody else does and make your camera target at the same Z value as your camera. (This is the same thing you'd do with a regular camera...)
The old way to deal with it was to render a much larger area and crop in post. If you can't pan then that's about it. So you have to render a lot of pixels that you will just throw away. Not good.
Why not render a region? On 3dsmax plugin we have an option to render only the pixels defined by a bitmap. Does it work?
  • 1
  • 3
  • 4
  • 5
  • 6
  • 7
  • 57
render engines and Maxwell

"prompt, edit, prompt" How will an AI r[…]