[Radiance-general] 360 degree rendering

[Raghuram Kalyanam]

Thanks Greg, that was very helpful. It worked. 

Best Regards,
Raghu

> On Jul 1, 2017, at 3:00 AM, Gregory J. Ward <gregoryjward at gmail.com> wrote:
> 
> Hi Raghu,
> 
> I had a look at your files.  Thanks for providing a link.
> 
> Line 48 of the file Components/Group_1.rad has a duplicate real argument count, which causes this file not be be included:
> 
> 	## guessed Material
> 
> 	void    glass   glass1_Group_1
> 	0
> 	0
> 	3       3       0.75 0.75 0.75
> 
> Deleting the second '3' above fixes the issue.  If this file was generated automatically, you should talk to the people who wrote that software.  It's also apparent that they aren't putting the proper dependencies into the "scene.rif" file.  This should include lines like:
> 
> 	objects= Materials/[Color_M06].mat Materials/[Color_M05].mat Materials/glass3.mat Materials/glass1.mat Materials/Default_Material.mat
> 	objects= Materials/[Color_M00].mat Materials/Lisanne_Boots.mat Materials/[Color_M04].mat Materials/Color_M1.mat
> 
> ...and so on.
> 
> Also, the "guessed" materials are terrible.  Most of them have 50% specular reflectance, which makes them metallic.  This is probably the main cause of noise in your images.  If you put in more reasonable material values, preferably with specularity less than 0.04 or so, your results will improve substantially.  Reserve specularities above 0.05 for metallic materials.
> 
> Cheers,
> -Greg
> 
>> From: Raghuram Kalyanam <kalyanam at rhrk.uni-kl.de>
>> Date: June 30, 2017 9:39:19 PM GMT+06:30
>> 
>> Hi Greg,
>> 
>> Thanks for answering, here are the files. I have tried out rvu also to check out, if normal image also the same (grainy). I guess you are right, it could be with the light sources. I am using only direct sun light. you can find it in sky.rad file.
>> 
>> The below command 
>> rvu -vtc -vh 360 -vv 130 -vp 0 0 1.25 -vd 1 1 0 -n 4 -ab 7 -as 1024 scene.oct
>> results  this image
>> 
>> Thanks & Regards,
>> Raghu
>> 
>>> On Jun 30, 2017, at 4:58 PM, Greg Ward <gward at lmi.net> wrote:
>>> 
>>> If you send me your input files and exact commands in a private message, I'll have a look. There's nothing in your parameters to explain your grainy results. My only guess at this point is that your light sources might be right triangles rather than rectangles, which screws up source sampling.
>>> 
>>> Cheers,
>>> -Greg
>>> 
>>> Sent from my iPad
>>> 
>>>> On Jun 30, 2017, at 9:08 PM, Raghuram Kalyanam <kalyanam at rhrk.uni-kl.de> wrote:
>>>> 
>>>> Hi Mark,
>>>> 
>>>> Thanks for answering, I tried with IPD 0 and its still same. I changed the default to also zero in cal file. Any idea how do i improve the graininess of the picture?
>>>> 
>>>> Best Regards,
>>>> Raghu
>>>> 
>>>>> On Jun 30, 2017, at 4:27 PM, Mark Stock <mstock at umich.edu> wrote:
>>>>> 
>>>>> Raghu,
>>>>> 
>>>>> An easy, though somewhat inefficient, way to get a single-eye view
>>>>> without digging into the script would be to change the IPD to 0.
>>>>> 
>>>>> Mark
>>>>> 
>>>>>> On 6/30/17, Raghuram Kalyanam <kalyanam at rhrk.uni-kl.de> wrote:
>>>>>> Thanks Greg. That was helpful. I generated the VR images. Could I get only
>>>>>> get the 360 degree projection only for one eye  using that cal file, if so
>>>>>> which variable should i tweak. Here
>>>>>> <https://seafile.rlp.net/f/5c3a3a6aeecb42faa3de/?dl=1> is a sample image i
>>>>>> wanted to produce.
>>>>>> 
>>>>>> Also there is lot of noise in the picture and appears grainy. I increased
>>>>>> -ab but it didn't improve much. i tried other parameters, here is a glimpse
>>>>>> of the command.
>>>>>> 
>>>>>> X=2048; Y=2048; cnt $Y $X | rcalc -f view360stereo.cal -e
>>>>>> "XD:$X;YD:$Y;X:0;Y:0;Z:1.25;IPD:0.06;EX:0;EZ:0" | rtrace -dp 4096 -ar 256
>>>>>> -ds .02 -dj .9 -dt .05 -dc .75 -dr 3 -ss 16 -st .01 -ab 5 -aa .03 -ad 1024
>>>>>> -as 512 -lr 12 -lw 1e-5 -x $X -y $Y -fac scene.oct  > out4.hdr
>>>>>> <https://seafile.rlp.net/f/7f343d2d476e43b2b546/?dl=1>
>>>>>> 
>>>>>> Best Regards,
>>>>>> Raghu
>>>>>> 
>>>>>>> On Jun 26, 2017, at 4:57 PM, Greg Ward <gregoryjward at gmail.com> wrote:
>>>>>>> 
>>>>>>> Hi Raghu,
>>>>>>> 
>>>>>>> Mark Stock created the attached file "view360stereo.cal" that makes a 360°
>>>>>>> stereo panorama in the standard format for such viewers.  This was checked
>>>>>>> into the Radiance HEAD back in January.
>>>>>>> 
>>>>>>> Cheers,
>>>>>>> -Greg
>>>>>>> 
>>>>>>> <view360stereo.cal>
>>