[Radiance-general] Desktop radiance Glare Analysis: Pcond

[Greg Ward]

That's a good practice, and one I follow as well.  Since pcond output is meant to be sent to a printer or low dynamic-range display device, converting it to a lower bit-depth at the same time makes perfect sense.  Some of the programs that convert to such formats have built-in tone-mappers that are similar to pcond but less sophisticated and faster.  See the ra_bmp '-e' option and the normtiff command.

-Greg

> From: Christopher Rush <Christopher.Rush at arup.com>
> Date: February 27, 2013 6:31:58 AM PST
> 
> I would suggest it is good practice when using pcond -h to save the image as a low dynamic range file rather than .pic or .hdr, since pcond is meant to create nice looking pictures, not useful data.
> e.g.
> pcond -h image.hdr | ra_ppm | convert ppm:- image.jpg
> 
> That way you never mix up .pic or .hdr files that contain adjusted and incorrect data values.
> 
----------
> From: "Sim Swee Ling Annie" <anniesim at surbana.com>
> Date: February 27, 2013 1:25:43 AM PST
> Hi all, 
> 
> I was trying to run glare analysis via Desktop Radiance for glazing with 2 different VLT: 
> 
> a) VLT = 30 
> b) VLT = 60 
> 
> The type used for these glazing are kept as default --> "glass". 
> 
> 1) For the first analysis, I have adjusted the exposure for both images in order to reflect more accurate visual view showing the differences between clear and darker glass. 
> 
> The glare index via UGR shows: 
> 
> a) VLT 30 = 36 
> b) VLT 60 = 44 
> 
> 2) For the second analysis, on top adjusting the exposure, I have also used "pcond -h" on the images. 
> 
> The glare index via UGR shows: 
> 
> a) VLT 30 = 0 
> b) VLT 60 = 0 
> 
> Looking at the comparison of the results above, I am not sure whether "pcond -h" should be used prior to glare analysis. Should the glare analysis be done on raw images without "pcond -h"? ......although it was stated that "pcond -h" will create a view that mimic human responses. 
> 
> Would really appreciate any sharing or help on this! 
> 
> Thanks in advance! 
> 
> Cheers, 
> 
> Ann