Dear Group
I have a question concerning the sequence for using pfilt and pcond. Since both of these programs actually adjust the exposure, it's hard for me to understand which approach is best. Any help would be useful.
Thanks
Marcus
_________________________________________________________________
Instant message with integrated webcam using MSN Messenger 6.0. Try it now FREE! http://msnmessenger-download.com
Marcus Jacobs wrote:
Dear Group
I have a question concerning the sequence for using pfilt and pcond. Since both of these programs actually adjust the exposure, it's hard for me to understand which approach is best. Any help would be useful.
As I understand it*, pfilt is primarily for image scaling and exposure adjustment (handled automatically if you use rad), whereas pcond is more of an image tweaker, manipulating the pixels in the image to make up for dynamic range shortcomings in various media, or color correction.
What are you interested in doing with your images? As I mentioned, rad is pretty good at getting the image exposed & scaled properly. I usually just run my images through pcond -h after that, and end up with a satisfactory image. But as with most aspects of Radiance, there is fantastic option complexity, if desired.
*Still learning, statements could be rubbish!
Adding my 2 cents to what Rob said...
Pfilt is an important part of the rendering process, performing anti-aliasing necessary to reduce the appearance of sampling noise and artifacts. The output from pfilt is generally the one you want to save or archive. Exposure adjustment is really a very minor function to pfilt, and is completely reversible.
Pcond is specifically for preparing the output for display or printing, and the resulting file should probably not be saved, or if it is, the input given to pcond should not be deleted. This is because, unlike pfilt, the exposure adjustment pcond does is not reversible. Often, I send the output of pcond directly to ra_tiff or another image converter, since pcond brings the entire dynamic range of the original into the 0-1 limits of most formats and displays.
-Greg
From: "Marcus Jacobs" <[email protected]>
Date: Wed Oct 1, 2003 11:38:32 AM US/PacificDear Group
I have a question concerning the sequence for using pfilt and pcond. Since both of these programs actually adjust the exposure, it's hard for me to understand which approach is best. Any help would be useful.
Thanks
Marcus
Thanks for the info Rob and Greg
One thing that I have questioned though is whether it would be better to use pfilt only for anti-aliasing and filtering and leave exposure adjustment only to pcond if I am dealing with a high dynamic range scene. The way I understand it* is that pfilt uses a 2 pass method (by default) to properly expose the image. The way I am reading into this is that pfilt in a way corrects the exposure of the picture output by rpict. To use pcond after pfilt essentially rexposes the exposed picture. Would it be better to use a filtered but unexposed picture as an imput to pcond?
Marcus
* I am only a layman (higher than a newbie but not an expert) so take this statement with a grain of salt :>)
From: Greg Ward <[email protected]>
To: [email protected]
CC: "Marcus Jacobs" <[email protected]>
Subject: Re: Sequence for using pfilt and pcond
Date: Wed, 1 Oct 2003 12:43:19 -0700Adding my 2 cents to what Rob said...
Pfilt is an important part of the rendering process, performing anti-aliasing necessary to reduce the appearance of sampling noise and artifacts. The output from pfilt is generally the one you want to save or archive. Exposure adjustment is really a very minor function to pfilt, and is completely reversible.
Pcond is specifically for preparing the output for display or printing, and the resulting file should probably not be saved, or if it is, the input given to pcond should not be deleted. This is because, unlike pfilt, the exposure adjustment pcond does is not reversible. Often, I send the output of pcond directly to ra_tiff or another image converter, since pcond brings the entire dynamic range of the original into the 0-1 limits of most formats and displays.
-Greg
From: "Marcus Jacobs" <[email protected]>
Date: Wed Oct 1, 2003 11:38:32 AM US/PacificDear Group
I have a question concerning the sequence for using pfilt and pcond. Since both of these programs actually adjust the exposure, it's hard for me to understand which approach is best. Any help would be useful.
Thanks
Marcus
_________________________________________________________________
Get MSN 8 Dial-up Internet Service FREE for one month. Limited time offer-- sign up now! http://join.msn.com/?page=dept/dialup
Hi Marcus,
I generally don't use the -2 option of pfilt for computing the exposure, even though it's the default. I generally know the exposure correction for my scene by the time I reach the filtering stage, so I just set it explicitly with the -1 and -e options. This is what rad does as well when you specify an EXPOSURE= value (as you should).
In either case, the exposure that's set by pfilt is completely reversible and doesn't affect the time taken or output produced by pcond. As a side note, using pcond without the -h option is pretty much the same as displaying the image with ximage -e auto.
-Greg
From: "Marcus Jacobs" <[email protected]>
Date: Wed Oct 1, 2003 1:54:48 PM US/PacificThanks for the info Rob and Greg
One thing that I have questioned though is whether it would be better to use pfilt only for anti-aliasing and filtering and leave exposure adjustment only to pcond if I am dealing with a high dynamic range scene. The way I understand it* is that pfilt uses a 2 pass method (by default) to properly expose the image. The way I am reading into this is that pfilt in a way corrects the exposure of the picture output by rpict. To use pcond after pfilt essentially rexposes the exposed picture. Would it be better to use a filtered but unexposed picture as an imput to pcond?
Marcus
* I am only a layman (higher than a newbie but not an expert) so take this statement with a grain of salt :>)