From: Dylan Penhale <dylanpenhale@(email surpressed)>
Subject: Max time in Batch Frames
   Date: Sun, 20 Nov 2005 21:20:05 -0800

When rendering Batch Frames, does Max Time check the overall batch or the
individual frames contained within? 

If it checks the over all batch how hard would it be to modify the submit to
check the individual frames contained within each batch?

Dylan Penhale
Systems Administrator
Fuel International
65 King Street
Newtown
Sydney
NSW 2042
 
Phone:  xxxxxxxxxx
Mobile: xxxxxxxxxx
Email:  Dylan@(email surpressed)
Web:    www.fuel-depot.com
 
___________________________________________________________________
The information in this email is confidential and may be privileged. 
The contents may not be disclosed, copied, printed, forwarded or 
disseminated, or used by anyone other than the addressee. If you 
are not the intended recipient, please contact the sender of this 
e-mail. His Best cannot accept any responsibility for the accuracy 
or completeness of this message has it has been transmitted over 
a public network. If you suspect that the message may have been 
intercepted or amended please call the sender immediately.
  

Dylan Penhale wrote:

When rendering Batch Frames, does Max Time check the overall batch or the

individual frames contained within?

	Overall batch.
	As far as rush is concerned, it counts the time the render script
	starts running until it ends. Everything inbetween is unknown to rush.

If it checks the over all batch how hard would it be to modify the submit to
check the individual frames contained within each batch?

	You'd have to add code to the script that watches the output of the renderer
	as it runs to determine the time one frame stops and the next starts.

	It would depend entirely on the output from the renderer in question,
	esp. the verbosity settings.

	For instance, you could start the renderer in a pipe (eg. open(PIPE, "Render -r sw .."))
	and read the pipe passively to watch for frame completion messages, and while
	watching, keep track of time. If too much time passes, you could kill the
	pipe, and requeue.

	In such a case, you'd probably want to leave the rush MaxTime: turned off,
	and make your own field "Frame Maxtime:", parse that, and set that as the
	timeout in the custom code that monitors the pipe.

	It's tricky stuff, and completely dependent on the renderer's output
	messages, and how well the renderer flushes its output (different on
	Windows and Unix). Also, in the case of submit-maya, the output can be
	radically different, as maya might be using Mental Ray to render, who's
	output is completely different from Maya's software renderer.

	But if you're using lightwave or some such, where the output is maybe
	more consistent, then it might be better suited.
	

--
Greg Ercolano, erco@(email surpressed)
Rush Render Queue, http://seriss.com/rush/
Tel: (Tel# suppressed)
Cel: (Tel# suppressed)
Fax: (Tel# suppressed)