Re: png output -- very long processing time?

From: wxtofly <wxtofly_at_nyahnyahspammersnyahnyah>
Date: Thu Jul 22 2010 - 16:38:13 MDT
Rick
Thanks for investigating,

The speed difference will make the direct png output unusable for most of my purposes.
Using ctrans then imageMagic to convert the ncgm files is much faster, as far as I can tell. 
The process I run is real-time and the hope was that saving a step or two would speed things up.
The better final image quality of direct png's is nice.  Makes me wonder if I can't tweak how I run imageMagick's convert.

TJ

On 07/22/2010 02:49 PM, Rick Brownrigg wrote:
I ran some quick tests of PNG vs. NCGM, and PNG was definitely slower,  
by about 4 times in my case. I can imagine the difference varies  
greatly by the visual complexity of the plot; the PNG driver must  
perform rasterization and anti-aliasing for everything, whereas I  
believe NCGM is a vector based format.  30 times though -- wow.

Is this a question out of curiosity, or is this speed difference  
causing you grief?

Rick

On Jul 22, 2010, at 3:25 PM, wxtofly wrote:

  
Hi,

Have I set it up wrong, or is png output expected to be about 10-30  
times slower than output to ncgm?

The times below are from start of ncl to finish of each plot and are  
thus culmulative.
As ncgm:
(0)     --- Single Parameter Plot end - Thu Jul 22 07:44:18 PDT 2010  
WallClockElapseTime= 1 sec
(0)     --- Single Parameter Plot end - Thu Jul 22 07:44:19 PDT 2010  
WallClockElapseTime= 2 sec
(0)     --- Single Parameter Plot end - Thu Jul 22 07:44:20 PDT 2010  
WallClockElapseTime= 3 sec
(0)     --- Single Parameter Plot end - Thu Jul 22 07:44:20 PDT 2010  
WallClockElapseTime= 3 sec
(0)     --- Single Parameter Plot end - Thu Jul 22 07:44:20 PDT 2010  
WallClockElapseTime= 3 sec
(0)     --- Single Parameter Plot end - Thu Jul 22 07:44:21 PDT 2010  
WallClockElapseTime= 4 sec
(0)     --- Single Parameter Plot end - Thu Jul 22 07:44:22 PDT 2010  
WallClockElapseTime= 5 sec
(0)     --- Single Parameter Plot end - Thu Jul 22 07:44:22 PDT 2010  
WallClockElapseTime= 5 sec
(0)     --- Single Parameter Plot end - Thu Jul 22 07:44:23 PDT 2010  
WallClockElapseTime= 6 sec
(0)     --- Single Parameter Plot end - Thu Jul 22 07:44:24 PDT 2010  
WallClockElapseTime= 7 sec
(0)     --- Single Parameter Plot end - Thu Jul 22 07:44:25 PDT 2010  
WallClockElapseTime= 8 sec
(0)     --- Single Parameter Plot end - Thu Jul 22 07:44:25 PDT 2010  
WallClockElapseTime= 8 sec

As png:

(0)     --- Single Parameter Plot end - Thu Jul 22 14:03:14 PDT 2010  
WallClockElapseTime= 27 sec
(0)     --- Single Parameter Plot end - Thu Jul 22 14:03:28 PDT 2010  
WallClockElapseTime= 41 sec
(0)     --- Single Parameter Plot end - Thu Jul 22 14:03:39 PDT 2010  
WallClockElapseTime= 52 sec
(0)     --- Single Parameter Plot end - Thu Jul 22 14:03:52 PDT 2010  
WallClockElapseTime= 65 sec
(0)     --- Single Parameter Plot end - Thu Jul 22 14:04:05 PDT 2010  
WallClockElapseTime= 78 sec
(0)     --- Single Parameter Plot end - Thu Jul 22 14:04:18 PDT 2010  
WallClockElapseTime= 91 sec
(0)     --- Single Parameter Plot end - Thu Jul 22 14:04:29 PDT 2010  
WallClockElapseTime= 102 sec
(0)     --- Single Parameter Plot end - Thu Jul 22 14:04:44 PDT 2010  
WallClockElapseTime= 117 sec
(0)     --- Single Parameter Plot end - Thu Jul 22 14:04:56 PDT 2010  
WallClockElapseTime= 129 sec
(0)     --- Single Parameter Plot end - Thu Jul 22 14:05:08 PDT 2010  
WallClockElapseTime= 141 sec
(0)     --- Single Parameter Plot end - Thu Jul 22 14:05:20 PDT 2010  
WallClockElapseTime= 153 sec
(0)     --- Single Parameter Plot end - Thu Jul 22 14:05:35 PDT 2010  
WallClockElapseTime= 168 sec
(0)     --- Single Parameter Plot end - Thu Jul 22 14:05:51 PDT 2010  
WallClockElapseTime= 184 sec

Might this have to do with the background image question of my prior  
post?

TJ
_______________________________________________
ncl-talk mailing list
List instructions, subscriber options, unsubscribe:
http://mailman.ucar.edu/mailman/listinfo/ncl-talk
    
_______________________________________________
ncl-talk mailing list
List instructions, subscriber options, unsubscribe:
http://mailman.ucar.edu/mailman/listinfo/ncl-talk

  


--

TJ's Blanchard Page    My Wx Widget  Ft Ebey (Whidbey) NWS Seattle ASP  
: MAPS/RUC2 | Lo Alt | MM5 Bham Obs Fire Winds Rasp Soundings
Winds Aloft   R♥sp    Visible Satellite  Fcst Discussion   Fraser Valley  Windgrams text
Blanchard Digital | ♥Meteogram ♦UW Satellite   | Loop ♥ Discussion With Images Windcast Blipspots
  Quick Radar NWS Radar 9 Meso-stations   Pressure gradients WA Meso Stations map Time/hght/wind
MesoW GoogleMaps aprs current Personal Wx Stations Wind Map now, sfc winds  |  bl winds


  RASP Regional Atmospheric Soaring Predictions for PNW      || Average BL wind loop Blanchard   
       Surface wind forecast Loops: | Animated Radar   |  | loop Blanchard | loop PNW | loop FtEbey | loop Tiger | loop FraserValley

       Layered by altitude Sea Level to 850mb (approx 5000 ft) looped by time wind forecasts:
Blanchard Pressure Level Loop | Tiger Pressure Level Loop | Fraser Valley Pressure Level Loop
All day hour-by-hour text RUC2/MAPS soundings
These have no graphics and are designed to be printed to take with you to the site.
After sunclipse there will be one run for the next day but it only has MAPS soundings.
  Woodside/Sumas/Bridal     Black Mt     Big Johnson     Stewart same as Blanchard     Ft Ebey     Skagit Valley     Tiger  
Blipsp⊕ts: | Baldy | BJ | Black | BlackLZ | Blanchard | Chelan | Iron | Saddle | Stewart | Tiger | FtEbey | Woodside | Bridal Falls
| Current Rapid Update Soundings for flying sites | Winds aloft Seatac & Vancouver | Visible Satellite zoomed | NWS forecasts |
Abbreviated pages for mobile devices. | Blanchard | Tiger | Ft Ebey | East Side
The Seattle Forecast Discussion "translated" or with maps or original form without maps (available sooner) Learn to read the maps. Soundings Tutorial
Cloud Forecast -- RUC based Meteogram for KBLI Bellingham    |   RUC based Meteogram for CYXX Abbotsford -- Cloud Forecast

TJ Olney Weather: Site Map

_______________________________________________
ncl-talk mailing list
List instructions, subscriber options, unsubscribe:
http://mailman.ucar.edu/mailman/listinfo/ncl-talk

raspicon16.png llapse1.png fireSky.gif discuss.jpg windgramico.png wbarb3.png
Received on Thu Jul 22 16:38:30 2010

This archive was generated by hypermail 2.1.8 : Wed Jul 28 2010 - 10:42:06 MDT