[Paraview] Animation with sudden large intervals between consecutive frames

Tue Jan 15 02:40:36 EST 2013

Good morning,

Well the caching was not the issue. The animation I started creating last
evening has thus far only written two 3 frames, so I will uncheck the
offscreen rendering next and try again. We have been playing with the
offscreen rendering before, but I beleive that had to do with stability
issues in one of the previous versions.

Tom

2013/1/14 Utkarsh Ayachit <utkarsh.ayachit at kitware.com>

> Interesting. I am not sure what it could be. Another option is to try
> disabling the use of offscreen rendering for screenshots (by
> unchecking "Use Offscreen Rendering for Screenshots" in the "Settings"
> dialog on the "Render View" page (at the bottom of that page).
>
> Utkarsh
>
> On Mon, Jan 14, 2013 at 11:06 AM, Tom Fahner <tom.fahner at gmail.com> wrote:
> > Hello Utkarsh,
> >
> > Well in fact I have used both. I had used a python script before, which
> gave
> > a similar behavior, but in that case the unusual large intervals were
> about
> > 15 minutes instead of 8 hours. I also noticed that there was something
> wrong
> > in my setup using the python script, so when I redid the animation I
> simply
> > used the GUI to check the setup more rigorously and decided not to use
> the
> > python script.
> >
> > A bit more info on the python script:
> >
> > With the python script I also assumed it may have been related to the
> actual
> > screen being rendered at the moment that the large intervals showed up.
> This
> > because I used VNC to log into the machine for rendering and the screen
> was
> > locked with ParaView running for most of the time. I just unlocked the
> > screen occasionally to have a look at the progress from time to time and
> I
> > could imagine this may have influenced the rendering. Later I also saw a
> 15
> > min. interval between two pngs which were created in the middle of the
> > night, meaning the screen was probably still locked, so I assumed that me
> > unlocking the screen could not have been the reason. Fifteen minutes
> between
> > two frames can be accepted if this happens once in 100 frames or so, but
> 8
> > hours is just too much.
> >
> > At the moment I started the animation from the GUI again, but with the
> > "Cache Geometry" unchecked as you suggested. I will check the results of
> > that setting tonight.
> >
> > Regards,
> > Tom
> >
> >
> > 2013/1/14 Utkarsh Ayachit <utkarsh.ayachit at kitware.com>
> >>
> >> Tom,
> >>
> >> I wonder if caching is causing these issues. Are you using the
> >> ParaView GUI or a Python script? If  the GUI, can you disable caching
> >> (uncheck "Cache Geometry" from the "Animation" page in the "Settings"
> >> dialog.
> >>
> >> Utkarsh
> >>
> >> On Mon, Jan 14, 2013 at 5:50 AM, Tom Fahner <tom.fahner at gmail.com>
> wrote:
> >> > Dear all,
> >> >
> >> > For some reason I noticed that ParaView has some large intervals
> between
> >> > writing a PNG for consecutive frames of an animation. Normally there
> is
> >> > about one minute between the timestamp of two PNGs, but sometimes
> there
> >> > is
> >> > suddendly a gap of 8 hours:
> >> >
> >> >  141K Jan 13 15:56 New_Volume_.0118.png
> >> >  141K Jan 13 15:58 New_Volume_.0119.png
> >> >  139K Jan 13 15:59 New_Volume_.0120.png
> >> >  139K Jan 13 16:00 New_Volume_.0121.png
> >> >  138K Jan 14 00:34 New_Volume_.0122.png
> >> >  139K Jan 14 00:35 New_Volume_.0123.png
> >> >  139K Jan 14 00:36 New_Volume_.0124.png
> >> >
> >> > There are multiple occasions where these large intervals happen but
> this
> >> > does not happen at a regular interval of the animation. There does
> seem
> >> > to
> >> > be a relation with the amount of memory that is used, since there is a
> >> > sudden decrease in memory used right after "Jan 14 00:34". Below I
> will
> >> > describe my animation and setup. I hope anyone can give an explanation
> >> > for
> >> > the behavior.
> >> >
> >> > During the weekend I have created an animation of a mixing tank using
> >> > volume
> >> > rendering of the concentration of some chemicals in the tank. It is a
> >> > reasonably large CFD simulation performed with OpenFOAM. There are
> about
> >> > 15
> >> > million cells (tetrahedrals and prismatic layer), but not extremely
> >> > large. I
> >> > saved the concentration every 0.25 seconds for a 120s simulation. We
> >> > have
> >> > ParaView 3.98.0 installed on this machine and it was the only program
> >> > running at the time. I have made the animation with ffmpeg after
> >> > ParaView
> >> > made the frames as consecutive pngs. Besides the volume rendering of
> the
> >> > concentration, the walls of the tank where shown with a fixed opacity
> of
> >> > 0.3
> >> > and the internal structure (some rotors and baffles) where present as
> >> > well.
> >> > Although the simulation use the MRF concept, I did mimick the rotation
> >> > of
> >> > the rotors using the transform filter. Using the "cool to warm" preset
> >> > for
> >> > visualization I could nicely set the opacity to 0 when the
> concentration
> >> > was
> >> > in the allowed range and it showed red in case of too high
> >> > concentration, of
> >> > blue when too low. The resulting animation is satisfactory, I just
> >> > wonder
> >> > what can be done to make sure these large intervals between writing
> >> > images
> >> > do not happen.
> >> >
> >> > Hope someone can help, if you need more information, please let me
> know.
> >> >
> >> > Regards,
> >> > Tom
> >> >
> >> > --
> >> > T.C. Fahner
> >> > e: tom.fahner at gmail.com
> >> >
> >> >
> >> > _______________________________________________
> >> > Powered by www.kitware.com
> >> >
> >> > Visit other Kitware open-source projects at
> >> > http://www.kitware.com/opensource/opensource.html
> >> >
> >> > Please keep messages on-topic and check the ParaView Wiki at:
> >> > http://paraview.org/Wiki/ParaView
> >> >
> >> > Follow this link to subscribe/unsubscribe:
> >> > http://www.paraview.org/mailman/listinfo/paraview
> >> >
> >
> >
> >
> >
> > --
> > T.C. Fahner
> > e: tom.fahner at gmail.com
> > t: +31-6-52642814
> > a: van Lodensteynstraat 24
> >     2612 SE Delft
> >     Netherlands
>

-- 
T.C. Fahner
e: tom.fahner at gmail.com
t: +31-6-52642814
a: van Lodensteynstraat 24
    2612 SE Delft
    Netherlands
-------------- next part --------------
An HTML attachment was scrubbed...
URL: <http://www.paraview.org/pipermail/paraview/attachments/20130115/68cf25e6/attachment.htm>