[ghost82]

Hi!
I have problems with exporting animation in ensight/ensight HPC.

I have a transient simulation of 2131 time steps.
When exporting from ensight (on win 7 64bit, ensight 10.1.4b), no matter what is the output file (CEI, mpeg, etc.) I have errors related (I think) to video drivers.
The animation is exported sometimes till time step 30, sometimes till time step 100 (it's random), then I receive an error telling me that that video driver has stopped working and it was restored or the cmd windows of ensight is full of error of this type:

Error: CvfChannel::recv: bad type 106 121929937
CvfChannel::recv_int64: short read len 8 actual -1
Client origLen = 0
ERROR: Client job->recv_ints64()

This happens on a first workstation with a gtx titan black and on a second workstation with a quadro 600 (same errors).

Can you help me please?
Is there a way to export from cmd the mpeg animation without open the gui?Maybe in batch mode?

[kevincolburn]

I can forward this onto the developers in the morning to take a look.
In the meantime, my suggestion would be to export as PNG files. You will get a single image per timestep. (You can use EnVe to sew the all of the PNG's back into a movie file quite easily). Exporting as individual images will certainly bypass any video issue if there is one.

[ghost82]

I will try as png but I can tell you that exporting as jpeg has the same errors.
It is a post processing of a barracuda vr simulation with a lot of particles.
Thank you for your help.

[kevincolburn]

If you have the same trouble with single image format output, then there is nothing to do with video driver issues (in terms of export). I would suggest contacting the main EnSight support (either support@ceisoftware.com) or call US +1 919 363 0883 to enlist the main support channel for help to resolve the issue.

-Kevin

[ghost82]

Unfortunately the problem is not solved, even in the last version 10.1.6a
And I cannot understand where the problem comes from: it can be ram, application overflow, server-client connection problem.
And I cannot also send the dataset since it is very big, and the problem appears random, it can be after 10, 100 or 200 time steps.
Sometimes, together with the errors in the console, ensight shows the window of the crash and I can see that the crash happens in Qtcore4.dll, or ntdll.dll, or StackHash_xxxx (xxxx are letters and/or numbers).
It seems that this happens only with dataset imported from barracuda, I have no problems for big fluent dataset.
Now I have 6000 steps (time step 0,1 s).

If you have some large barruda datasets can you test the animation exporting function?

As you can imagine it is very annoying to export 20-50 images at a time when the total is 6000 (multiplyied by the number of variables to animate in each video)
But for now it is the only solution for me.

The crashes happens both on my 2 workstations: both have ecc ram (different brand, but same mobo): the first has 64 gb of ram, the second 32 gb.
Sometimes the last png screenshot is the domain without particles, sometimes not.

I don't know what was the version of barracuda (I'm doing only the post-processing), but I'm quite sure it was the last, v. 17.0.

Another thing is that when I save as png shots, 000 and 001 files are for the same time step!is it normal?

Should I write to support even if I can't provide the dataset?

[kevincolburn]

Couple of questions for you, to try to narrow down what you are doing, and what is causing the problem.

1. Are you Creating a Flipbook first, and saving the flipbook to file? Or saving just the time domain to animation file?
2. What export format are you exporting to ?
3. When you get to the point of saving the animation, do a File -- > save -- > commands from this session. Send me this small text file (.enc), and a screenshot of your File -- > Export -- > Animation window (so I can see what you are doing after the command file).

I have some Barracuda datasets, but only with about 500 steps, not 5000.

-Kevin

[ghost82]

Thank you Kevin for your reply.
I'm post processing a fluid bed reactor and I did this:

- Opened ensight, and load the first case: particles_ResTime.case (this set comprises particles_ResTime.case and particles_ResTime.geo0000 to particles_ResTime.geo5999).
This file has zone "particles" and variable "residence time".
- Load a second case struttura.case (this is the cad file of the external geometry and the dataset comprises struttura.case and struttura.geo)
- Load a third case reti0.case (this is the cad file of internal meshes and the dataset comprises reti.sos, reti0.case, reti0.geo)
- Load a fourth case setti0.case (this is the cad file of other internals and the dataset comprises setti.sos, setti0.case, setti0.geo)

Then, after setting colours, notes, background etc. I saved the context file.
Then file->export->animation: png, check on play, output filename, custom dimensions (744pxx752px) and click OK.

So, answering your first question, no flipbook.
Second answer: png, but it seems the error is independant of the output file format

[kevincolburn]

Ghost82:

I think that we have been able to replicate an issue with a similar type of operation. We are not entirely sure this is the same problem that you are experiencing. If you could send me your contact information, we can raise an official support ticket for this case and be able to provide updated information and resolve the issue.

Regards,

Kevin

[ghost82]

Thank you
I've just sent my personal email.
I'm also attaching some types of errors I encountered during export, maybe they are useful.