# Volume rendering composite datasets
**URL:** https://discourse.paraview.org/t/volume-rendering-composite-datasets/1073
**Category:** ParaView Support
**Created:** [December 20, 2018, 8:51pm UTC](https://discourse.paraview.org/t/volume-rendering-composite-datasets/1073 "2018-12-20T20:51:24Z")
**Posts on this page:** 4
**Page:** 1
### Author:  [@Andy\_Bauer](https://discourse.paraview.org/u/Andy_Bauer)
#### Post date: [December 20, 2018, 8:51pm UTC](https://discourse.paraview.org/t/volume-rendering-composite-datasets/1073/1 "2018-12-20T20:51:25Z")
Hi,
I’m getting confused with which types of vtkDataSets stored in composite datasets can be volume rendered. If I have a multiblock of vtkImageDatas the GUI doesn’t provide me with the volume representation. For a multiblock of vtkUnstructuredGrids though I do get the volume representation option but it doesn’t appear correct in the following image:
I’ve attached the state file from ParaView 5.6 [volume\_and\_surface\_representations.pvsm](https://discourse.paraview.org/uploads/default/original/1X/8025885a174d7e2febf65996048800f88182c9ab.pvsm) (681.8 KB) to help show the issue. For a multiblock of vtkStructuredGrids I also get the volume representation option but nothing is shown in the 3D render view and I get the following error when I try to use it:
> ERROR: In /home/buildslave/dashboards/buildbot/paraview-pvbinsdash-linux-shared-release\_superbuild/build/superbuild/paraview/src/VTK/Common/ExecutionModel/vtkCompositeDataPipeline.cxx, line 164
> vtkPVDataRepresentationPipeline (0x6989ef0): Can not execute simple algorithm without output ports
> ERROR: In /home/buildslave/dashboards/buildbot/paraview-pvbinsdash-linux-shared-release\_superbuild/build/superbuild/paraview/src/VTK/Common/ExecutionModel/vtkExecutive.cxx, line 782
> vtkPVCacheKeeperPipeline (0x65782f0): Algorithm vtkPVCacheKeeper(0xfc30e30) returned failure for request: vtkInformation (0x6578860)
> Debug: Off
> Modified Time: 9853171
> Reference Count: 1
> Registered Events: (none)
> Request: REQUEST\_DATA\_OBJECT
> ALGORITHM\_AFTER\_FORWARD: 1
> FORWARD\_DIRECTION: 0
I’m not sure if it matters or not but for the multiblock of vtkStructuredGrids the extents are nearly the same for each block even though the actual geometry of each block doesn’t overlap. This dataset is a probably a bit big to share here but I have a similar dataset that I need to get to @cory.quammen anyway that can be used to duplicate the issue.
Thanks,
Andy
---
### Author:  [@Antech](https://discourse.paraview.org/u/Antech)
#### Post date: [December 24, 2018, 8:32am UTC](https://discourse.paraview.org/t/volume-rendering-composite-datasets/1073/2 "2018-12-24T08:32:46Z")
Hello. I noticed this issue even in old versions like 5.2. Seems like i’s there since that times (around 2015). ParaView not always handles multi-block datasets, there are similar issues with other filters (like it doesn’t “see” some variable although Auto Convert is on until you apply Extract Block).
---
### Author:  [@wascott](https://discourse.paraview.org/u/wascott)
#### Post date: [January 2, 2019, 1:19am UTC](https://discourse.paraview.org/t/volume-rendering-composite-datasets/1073/3 "2019-01-02T01:19:07Z")
Andy,
I may be missing something, but you know that ParaView only volume renders the first block for unstructured grids? (Yep, my users get caught with this one all the time.) Use Merge Blocks filter, then volume render. Can is perfect for replicating this issue. Since the first block is the plate (i believe), all variables are the same, and ParaView doesn’t volume render as you would expect.
Andrew,
If you could post an example of other filters not seeing variables from different blocks, I could take a look at it? Be sure to include a dataset, and detailed instructions…
Alan
---
### Author:  [@Andy\_Bauer](https://discourse.paraview.org/u/Andy_Bauer)
#### Post date: [January 2, 2019, 12:43pm UTC](https://discourse.paraview.org/t/volume-rendering-composite-datasets/1073/4 "2019-01-02T12:43:21Z")
Thanks Alan. That’s probably the same behavior for a multiblock of image data as well as that explains what I’m seeing in the provided pvsm file.