# Stream Tracer - from surface

**URL:** https://discourse.paraview.org/t/stream-tracer-from-surface/2996
**Category:** ParaView Support
**Created:** [November 18, 2019, 4:11pm UTC](https://discourse.paraview.org/t/stream-tracer-from-surface/2996 "2019-11-18T16:11:34Z")
**Posts on this page:** 10
**Page:** 1

<div class="post-metadata">

### Author: ![Dec](https://discourse.paraview.org/letter_avatar_proxy/v4/letter/d/e79b87/32.png) [@Dec](https://discourse.paraview.org/u/Dec)
#### Post date: [November 18, 2019, 4:11pm UTC](https://discourse.paraview.org/t/stream-tracer-from-surface/2996/1 "2019-11-18T16:11:34Z")

</div>

Hi All,

Apologies if this has been covered, however I was unable to find it if it is.

I would like to know how to seed streamlines from a surface, specifically any inlet within a solution.

To simplify, if we have a rectangular section with an inlet at one end, is it possible to seed stream lines from that surface rather than using the spherical point location or line source?

Thanks

---

<div class="post-metadata">

### Author: ![mwestphal](https://discourse.paraview.org/user_avatar/discourse.paraview.org/mwestphal/32/17_2.png) [@mwestphal](https://discourse.paraview.org/u/mwestphal)
#### Post date: [November 18, 2019, 4:20pm UTC](https://discourse.paraview.org/t/stream-tracer-from-surface/2996/2 "2019-11-18T16:20:44Z")

</div>

Yes, StreamTracerWithCustomSource.

---

<div class="post-metadata">

### Author: ![Dec](https://discourse.paraview.org/letter_avatar_proxy/v4/letter/d/e79b87/32.png) [@Dec](https://discourse.paraview.org/u/Dec)
#### Post date: [November 18, 2019, 4:49pm UTC](https://discourse.paraview.org/t/stream-tracer-from-surface/2996/3 "2019-11-18T16:49:33Z")

</div>

> [@mwestphal](#):
>
> StreamTracerWithCustomSource

Thanks Mathieu,  
I had not seen it, this software is huge and powerful…

much appreciated!

---

<div class="post-metadata">

### Author: ![Dec](https://discourse.paraview.org/letter_avatar_proxy/v4/letter/d/e79b87/32.png) [@Dec](https://discourse.paraview.org/u/Dec)
#### Post date: [November 18, 2019, 5:56pm UTC](https://discourse.paraview.org/t/stream-tracer-from-surface/2996/4 "2019-11-18T17:56:34Z")

</div>

What are the main contributors to causing it to calculate the very slowly? my grid is less than 1\*10^6 cells, albeit with 3 individual inlets and 1 single outlet…

---

<div class="post-metadata">

### Author: ![mwestphal](https://discourse.paraview.org/user_avatar/discourse.paraview.org/mwestphal/32/17_2.png) [@mwestphal](https://discourse.paraview.org/u/mwestphal)
#### Post date: [November 19, 2019, 9:30am UTC](https://discourse.paraview.org/t/stream-tracer-from-surface/2996/5 "2019-11-19T09:30:30Z")

</div>

How many seeds (points) do you have in your custom source ?  
Are you running ParaView in parallel ?

If you want to have a quick look into the flow, you may want to use the StreamLines representation :

- Tools -\> ManagePlugin -\> StreamLinesRepresentation -\> Load
- Create a new view
- Use the new StreamLine representation instead of surface/outline/etc

---

<div class="post-metadata">

### Author: ![Dec](https://discourse.paraview.org/letter_avatar_proxy/v4/letter/d/e79b87/32.png) [@Dec](https://discourse.paraview.org/u/Dec)
#### Post date: [November 19, 2019, 11:38am UTC](https://discourse.paraview.org/t/stream-tracer-from-surface/2996/6 "2019-11-19T11:38:09Z")

</div>

Yes I see now the influence of the number of seed points is substantial.

I have tried to use the Mask Points function (which can reduce the # of points??) in the workflow;

- celldata.vtu
- CellDatatoPointData
- Mask Points
- StreamTracerWithCustomSource  
- Not sure of the flow here, do you select input source as CellDatatoPointData or MaskPoints?  
- And also do you select Seed Source as your custom seeding surface?

---

<div class="post-metadata">

### Author: ![mwestphal](https://discourse.paraview.org/user_avatar/discourse.paraview.org/mwestphal/32/17_2.png) [@mwestphal](https://discourse.paraview.org/u/mwestphal)
#### Post date: [November 19, 2019, 1:13pm UTC](https://discourse.paraview.org/t/stream-tracer-from-surface/2996/7 "2019-11-19T13:13:03Z")

</div>

> Not sure of the flow here, do you select input source as CellDatatoPointData or MaskPoints?

Input should be celldata.vtu, Seed source should be MaskPoints

---

<div class="post-metadata">

### Author: ![Dec](https://discourse.paraview.org/letter_avatar_proxy/v4/letter/d/e79b87/32.png) [@Dec](https://discourse.paraview.org/u/Dec)
#### Post date: [November 19, 2019, 1:21pm UTC](https://discourse.paraview.org/t/stream-tracer-from-surface/2996/8 "2019-11-19T13:21:01Z")

</div>

Uh so do the mask points filters have to be applied to each surface of interest “for seeding” individually?

---

<div class="post-metadata">

### Author: ![mwestphal](https://discourse.paraview.org/user_avatar/discourse.paraview.org/mwestphal/32/17_2.png) [@mwestphal](https://discourse.paraview.org/u/mwestphal)
#### Post date: [November 19, 2019, 1:25pm UTC](https://discourse.paraview.org/t/stream-tracer-from-surface/2996/9 "2019-11-19T13:25:21Z")

</div>

Maybe the easiest way would be to append all your seed surface together using AppendDataSet filter, then use MaskPoint on that.

---

<div class="post-metadata">

### Author: ![momentumv](https://discourse.paraview.org/user_avatar/discourse.paraview.org/momentumv/32/1686_2.png) [@momentumv](https://discourse.paraview.org/u/momentumv)
#### Post date: [November 26, 2019, 7:25pm UTC](https://discourse.paraview.org/t/stream-tracer-from-surface/2996/10 "2019-11-26T19:25:15Z")

</div>

Late to the party, but I’ve sometimes used a decimate/resample instead of MaskPoint
