# S receiver functions Peaks

**URL:** <https://discourse.obspy.org/t/s-receiver-functions-peaks/1715>\
**Category:** Related: rf\
**Created:** [May 22, 2023, 10:24pm UTC](https://discourse.obspy.org/t/s-receiver-functions-peaks/1715 "2023-05-22T22:24:43Z")\
**Posts on this page:** 3\
**Page:** 1

<div class="post-metadata">

**Author:** ![rafa0110](https://discourse.obspy.org/user_avatar/discourse.obspy.org/rafa0110/32/575_2.png) [@rafa0110](https://discourse.obspy.org/u/rafa0110)\
**Post date:** [May 22, 2023, 10:24pm UTC](https://discourse.obspy.org/t/s-receiver-functions-peaks/1715/1 "2023-05-22T22:24:43Z")

</div>

Hello!

I am working with the SRF and I have noticed that the move-out correction causes some peaks for later times(stream3c.moveout(phase=‘Sp’)). I think it is caused specifically by the stretch\_delay\_times() .  
Are this peaks just expected because of the stretching manipulation of the trace? If they are artifacts noise, do you know how to remove it? Will be reasonable to just low-pass again to get rid of them?

I tested different values for the slowness, but it didn’t disappear. On my script. I am using the slowness=6.4.

> stream3c.rf(method=‘S’, deconvolve=‘time’)  
> stream3c.moveout(phase=‘Sp’)  
> stream.extend(stream3c)

Here are some examples of my issue:

 ![4](https://discourse.obspy.org/uploads/default/original/1X/83ef30b9ce7b88d63fccbe4db5439d1de8f50983.png)

 ![SRF_4C.BAJ7m](https://discourse.obspy.org/uploads/default/original/1X/0cf8b3ffe97c9a7651e77e986ab25117421bde49.jpeg)

Thanks.

---

<div class="post-metadata">

**Author:** ![trichter](https://discourse.obspy.org/user_avatar/discourse.obspy.org/trichter/32/226_2.png) [@trichter](https://discourse.obspy.org/u/trichter)\
**Post date:** [May 24, 2023, 9:27am UTC](https://discourse.obspy.org/t/s-receiver-functions-peaks/1715/2 "2023-05-24T09:27:36Z")

</div>

Hi! Nor sure what is going on. If you see these peaks before the moveout (at different times) everything is working as expected. Looks more like a wrap-around effect from the deconvolution. Couldn’t you simply use a longer time window for the response components and trim again after the deconvolution?

---

<div class="post-metadata">

**Author:** ![rafa0110](https://discourse.obspy.org/user_avatar/discourse.obspy.org/rafa0110/32/575_2.png) [@rafa0110](https://discourse.obspy.org/u/rafa0110)\
**Post date:** [May 24, 2023, 11:24am UTC](https://discourse.obspy.org/t/s-receiver-functions-peaks/1715/3 "2023-05-24T11:24:59Z")

</div>

Sure! I will try! Thanks Tom!!
