# Detector Noise For data simulation

**URL:** <https://ask.igwn.org/t/detector-noise-for-data-simulation/1104>\
**Category:** Data Analysis\
**Created:** [September 23, 2024, 10:40am UTC](https://ask.igwn.org/t/detector-noise-for-data-simulation/1104 "2024-09-23T10:40:09Z")\
**Posts on this page:** 4\
**Page:** 1

<div class="post-metadata">

**Author:** ![Rao](https://avatars.discourse-cdn.com/v4/letter/r/54ee81/32.png) [@Rao](https://ask.igwn.org/u/Rao)\
**Post date:** [September 23, 2024, 10:40am UTC](https://ask.igwn.org/t/detector-noise-for-data-simulation/1104/1 "2024-09-23T10:40:09Z")

</div>

How one can download the detector noise (without signals or glitches) from all 3 observations ?

---

<div class="post-metadata">

**Author:** ![jonah](https://yyz1.discourse-cdn.com/flex031/user_avatar/ask.igwn.org/jonah/32/10_2.png) [@jonah](https://ask.igwn.org/u/jonah)\
**Post date:** [September 23, 2024, 11:17pm UTC](https://ask.igwn.org/t/detector-noise-for-data-simulation/1104/2 "2024-09-23T23:17:31Z")

</div>

@Rao Detector data is available through the GWOSC website. For example, data from O3b can be downloaded from this page:

> **[Archive for O3b\_4KHZ\_R1 dataset](https://gwosc.org/archive/O3b_4KHZ_R1/)**
>
> Gravitational Wave Open Science Center

I’m not sure what it means to get noise without signals or glitches. When the data are recorded, the noise, glitches, and signals are all mixed together. You can look for times away from signals, and/or subtract out best fit waveforms. But, glitches are present in most data segments.

If you instead prefer purely Gaussian noise, you should simulate noise. Here are some examples of how to do that:

[https://pycbc.org/pycbc/latest/html/noise.html](https://pycbc.org/pycbc/latest/html/noise.html)

---

<div class="post-metadata">

**Author:** ![Rao](https://avatars.discourse-cdn.com/v4/letter/r/54ee81/32.png) [@Rao](https://ask.igwn.org/u/Rao)\
**Post date:** [September 24, 2024, 4:50am UTC](https://ask.igwn.org/t/detector-noise-for-data-simulation/1104/3 "2024-09-24T04:50:01Z")

</div>

@jonah I’am very glad for your reply, I understand that even if we download data the times before signals there might be chance of containing glitches, so I’ll prefer subtract out best fit waveforms.  
do we have any tutorial for that ? or any video demonstration?  
I have found out this paper, [Glitch subtraction from gravitational wave data](https://iopscience.iop.org/article/10.1088/1361-6382/acd0fe/ampdf)

---

<div class="post-metadata">

**Author:** ![jonah](https://yyz1.discourse-cdn.com/flex031/user_avatar/ask.igwn.org/jonah/32/10_2.png) [@jonah](https://ask.igwn.org/u/jonah)\
**Post date:** [September 24, 2024, 7:27pm UTC](https://ask.igwn.org/t/detector-noise-for-data-simulation/1104/4 "2024-09-24T19:27:45Z")

</div>

@rao There are few different tutorials that show how to construct a best fit waveform, which you can subtract from the data.

See the waveform tab, here:  
[http://peviewer.igwn.org](http://peviewer.igwn.org)

Or, you can see another approach, here:

> <https://github.com/gwosc-tutorial/Data_Guide/blob/master/Guide_Notebook.ipynb>

Good luck!