Erdi - from your image it looks like you are generating the synthetic data using a zero-phase wavelet but are then picking using a first-break approach (assuming that the arrow represents the pick on the 4th trace). I would suggest either picking on a peak using a threshold or modelling using a minimum phase wavelet. As Rob says, you could shift the data down with sushift but I suspect that the modelling has not created any data before T=0 so this might not work.

Another alternative would be to set a negative start time in the modelling if this is an option.

Regards John Chamberlain

On 06/08/20 06:59, rob@xsgeo.com wrote:

I would use sushift to delay the traces and then subtract the delay from the subsequent picks generated by sufbpickw

Regards

Rob

*From:*Seisunix seisunix-bounces@mailman.seismic-unix.org *On Behalf Of *Erdi Apatay *Sent:* Wednesday, 5 August, 2020 11:30 PM *To:* seisunix@mailman.seismic-unix.org *Subject:* [Seisunix] VSP Processing

Hello All

I am trying to pick first break arrivals in Zero Offset VSP on the Crewes toolbox and another software. My problem is that I start picking from 4th trace because I cannot see the first three traces when I look at my data, which results in wrong velocities.

My data is passive zero-offset VSP data, utilizing ambient noise sources. I created this by using a modeling code.

I added one of my data and the problem's screenshot

Can I fix this on Seismic Unix?

---

Seisunix mailing list Seisunix@mailman.seismic-unix.org https://mailman.seismic-unix.org/listinfo/seisunix