Difference in Differences with Lags and Leads

[Erdem Yilmaz]

I have a panel dataset for songs which can be remixed at different time periods. I am interested in the question of whether derivative products create a upstream competition for their parent products. Here is my baseline regression:
Yit = α + β1 × Treatment*postit + γi + δt + ϵit, where Yit is number of daily streams, γi is song and δt is time fixed-effects.
I also want to run a placebo test by adding leads and lags, similar to Autor (2003). As far as I understood, these leads and lags are the interactions of the treatment variable with time dummies, rather than actual leads and lags. Currently, I am using reghdfe command rather than xtreg,fe as my dataset is huge and I need to add time fixed effects. Here is the code that I use when I run difference in differences analysis:

reghdfe log_daily_streams treatment, absorb(song_id day) vce(cluster song_id)

Here is how my data looks like:

song_id  n_streams       day                 treatment
   1           30            May01              0
   1           34            May02              0
   1           32            May03              0
   1           32            May04              0
   1           56            May05              1
   1           67            May06              1
   1           45            May07              1
   2           23            May01              0
   2           25            May02              0
   2           27            May03              0
   2           27            May04              0  
   2           28            May05              0  
   2           23            May06              0
   2           23            May07              0

How could I run something similar to what Author(2003) has done?