Hi ​@MqMq 

Let me know if these resources help you:
Working with Raster Bands: GitHub
 

Hi ​@elyhienrich ,

First of all, thank you for your reply.

I only downloaded the reflectance images and I cannot download visual images now. Therefore, I have to use reflectance image to get the true-color image. Do you have any other recommendation?

Hi ​@MqMq 

There are two Python-based approaches to render true color from PlanetScope reflectance imagery.

Option A — Using Sentinel Hub Process API (PlanetScope ARD in BYOC)

• This uses an evalscript that maps red/green/blue to RGB and scales reflectance (DN ~ 0–10000) to 0–1 with a small brightness boost. Replace your BYOC collection ID, bbox/geometry, and time interval as needed. 

Option B — Local GeoTIFF (Analytic SR) with Rasterio and NumPy

• If you already have PlanetScope analytic SR GeoTIFFs (scaled by 10000), map bands to RGB and apply a simple display stretch. Try clipping and a non-linear curve for best visual results; a simple linear scaling works as a start. 
  
   

I would recommend using the Ask Ai widget you can find in the banner here on Planet Community, you can share how your bands are named/ordered and whether you’re using local GeoTIFFs or Sentinel Hub BYOC, and the AI widget can tailor the python code to your exact setup.