From the plain figures, noise-wise the Rigol appears to be 10dB worse than the Siglent. But there is one little detail that may affect the relation somewhat: The Rigol DSA accepts 10dB higher signals at the mixer before producing significant distortion than the Siglent. This means, at the same signal level, you can run the Rigol with 10dB less attenuation than the Siglent which puts the noise level approximately on par. Of course, at very low signal levels, the Siglent is better off. Also, the Siglent SSA is the more modern machine with all the benefits of better semiconductors and the like. Anyway, this thread isn't covering the differences of the Rigol vs. Siglent spectrum analyzers, I just wanted to point out in my previous post that there still may be more questions than answers regarding the front-end designs.
Cheers,
Thomas