I use 0.1% Sudan Black B in my brain immunofluorescence staining, even though it reduces autofluorescence of background but also gives low non-specific staining. It looks like neuronal marker staining.
Well 1) I think my previous answer doesn't make sense, because in case of immunofluorescence, SBB should rather quench the lipofuscein signal and 2) it might be the filter combination you're using, Sudan Black performs worse in red and far-red. You could try a sudan black analouge such as TrueBlack, which should be more specific against autofluorescence (but is also more expensive). Worked well on my sections, except for the far-red channel, where it didn't improve the overall signal-to-noise ratio.
But from your images (those are withour antibody-stain, right?), it doesn't really look like you improve the signal-to-noise ratio with the SBB protocol.
I don't know about your staining protocol, but if you want to improve the signal-to-noise ratio, you could also take measures to increase the signal rather than reduce the noise, i.e. different antibody concentrations, longer incubation times, signal boosting kits...
Both are with Anti-Dopamine Receptor 2 Antibody (1:500). The photos show midbrain. In references there are not many receptors like on this photo with SBB. For me more authentical is without SBB. Because of that I claim that SBB gives non-specific staining and I don't know why? Of course, SBB should reduce autogluorescence of the backgrand, and it is, but also gives (extra) low-staining of cells in the background.
In my protocol, after secondary antibody incubation, 0.1% Sudan Black B was incubated through 20 min, then washed three times.