At this point, I've reconstructed about a million fossil birds (and other vertebrates) and it pains me that this stuff is sitting in my files, UNPUBLISHED. It's because I >>can't<< make time to finish my textbook, not yet anyway. Let's talk about Paleogene birds. So.... (thread) 
How do you reconstruct Paleogene #birds? The good news is that some of the relevant taxa - eg, those from Messel - are known from articulated skeletons, sometimes with plumage preserved and even feather patterns and melanosomes revealing colour... [pic Gerald Mayr/Volker Wilde]
Based on the skeleton, you take a load of measurements and get a skeleton scaled about right. My drawings at this point are really sketchy and look pretty goofy... This is the tiny Gracilitarsus, a hummingbird-sized relative of hoopoes.
If there are feathers preserved, you take measurements from them too, and add them to the reconstruction. And I also add plumage thickness and details inferred from what's present in living birds...
If colour patterns and colours are preserved on the fossils, they get added to. If not, I make inferences based on living birds that are (a) related and/or (b) ecologically similar to the fossil animal. So we end up with this sort of result (this shows the seriema-like Salmila)..
There's some photoshop editing at this point, wherein I clean up mistakes, smudges (I draw with pencil and don't use fixer) and such. Interesting issue: Paleogene birds often ending up looking really samey! It's funny...
The plumage ends up obscuring the details that made the bird interesting! (this affects ALL feathered #dinosaurs, by the way. The details we like showing in reconstructions would virtually all have been obscured by plumage)... (here's a #Velociraptor)...
Anyway, I've done a ton of these things. They're all done for my in-prep textbook The Vertebrate Fossil Record (#TetZooBigBook), which I've been working on for years and still have yet to finish. If you're interested you can support this project here ... patreon.com/TetZoo
* too
