An anonymous post I read recently:
“The fact is that the singing on those old recordings was much more flexible in every way (as we have said–cleaner coloratura, ability to decrescendo on high notes, etc.) and it can’t be because of hall size or, apparently, orchestra size, what happened to beautiful singing between then and now?”
Perhaps the the thing that changed the most is the training. People don’t apprentice with masters anymore. They go to postsecondary music schools and have weekly lessons for 30 weeks per year. And how is the content of the lesson different? Well, it seems to be largely comprised of “breath support” and “resonance” taught by “artist teachers” without a rigorous grounding in pedagogy.
Second, I think that the invention of sound recording has led people to try to emulate sounds from outside themselves rather than learning how to make a free sound, from the inside out, under the frequent guidance of a teacher who themselves came from that type of process, and who was their only means of objective feedback. Can you imagine being someone like Patti or Caruso and having all of your training and the beginning of your career without EVER hearing your voice played back to you? Once you were done with your training, how would you know whether you were singing well or not?
Recording practicalities: In the beginning of the era of sound recording, strings did not record well, so there were some pretty strange orchestrations, using small bands of wind instruments, to accompany the singers. The earliest recordings very rarely used a standard opera orchestra. Also the singers were standing right at the opening of the recording horn, making them extremely “out front” in the recording. The lightest of voices could easily be made audible this way without actually raising the level on a specific track (since “tracks” didn’t even exist back then). This would allow for the recording of a very different kind of vocal sound than we are used to. Also, many orchestral instruments are louder than they used to be. Most of the winds are better-engineered, flutes went from wood to metal, oboes went from simple or Viennese to French type, strings went from gut to steel-wound, etc.
Also, some inaccurate, though reasonable, deductions were made about registration after the invention of the laryngoscope and the observation of phonation with it. The two-register idea of the old Italians got sliced and diced many different ways since the 1800s.
The ways in which our musical and scientific development has unfolded over the last century guaranteed that teachers and singers are doing things differently now. We have lost individuality and vocal freedom in our desire to emulate generally heralded ideal sounds.