Stefano Bertoncello from Italy posted at his blog an article about Nimbus’ techniques to play old 78s.

http://twogoodears.blogspot.com/2010/02/nimbus-prima-voce-series.html

http://www.wyastone.co.uk/nrl/pv_transfer.html

I think have problems with it. The upper frequency response of the old 78 is somewhere around 5kHz -8kHz. The horn looks like a hyperbolic 70-80Hz horn from 1.5cm-3cm. This type of horn shall act as low pass filter and shall roll off a few dB even at 5kHz. Generally I is a good idea to get out of the 78s sound as much HF as possible.  So, I do not like the idea that Nimbus' proposes. In my view a full spectra of 78s bandwidth can’t be handled by a single horn. It is fine to pick up sound from horn with microphone but I would like to see different method. The sound from the disk shall be recorded via microphone twice. First run with short HF horn and second run with bass horn as Nimbus is doing now. Then the timing marks of the first (HF) and second (LF) runs need to be synchronized and signal dubbed as sum of both runs. The key in it is to find the right analog high-pass filter for the HF run. In my view with proper setting of this filter and with very high level of speed stabilization I think it would allow to get an ultimate 78s reading…