A ferrite/pearlite microstructure can be obtained by heating the steel above the Ac₃ temperature (often called the upper critical temperature) so that a uniform austenite is formed.  The steel is then cooled in a controlled manner so that ferrite + pearlite is formed, and not martensite or bainite.  Rapid cooling (quenching) produces martensite.  An extended dwell below the Ac₁ (lower critical temperature) but above the Ms temperature (martensite start) can produce bainite.

So, for an alloy steel such as 655H13, the upper critical temperature will be ~ 800 C.  Full annealing or normalizing are two techniques that produce ferrite + pearlite structures.  Normalizing is usually performed by heating about 55 C above the Ac₃ followed by air cooling.  Full annealing is typically performed just above the Ac₃ temperature and then slow cooled (furnace cooled).  The slower cooling of annealing results in high temperature transformation to ferrite + pearlite and coarser microstructures than does normalizing.

Full anneal: 800-850 C then furnace cool
Normalize: 890-955 C then air cool

Keep in mind that a ferrite + pearlite microstructure can be produced by the steel mill after casting and during the rolling of this grade.  The transformation of austenite to ferrite + pearlite is controlled during the rolling process in order to obtain the desired properties (ferrite grain size, pearlite colony size, colony distribution, etc.).

It is true to say that annealing/normalising will generally produce a ferrite/pearlite structure in most steels. However the high nickel and chromium content makes this difficult in this grade of steel. Isothermal treatments are more common provided the temperatures and times are selected carefully. Any thoughts?

I just found some time-temperature transformation curves for this grade (old designation of EN 36) in Atlas of Time-Temperature Diagrams for Irons and Steels from ASM.  I knew that the kinetics would be slow due to the high Ni content, but apparently the tranformation is EXCEEDINGLY slow.  90% transformation from an austenitizing temperature of 860 C requires something like 9 hours at 600 C, or 20-60 minutes at 440 C.  The Ms temperature is around 410 C so this complicates using the 440 C lower limit.

All of this explains why the normalizing temperature may be as high as 955 C for this alloy-- air cooling from this temperature should prevent martensite formation but allow enough dwell time to transform most of the austenite to ferrite + pearlite.  Isothermal transformation definitely looks like a good idea for this alloy, especially if the lower value of 440 C is to be used.  I hope this was helpful.  Feel free to post again if you want to discuss further.