Sat Feb 09, 2013 11:36 am
Whilst it is true that resampling to an octave below or indeed to an octave above GENERALLY doesn't make a valid stop, there are always exceptions. I have successfully created many pitch shifted stops in this manner which sound absolutely correct in terms of transients, harmonics and releases.
However, it is very rare to find a stop that transposes successfully by pitch shifting alone and usually it is necessary to edit the harmonics and adjust the attack by various techniques in order to produce a seamless and natural continuation of the stop to its new extremes. However, it can be done.
It is possible also, to create a valid and convincing extended stop by using the bottom octave of another similar stop transposed down and then tacked on to the original with the break carefully matched. However, it does take time and lots of experimentation to achieve any valid result.
In reality, what is far more difficult is to duplicate a convincing continuation of the stop's stereo field regardless of how successful any pitch shifting might be. Fortunately many bottom octaves are on separate chests and in different positions so distinguishable breaks are less evident since they occur naturally in the first place. It might not be 100% authentic but it can be sufficiently close to be convincing.
All of what I have written above applies for the most part to dry samples or at least samples recorded fairly close with little ambience. When wet samples are used, regardless of any success in pitch shifting per se, the additional factor of a completely altered reverberation characteristic has to be taken into account and this will definitely make a convincing result hard to achieve.
However, going back to the original; question, the pitch shift here is not extreme and apart from merely pitch shifting the final note to accommodate the additional two notes, one method would be to take the E3 and let that serve as E3 and F3 then take F3 and let it serve as F#3 and G3. If it is in stereo and in sides then the different field will cover once the sides are swapped.
Alternatively, something like D#3 can be pitch shifted to the F# and E3 up to G3. this will create a more believable individual sample effect. Whether you wish to alter the length of the pitch shifted sample is up to you - sometimes it is effective but at other times it can introduce strange artifracts depending on the editing software used.