Tweaking Sample Sets

I have the nasty history of purchasing sample sets and never really adjusting them to my room and speakers. I get busy playing them and only stop for individual notes that might need adjustment. I want to start concentrating on maximizing the sound potential of my sets, but I'm not sure where to start. What steps do the rest of you follow to tweak your ranks and notes? Do you use any particular "voice" as a base for comparison in setting volume levels for other ranks, and how do you adjust brightness or balance across the keyboard range? I am particuarly concerned with the volume/brightness across the full keyboard range. Thanks in advance.

John

I don't tweak. I use ARC to correct for room resonance and that's it.

Same for me. ARC2 (which nowadays costs around EUR 200) comes with a measurement microphone, and works as a plugin, I use it with Reaper which costs around EUR 60.

On the Dutch forum I just learnt about REW, which is free. I experimented a bit with REW. Similar to ARC2, you need a measurement microphone to record a frequence sweep of every speaker. Next, REW calculates what corrections you need to do. Finally, I import the values that REW has calculated into the ReaEQ plugin of Reaper.

If this sound complicated, sorry But it is worthwile. The sound improves dramatically after correction.

If the same notes need tweaking in multiple ranks that is a reason to look at overall equalization, be it via ARC or REW, the latter being my preference when looking at multichannel setups. I'm blessed with a sloping ceiling and a complex room shape which helps a lot with those issues, especially in the pedal range.

The tweaking that this posts asks about is a different matter. I tweak a lot of individual notes, more in the direct, dryer samples than in diffuse samples as they tend to blend better in the natural acoustic. It's also possible that the more diffuse samples get more voicing attention by sample set creators, but I have no way to know if this is the case.

For notes that are too high or low in level I almost always find that the interface level meters correspond with what I hear. That is, the level variations are really present in the sample set and not a product of room resonances.

I hear more problems with extreme variations in the brightness of individual notes. Whether this is the product of over aggressive sample manipulation I cannot say. In any case I find myself manipulating the brightness parameters nearly as much as the aptitude (level).

I avoid modifying overall rank levels unless I have a special reason such as producing a stronger tuba or trumpet if a big trumpet is not available. Otherwise I tend to hope the sample set producer makes careful notes about the relative SPL of various ranks and keeps the different ranks at the correct relative levels. I am a technical person first and a musician second so I don't presume to "correct" the balance of a principal chorus as an example.

I tend to tune only the worst offending notes when I run into them. I have questions at times as to whether celesting ranks maintain their correct frequency relationships as they begin to howl in the higher octaves. So far I have just been careful to not use much other than a string celeste when playing in the upper registers.

I'll be interested in hearing other observations.

Hi,
Here are some of my voicing hints and guidelines, (I hope they are somewhat clearer than mud).
I always start with the 8' Principal (Diapason, Montre) in the principal division (Great, Hauptwerk, .) and get its volume right for the room and then set the other stops in relation to it)
Avoid using major and minor scales. The dominant and tonic and leading tone can have the psychological effect of causing some notes to sound louder and softer even though they aren’t.
Use instead, all of the following, which don’t have a sense of tonal center and therefore will not cause a note to sound stronger or weaker just because of its position in the scale.
1. Whole tone scales, need to use two to cover all the notes.
2. Fully diminished seventh chords, need to use three to cover all notes.
(C-eflat g flat a; c3- e-g b flat; d-f-a flat-b)
3. Augmented triads, need to use four to cover all the notes.
(C-e-g#, C#-f-a; d-f#-a#; e flat, g-b)
4. Throughout using each of the above, you can use a chromatic scale to make sure all the individual notes are matching up. You don’t want to voice the entire C and C# side of a stop only to find out the C# you started on was a little soft compared to the C you started on and therefore you now have an entire C# side that is too soft.

Baroque style voicing will be more even as the scales, chords ascend the register.
Romantic voicing tended to have “treble ascendency”, i.e. the scales would get slightly louder toward the treble end, the idea being that music that placed a greater emphasis on accompanied melody needed the treble end to sing out over the lower register. BUT BE CAREFUL, it is easy to overdo this, it needs to be subtle.

Setting balance between stops can be a matter of taste and style, however in general you want the higher pitched stops to be somewhat softer than the main 8' stops, more so in Romantic style organs, in baroque style the stops would be closer the to the same volume across the board i.e. tenor c of the 2 would be almost as loud as treble c of the 8'. However, this is a very general statement and much variation can be sound in practice.
You can hold notes and chords of the 8' in the treble or middle octave, then compare to the same note or chord of the 4' and octave lower and the 2' yet another octave lower. Check to see that each stop can be heard contributing something to ensemble (with the exception of the softest strings, Flauto Dolces, Dulcainas, etc.) I.E. Draw the 8', 4' and 2' Principals and make sure the 8' and 4' flutes can be heard when you add them.
Build up the ensemble and make sure each stop is adding what you want as you go along, i.e. 8' Flute, 4' flute, 8' principal, 4' octave, 2' flute,(if there is one) 2' octave, Mixture(s) Reed(s)
If it is a romantic style instrument, check that each 8' is making its contribution to the Fonds, i.e. Bourdon 8', Salicional (Gambe) 8', Harmonic Flute 8', Montre 8' and even 8' Oboe in a swell division.

Finally, Hold some notes, chords, open fifths in each octave of the keyboard and build up the ensemble to make sure every stop is making the same contribution to the overall ensemble in each octave. For example, you may find that although the 4' Octave, or Mixture sounded even throughout its compass as you were voicing it, but that it ends up making a much stronger contribution in the tenor octave than the treble octave, You will have to decide whether to gradually taper up the treble end or taper down the lower end.
In HW you can use the brightness level to adjust individual notes that seem too dull or bright, but you will need to go back and check the volume (amplitude) as changing the brightness will affect the overall volume).
John

John - exactly what I needed, thank you. I have used calibration devices for speaker adjustments, but your note on treble ascendency was most valuable. I find many of my sets "muddy" in the lower octaves and somewhat too loud in the upper. I am anxious to get started this evening following your advice.