Wed Jan 08, 2020 12:15 pm
I chose to upgrade my current Hauptwerk Windows PC to HW5 in December in order to gain experience with the new version even though I knew that I would be moving everything to a new Mac platform in January. Regarding HW5, many changes and advantages have been described, but I want to share a couple that I was not expecting.
I finished the “migration” to HW5 on my PC late on the evening of 23 December. Rather than spend time devising HW5 audio routings and mixes for an outgoing computer/audio system, I let the program “transcribe” what I had been using. Although it was too late for household purposes for me to sit down and seriously play a sample set through the speakers, I was too curious about the success of the “transcribed” channel assignments to go to bed without testing them. I loaded an old friend, the Marcussen/Moerdijk sample set, and registered a single 8-ft. flute on each keyboard. At modest volume I pressed one central note of the Rygpositiv, and my eyes popped. I had never before heard such “body” and “presence” from this stop. I had been happy with this set since 2010, and being among my oldest sets, it had also been among the most audibly improved sample sets following my inclusion of DiracLive phase/time correction in my system. Now, more than ever before, HW5 made it seem like I could be holding the pipe in my hand. Yes, the sound of each of the three stops emerged from the correct speaker pair, but the big news was the ravishing improvement in sound QUALITY for all three of them.
I learned over the next couple of days that this entire sample set (as well as others) sounds simultaneously brighter and clearer, making the previous HW4 sound seem to have been relatively “washed out.” Everything was more substantive, “real.” The NATURE of the improvement in sound was similar to the improvement evoked by DiracLive, but it apparently takes place upstream of where DiracLive can have an effect.
It is possible that Hauptwerk 5, with its extensive audio capabilities, is able to circumvent an unknown (misphased?) deficiency in my Windows 7 system. Remember, the only modification I made was to replace HW4 with HW5: the General and Organ settings, including polyphony, remained the same. I will never know whether the same contrast between HW4 and HW5 could occur in my upcoming Mac system. In any case, it is encouraging and rewarding to know the possibility of such a profound, unexpected HW5 improvement.
Unfortunately, there is more to the story. I did learn of one apparent HW5 detriment, or at least complication, to the Marcussen/Moerdijk sample set. The sole 8-ft. flue of the Brystvaerk, a Traegedackt, is aberrantly affected by its Tremulant under HW5 (but not 4). Between the F# and G above middle C (MIDI 66/67) there is a large jump (2-fold?) in volume going up the scale IFF the Tremulant is drawn. In other words, the apparent loudness of the individual notes at and above MIDI 67 toggles with the Tremulant status: off = normal, on = (too) loud. The effect is stronger for some notes than for others. At a minimum, my guess is that the age of this sample set (HW2?) and the sophisticated stereo tremulant modeling algorithm of HW5 do not always play well together. I did not test systematically, but I think there is no other stop in the sample set that behaves the same way. The solution I found was to adjust the “tremulant mod: amplitude stereo bal” for this stop in the HW5 voicing tools. MIDI 67 and many notes above it needed to be set to the minimum slider position (down 300); for others, the setting was less extreme. For one note, the best setting was in the other direction: up 114. These adjustments succeeded pretty well in evening out the loudness of the entire rank when the Tremulant is activated.
I finished the “migration” to HW5 on my PC late on the evening of 23 December. Rather than spend time devising HW5 audio routings and mixes for an outgoing computer/audio system, I let the program “transcribe” what I had been using. Although it was too late for household purposes for me to sit down and seriously play a sample set through the speakers, I was too curious about the success of the “transcribed” channel assignments to go to bed without testing them. I loaded an old friend, the Marcussen/Moerdijk sample set, and registered a single 8-ft. flute on each keyboard. At modest volume I pressed one central note of the Rygpositiv, and my eyes popped. I had never before heard such “body” and “presence” from this stop. I had been happy with this set since 2010, and being among my oldest sets, it had also been among the most audibly improved sample sets following my inclusion of DiracLive phase/time correction in my system. Now, more than ever before, HW5 made it seem like I could be holding the pipe in my hand. Yes, the sound of each of the three stops emerged from the correct speaker pair, but the big news was the ravishing improvement in sound QUALITY for all three of them.
I learned over the next couple of days that this entire sample set (as well as others) sounds simultaneously brighter and clearer, making the previous HW4 sound seem to have been relatively “washed out.” Everything was more substantive, “real.” The NATURE of the improvement in sound was similar to the improvement evoked by DiracLive, but it apparently takes place upstream of where DiracLive can have an effect.
It is possible that Hauptwerk 5, with its extensive audio capabilities, is able to circumvent an unknown (misphased?) deficiency in my Windows 7 system. Remember, the only modification I made was to replace HW4 with HW5: the General and Organ settings, including polyphony, remained the same. I will never know whether the same contrast between HW4 and HW5 could occur in my upcoming Mac system. In any case, it is encouraging and rewarding to know the possibility of such a profound, unexpected HW5 improvement.
Unfortunately, there is more to the story. I did learn of one apparent HW5 detriment, or at least complication, to the Marcussen/Moerdijk sample set. The sole 8-ft. flue of the Brystvaerk, a Traegedackt, is aberrantly affected by its Tremulant under HW5 (but not 4). Between the F# and G above middle C (MIDI 66/67) there is a large jump (2-fold?) in volume going up the scale IFF the Tremulant is drawn. In other words, the apparent loudness of the individual notes at and above MIDI 67 toggles with the Tremulant status: off = normal, on = (too) loud. The effect is stronger for some notes than for others. At a minimum, my guess is that the age of this sample set (HW2?) and the sophisticated stereo tremulant modeling algorithm of HW5 do not always play well together. I did not test systematically, but I think there is no other stop in the sample set that behaves the same way. The solution I found was to adjust the “tremulant mod: amplitude stereo bal” for this stop in the HW5 voicing tools. MIDI 67 and many notes above it needed to be set to the minimum slider position (down 300); for others, the setting was less extreme. For one note, the best setting was in the other direction: up 114. These adjustments succeeded pretty well in evening out the loudness of the entire rank when the Tremulant is activated.
Don Vlazny