A Hidden [Skinner] Surprise in Every HW 6?

I was one of the early birds that needed a temporary licensing ‘fix’ for the HW 6 perpetual upgrade on 19 Nov 20. While thanking François for his prompt help, I mentioned to him that I now know why I bought 16 cores and an RME UFX+ [as well as Dirac2 processor/crossovers for all channels] late last year. The sonic result with a wholly unfettered Hauptwerk 6 is fabulous. Incredible. The only ‘compromise’ I have needed to make for HW 6 was to increase my (single page) buffer size from 512 to 1024, though I may have been on the edge of needing to do so with HW 5. Even so, latency continues to be a non-issue: with my previous Windows 7 HW computer, I could “feel” the difference between an estimated 6ms and 11ms latency, but with the current Thunderbolt connection to the RME, I detect no delay at all despite the fact that I CAN “feel” the 10ms delay that I SET for the antiphonal division of one sample set.

Back to the titled topic: although I am not as much a fan of “American classic” organs as are many valued contributors to the Forum, the Mt. Carmel Skinner was among my earlier sample set purchases after I began working with Hauptwerk in 2010, and I have had genuinely rewarding experiences with it, particularly playing slower pieces not so affected by the single-release full natural reverberation. Since then my Hauptwerk data folder has always included, as far as I know, the items 000115 and 000116 for the “Skinner Organ Installation” ID 01 and ID 02, plus the slightly later addition of 000119, “E.M. Skinner Update 1.02.” However, after installing HW 6, I was quite surprised to find data files that were previously absent: 000117 and 000118, “MasterWorks E.M. Skinner” 01 and 02. I already suspected that something unusual was happening with the Mt. Carmel sample set, because among 167 caches that I needed to regenerate for HW 6, it was the only one that exhibited a significant REDUCTION in size, from 4.49 (HW 5) to 3.65 GB. (Only 3 other caches showed any reduction at all.)

I spent most of my organ time today with a different sample set, but I did load Mt. Carmel for a while, and I could distinctly cause and hear at least two different releases: one quick with echoes and one familiarly long-tailed. IS THAT TRUE?!! It appears that I will need to rethink and revise the truncation/IR setup that I devised for Mt. Carmel under HW 5. Unless I am completely daft about this entire matter….

In any event, I do express great thanks to Martin Dyde, et al., for this stellar development in Hauptwerk sound. And I also acknowledge the stunning blessings that have coalesced to make this all possible for me.

MasterWorks E.M. Skinner

MasterWorks E. M. Skinner is just the name of MDA's release (many years ago) of the entirely dry version of the Mt. Carmel organ. They still sell both versions: the wet and dry (their current website/store offerings reflect the MasterWorks name for the dry version).

Although many years ago now MDA said that a new, updated version (actually an entirely new recording) of Mt. Carmel in its wet version was coming (with multiple releases, among other newer technology), that has not yet appeared, and given the extreme delay now, it feels rather unlikely that it ever will. (It was certainly much anticipated by many, but I think most of us anticipating it have long given up hope.)

One of the forum members had even produced a reverse-engineered version to add "fake" multiple releases (created with convolution reverb), but MDA did not permit it to exist for more than a few days and said that a wet version, done officially by them, would be forthcoming. That was many years ago now.

Thank you, Carson. (And thank you for many other contributions in and outside the Forum!)

Yes, I knew the instant that I saw "MasterWorks E.M. Skinner" in my data list that these two entries referred to the dry version. However, I never purchased that version, and I never saw these files in my HW installation prior to HW 6.

The surprise for me is that I am now hearing alternative release tails after loading the wet version. [Was I simply missing them all along? I don't understand how application of an IR can create the distinction.] Since I have both of my at-home invalids in bed, I fired up the organ to again listen and confirm, this time through AKG K812 headphones. Moreover, I also asked my older daughter to listen through the headphones and describe to me any difference she heard AFTER I lifted my hands from holding either long or very short chords. Her primary talents are not in music, and she knows nothing about organs or electronics. When I asked her whether she heard a difference, she responded, "It depends what you mean." I said I was not trying to shade her response. Her next sentence regarding the difference she heard actually included the word, "echo." She independently heard the same difference I am hearing, even though the sample set is still swimming inside the IR treatment I gave it under HW 5. (I have not yet looked up the details, but as I recall, I selected a partial tail truncation for loading the set and a moderately reverberant IR for amplification.)

I guess that I am trying to understand this unexpected bonus in my HW 6 experience. And I still need to decide how I want to hone it - if at all.

Thanks again.

[mdyde]

Hello Don,

I was one of the early birds that needed a temporary licensing ‘fix’ for the HW 6 perpetual upgrade on 19 Nov 20. While thanking François for his prompt help, I mentioned to him that I now know why I bought 16 cores and an RME UFX+ [as well as Dirac2 processor/crossovers for all channels] late last year. The sonic result with a wholly unfettered Hauptwerk 6 is fabulous. Incredible. The only ‘compromise’ I have needed to make for HW 6 was to increase my (single page) buffer size from 512 to 1024, though I may have been on the edge of needing to do so with HW 5. Even so, latency continues to be a non-issue: with my previous Windows 7 HW computer, I could “feel” the difference between an estimated 6ms and 11ms latency, but with the current Thunderbolt connection to the RME, I detect no delay at all despite the fact that I CAN “feel” the 10ms delay that I SET for the antiphonal division of one sample set.

Excellent. Glad you like it. Thanks again.

Back to the titled topic: although I am not as much a fan of “American classic” organs as are many valued contributors to the Forum, the Mt. Carmel Skinner was among my earlier sample set purchases after I began working with Hauptwerk in 2010, and I have had genuinely rewarding experiences with it, particularly playing slower pieces not so affected by the single-release full natural reverberation. Since then my Hauptwerk data folder has always included, as far as I know, the items 000115 and 000116 for the “Skinner Organ Installation” ID 01 and ID 02, plus the slightly later addition of 000119, “E.M. Skinner Update 1.02.” However, after installing HW 6, I was quite surprised to find data files that were previously absent: 000117 and 000118, “MasterWorks E.M. Skinner” 01 and 02. I already suspected that something unusual was happening with the Mt. Carmel sample set, because among 167 caches that I needed to regenerate for HW 6, it was the only one that exhibited a significant REDUCTION in size, from 4.49 (HW 5) to 3.65 GB. (Only 3 other caches showed any reduction at all.)

I spent most of my organ time today with a different sample set, but I did load Mt. Carmel for a while, and I could distinctly cause and hear at least two different releases: one quick with echoes and one familiarly long-tailed. IS THAT TRUE?!! It appears that I will need to rethink and revise the truncation/IR setup that I devised for Mt. Carmel under HW 5. Unless I am completely daft about this entire matter….

Whatever gave that impression/appearance, it definitely wasn't due to any change in Hauptwerk v6 above v5 -- there are no differences at all in the way it handles organs, installation packages, release samples, or any given sample sets.

As Carson mentioned, there are 'wet' and 'dry' versions of the MDA Skinner (with the dry version called the 'MasterWorks Skinner'). Downloads for both were published by MDA when Hauptwerk v5 was released:

viewtopic.php?f=10&t=17988

... so maybe you had downloaded and installed both at that time (or at at least some of the parts of the dry one). They're separate/self-contained sample sets, although I recall that there were some spurious (but harmless) installation packages included in the downloads for those v5-compatible versions of the Skinner at the time: viewtopic.php?f=1&t=18003&p=135877#p135877

No official version of MDA's Skinner sample sets has ever been released with multiple release samples to my knowledge.

I think you must just have some setting difference (e.g. different reverb selected, or different release truncation settings) since you last loaded that particular sample set, or are perhaps misremembering how its releases sounded in that regard. Alternatively, I suppose it's possible that your MIDI keyboard could be triggering the notes twice (e.g. key contact bounce), giving a sort of brief 'blip' of the pipes upon release. You could maybe see whether the effect is still audible by instead clicking on the virtual keys and/or if the reverb is temporarily disabled.

Anyway, whatever it is, at least you like effect!

Thanks again.

Thank you, Martin. Sorry to be a bother. I still have no idea how the dry files ended up on my computer, though a careful look at my records indicates that the last I can GUARANTEE their absence was in 2019, under HW 4, not 5. I do not seek unauthorized copies of anything.

I also do not understand the "effect" that I (we) hear: there is definitely more than one "echo" per sharp attack/release. I don't know whether I "like" it, but I can at least report that it is more audibly clear under HW 6. And it is different from what I expect.

[mdyde]

Thanks, Don.

I still have no idea how the dry files ended up on my computer, though a careful look at my records indicates that the last I can GUARANTEE their absence was in 2019, under HW 4, not 5.

In that case, it was almost certainly simply due to the spurious MasterWorks Skinner packages that were accidentally included at in the v5-compatible version of the wet Skinner's downloads around the time of Hauptwerk v5's release (as detailed in the forum topic I mentioned: viewtopic.php?f=1&t=18003&p=135877#p135877 ). Anyway, it's harmless (aside from occupying some disk space unnecessarily), and you can just uninstall the spurious packages if you like.

[mdyde]

Hello Don,

I also do not understand the "effect" that I (we) hear: there is definitely more than one "echo" per sharp attack/release. I don't know whether I "like" it, but I can at least report that it is more audibly clear under HW 6. And it is different from what I expect.

For good measure, I've just loaded the wet MDA Skinner sample set in Hauptwerk v5, and compared it carefully to v6, and I can confirm that (as I would expect) there are definitely no changes in the way its release samples behave or sound.

Try the tests I suggested (disable any additional reverb that you might have applied, and try clicking on the virtual keys using the mouse, instead of playing via MIDI). My guess is that you had an additional reverb applied that is giving the impression you describe. (There definitely haven't been any relevant changes in Hauptwerk v6, aside from the potential improvements in overall clarity from the new 'higher definition pitch-shifting' and/or 96 kHz audio options).

Dear Martin –

In case you choose to monitor this thread for my test results (appreciated but not necessary), I can report that disabling the IR that I’ve been using with the Mt. Carmel sample set does eliminate the “echos” (perhaps more accurately, oscillations in reverb energy or focus) that I have been perceiving. With the IR defeated, I also hear the result of my having partially truncated the set’s own reverb. However, the IR’s reverb oscillations are, curiously, still apparent to me ONLY in the tails of short/sudden chords or notes; in contrast, long-held chords or notes seem to decay “smoothly.” Overall, you are correct: I have come to generally “like” this IR result, though I may eventually do a little tweaking. By the way, I experience NO difference between playing from the keyboards and inputting notes from the "Console" computer screen. (My physical keyboards consist of a fine UHT stack.)

If you don’t mind my “piling on,” this reminds me of a different, unrelated observation that I made soon after acquiring HW 5, but which I never pursued in the Forum. It continues to be apparent to me in HW 6. It is that in at least two of my “oldest” sample sets (Cappel and Forcalquier), there is a steady oscillation introduced into the sound of at least some stops that very much resembles the action of a tremulant even when the actual tremulant that belongs to the set is not applied. The effect (which, again, is new to HW 5/6) is particularly strong in Cappel, for which I had to turn the wind model down to 40% in order to adequately eliminate it. (I then restored some of the set’s tremulant strength using the voicing tools.) The Forcalquier set required wind model reduction only to 70%, with no further voicing adjustment.

Although reducing the wind model level may be unfortunate for these sample sets, I readily accepted it as a cost of the improvements in sound that are otherwise so overwhelming for ALL sample sets.

Thank you!

[mdyde]

Thanks, Don.

In case you choose to monitor this thread for my test results (appreciated but not necessary), I can report that disabling the IR that I’ve been using with the Mt. Carmel sample set does eliminate the “echos” (perhaps more accurately, oscillations in reverb energy or focus) that I have been perceiving. With the IR defeated, I also hear the result of my having partially truncated the set’s own reverb. However, the IR’s reverb oscillations are, curiously, still apparent to me ONLY in the tails of short/sudden chords or notes; in contrast, long-held chords or notes seem to decay “smoothly.” Overall, you are correct: I have come to generally “like” this IR result, though I may eventually do a little tweaking. By the way, I experience NO difference between playing from the keyboards and inputting notes from the "Console" computer screen. (My physical keyboards consist of a fine UHT stack.)

Yes -- added reverb can certainly give perceived 'echoes' (room reflections), which might be heard more clearly on brief sounds.

If you don’t mind my “piling on,” this reminds me of a different, unrelated observation that I made soon after acquiring HW 5, but which I never pursued in the Forum. It continues to be apparent to me in HW 6. It is that in at least two of my “oldest” sample sets (Cappel and Forcalquier), there is a steady oscillation introduced into the sound of at least some stops that very much resembles the action of a tremulant even when the actual tremulant that belongs to the set is not applied. The effect (which, again, is new to HW 5/6) is particularly strong in Cappel, for which I had to turn the wind model down to 40% in order to adequately eliminate it. (I then restored some of the set’s tremulant strength using the voicing tools.) The Forcalquier set required wind model reduction only to 70%, with no further voicing adjustment.

Although reducing the wind model level may be unfortunate for these sample sets, I readily accepted it as a cost of the improvements in sound that are otherwise so overwhelming for ALL sample sets.

The wind model is processed at a higher time resolution in Hauptwerk v5+ (which in general gives more realistic, and smoother, results), so yes -- it can make a small amount of difference to modelled wind behaviour. In the longer term, maybe those two sample sets would benefit by some slight tweaking in their organ definitions for optimal results in v5+, but in the meantime, tweaking the depth of the user wind model depth setting (as you did) would be the best option.