Thu Oct 14, 2010 12:27 am
I have found the recent threads which touched on the above to be very good reading.
There are many competing requirements for multi channel configurations - effective acoustic mixing, removal of dissonance, creating a soundfield that might mimic that which pipes might, creating the 'feel' of pipes, and most of all, creating the overall illusion of a pipe organ.
I consider some of these competing requirements. For instance, to remove the dreaded 3rds/5ths IMD/Dissonance, issue really requires that one moves to at least a 5 speaker group, and preferably a 12 speaker group (using cyclic ranks and octaves cycled). But, if I were to consider my installation, which currently stands at 39 channels (mostly 2 channel groups plus the aux and bass channels), I would then lose the ability to terrace the ranks.
Recently, the impending visit of an acclaimed organist, spurred me on to work more on the terracing of ranks. This paid immense dividends in achieving 'that' sound. However, if I were to move to larger multi channel groups, I lose the ability to do this as too many ranks would be in the same location. Of course, if I were to build a much larger music studio, and install 100 or so channels, then I could achieve what I am achieving now, but some IMD/dissonance (however, I might as well install the real thing instead!). The other aspect which would go MIA would be the ability to have all celestes physically removed from the same physical location as the unisons.
So, I am very hopeful that once V4.00 is released and bedded in that Martin will be able to wave his magic wand and come up with a hybrid cyclic outcome which will at least remove the most critical 3rds and 5ths in 2 channel groups. By "most critical" I mean those intervals where the ear is most sensitive from just below note 48 through to 72.
As an aside I have reconsidered my configuration using a mix of groups that are larger than 2 channels. It has been interesting. In more than one instance, to take say, four 2 channel groups with 8 ranks (two each) and turn them into a single 8 channel group, I lose acoustic mixing (they need to be too close together), and most especially, terracing, but gain IMD/dissonance improvement. So, far the latter has not been worth the loss of the former.
While I have been doing battle with the challenges of multi-channel, it has occurred to me, given the success of the Allen Q311, that maybe there is another way. The Allen Q311 uses a virtual soundscape into which pipes are postioned. To me, it is the most musical of all of the Allen theatre models. IMHO ,their multi channel theatre instruments fail, because they simply do not have enough channels to do the job, and they lose the effectiveness of the Lake Technology system (now owned by Dolby) which is utilised to deliver the Acoustic Portrait in the Q311 (as it works best in a stereo soundfield). So, stick with me, as I work this one through.................
The other reality for me, in the context of a VPO is that a high quality 3 way speaker clobbers the performance of a two way every time. Such that, a 4 speaker group of two ways, is not as effective on every audio front (except acoustic mixing) especially, timbral accuracy, mid range energy, authenticity (the 'illusion' again!) and so on. If I had the money all of my two ways would be replaced with 3 ways! But, where does it end?
And, so, if one were to take this to the absolute limit, I have to wonder whether it might be better to have a virtual pipe organ, in a completely virtual soundscape, and virtual space?
If one was to use all of the investment in a medium to large multi channel VPO running across, say 3 channels of very high quality 3 way speakers at the front, and at least 5 channels (inc. sub) for reverb, at the rear (so some two ways would survive!), might not this be a better way?
Could the dissonance issues be dealt with digitally? I don't know. Are they even an issue in this context?
Given the work done by some, like Paul Jacyk using just two channels and separate virtual chamber spaces which then feed into virtual reverb spaces, it seems to me, that having the ability for a HW user (as opposed to a HW Sample Set Creator) to place the ranks L - R and front to back (i.e. terracing) in a chamber, or multi chamber soundscape might be very effective. Would this be difficult to achieve in HW? I have no idea. But, as the facilities of an ODF permit the placement of pipes L to R within across a chest. It seems conceivable that front to back might also be. Is it?
Ian McLean
There are many competing requirements for multi channel configurations - effective acoustic mixing, removal of dissonance, creating a soundfield that might mimic that which pipes might, creating the 'feel' of pipes, and most of all, creating the overall illusion of a pipe organ.
I consider some of these competing requirements. For instance, to remove the dreaded 3rds/5ths IMD/Dissonance, issue really requires that one moves to at least a 5 speaker group, and preferably a 12 speaker group (using cyclic ranks and octaves cycled). But, if I were to consider my installation, which currently stands at 39 channels (mostly 2 channel groups plus the aux and bass channels), I would then lose the ability to terrace the ranks.
Recently, the impending visit of an acclaimed organist, spurred me on to work more on the terracing of ranks. This paid immense dividends in achieving 'that' sound. However, if I were to move to larger multi channel groups, I lose the ability to do this as too many ranks would be in the same location. Of course, if I were to build a much larger music studio, and install 100 or so channels, then I could achieve what I am achieving now, but some IMD/dissonance (however, I might as well install the real thing instead!). The other aspect which would go MIA would be the ability to have all celestes physically removed from the same physical location as the unisons.
So, I am very hopeful that once V4.00 is released and bedded in that Martin will be able to wave his magic wand and come up with a hybrid cyclic outcome which will at least remove the most critical 3rds and 5ths in 2 channel groups. By "most critical" I mean those intervals where the ear is most sensitive from just below note 48 through to 72.
As an aside I have reconsidered my configuration using a mix of groups that are larger than 2 channels. It has been interesting. In more than one instance, to take say, four 2 channel groups with 8 ranks (two each) and turn them into a single 8 channel group, I lose acoustic mixing (they need to be too close together), and most especially, terracing, but gain IMD/dissonance improvement. So, far the latter has not been worth the loss of the former.
While I have been doing battle with the challenges of multi-channel, it has occurred to me, given the success of the Allen Q311, that maybe there is another way. The Allen Q311 uses a virtual soundscape into which pipes are postioned. To me, it is the most musical of all of the Allen theatre models. IMHO ,their multi channel theatre instruments fail, because they simply do not have enough channels to do the job, and they lose the effectiveness of the Lake Technology system (now owned by Dolby) which is utilised to deliver the Acoustic Portrait in the Q311 (as it works best in a stereo soundfield). So, stick with me, as I work this one through.................
The other reality for me, in the context of a VPO is that a high quality 3 way speaker clobbers the performance of a two way every time. Such that, a 4 speaker group of two ways, is not as effective on every audio front (except acoustic mixing) especially, timbral accuracy, mid range energy, authenticity (the 'illusion' again!) and so on. If I had the money all of my two ways would be replaced with 3 ways! But, where does it end?
And, so, if one were to take this to the absolute limit, I have to wonder whether it might be better to have a virtual pipe organ, in a completely virtual soundscape, and virtual space?
If one was to use all of the investment in a medium to large multi channel VPO running across, say 3 channels of very high quality 3 way speakers at the front, and at least 5 channels (inc. sub) for reverb, at the rear (so some two ways would survive!), might not this be a better way?
Could the dissonance issues be dealt with digitally? I don't know. Are they even an issue in this context?
Given the work done by some, like Paul Jacyk using just two channels and separate virtual chamber spaces which then feed into virtual reverb spaces, it seems to me, that having the ability for a HW user (as opposed to a HW Sample Set Creator) to place the ranks L - R and front to back (i.e. terracing) in a chamber, or multi chamber soundscape might be very effective. Would this be difficult to achieve in HW? I have no idea. But, as the facilities of an ODF permit the placement of pipes L to R within across a chest. It seems conceivable that front to back might also be. Is it?
Ian McLean
