It is currently Sat Oct 16, 2021 1:33 am


Why no vertical channels in samplesets?

Existing and forthcoming Hauptwerk instruments, recommendations, ...
  • Author
  • Message
Offline
User avatar

Jan Loosman

Member

  • Posts: 357
  • Joined: Sat Dec 20, 2008 4:33 pm
  • Location: The Hague, Netherlands

Why no vertical channels in samplesets?

PostWed Jul 29, 2020 8:18 am

Dolby atmos has been discussed on this forum in different topics. I have a Dolby atmos receiver in my home cinema setup and the additional height channels ad a lot to the overall sound. It gives more space, more sense of reality.

These days sample set producers are expanding their channels with direct, mid, diffuse ,rear in their latest creations.
But al these channels remain in the same horizontal plane!!!!
Why do they do that?, aren’t they forgetting a dimension? Are churches and cathedrals not buildings with also a very large vertical component.
So why not use one off the three front perspectives and feed it with a genuine sample recording of the front height with mics high placed or directed to the ceiling. this could ad a lot for a more genuine reproduction off a church organ.
This would have the same effect as a rear speaker but then vertically.

The drawback is that we have to add another speaker pair at the ceiling above our Hauptwerk setup.
I think rather small Cambridge audio mini minx speakers or other small speakers can do the job.

Regards Jan
Last edited by Jan Loosman on Thu Jul 30, 2020 8:37 am, edited 1 time in total.
Offline

1961TC4ME

Member

  • Posts: 3074
  • Joined: Mon Feb 07, 2011 1:45 pm
  • Location: Lake Minnetonka, Minnesota

Re: Why no vertical channels in samplesets?

PostWed Jul 29, 2020 10:59 am

Very interesting topic, Jan.

I too have pondered and looked at ways to expand the sound field vertically in order to achieve the best possible sound (or better said, more realistic sound). As you say and I too have come to realize, there's much height to the sound you hear in a real church that I felt I was missing at home. You go into a larger church / cathedral with a good acoustic, and depending on where you sit the sound of the organ also seems to come from a higher position and above, in some cases almost like a surrounding sound that's raining down on you. It's a very pleasing effect that I've wanted to duplicate at home, but has alluded me in the past. It isn't possible to me anyways having speakers positioned just to the front and the rear to achieve this as I've tried everything (including standing on my head :lol: ) with results that haven't done the job, and to me we are leaving out an important part in the sound by taking the front / rear speaker only approach. What got me thinking on a new method is I've taken a cue from a few of the local movie theaters we visit from time to time, and have noted how they position speakers, both in front, and also to the sides higher up along the walls which gave me an idea to try. I'm in a fairly small space of approx. 10' wide by 12' long with a ceiling height of just over 8'. My theory was to start by keeping my front and rear speakers as positioned but also add one pair of speakers up high near the ceiling directly off to my left and right side to fill in that high vertical and off to the side sound which I can control the volume of and also choose which part of the signal I send to these speakers. I've tried sending the rear / surround signal only to them which is the most distant signal, but have found so far that a signal mix of all works best (with the exception of direct not being included). My theory being by the time the sound travels so far in the large church coming off the ceiling and sides, the sound is becoming more distant / dispersed sounding, but not like it would be if you stood way at the back of the church, so I'll call it the mid somewhat dispersed signal. Then, with careful volume adjusting from there it opens up the vertical sound field considerably. I'll put it this way, since I've implemented this arrangement I won't go without it. Not exactly what you have in mind of recording a high and somewhat distant position and then having the option to send it to another pair of speakers positioned in a certain way, but then I do wonder doing that and what I'm doing if it would make that much of a difference. I've contemplated adding a second pair of side speakers and positioning them off to the sides again near the ceiling and just behind me, but I wonder what benefit it would add or if it might then detract from the overall sound.

I think part of the key to making it work is setting things up as you would imagine yourself sitting in the church. I try to strive for a sound that would place me virtually about a third of the way back in the church from the organ.

Interesting topic for sure and I'm always looking for new ways to improve the sound.

Marc
Last edited by 1961TC4ME on Wed Jul 29, 2020 12:14 pm, edited 1 time in total.
Offline

josq

Member

  • Posts: 870
  • Joined: Tue Jan 24, 2012 12:11 pm

Re: Why no vertical channels in samplesets?

PostWed Jul 29, 2020 11:53 am

Since HW5 we can route 4 "output perspectives": front&rear main, and front&rear upper. So it looks like Martin has nicely anticipated the demand for adding a vertical dimension to the sound.

I doubt however if extra recording positions are required. Rather, we can route a rank both to the main&upper speakers. We can do some virtual vertical positioning of the pipes by adjusting the balance between main&upper, just like we can adjust the horizontal positioning (left & right, front and back) by adjusting the stereo balance and the balance between front&rear.
Offline

kasterling

Member

  • Posts: 19
  • Joined: Thu Feb 20, 2020 7:04 pm

Re: Why no vertical channels in samplesets?

PostWed Jul 29, 2020 9:12 pm

josq wrote:Since HW5 we can route 4 "output perspectives": front&rear main, and front&rear upper. So it looks like Martin has nicely anticipated the demand for adding a vertical dimension to the sound.

I doubt however if extra recording positions are required. Rather, we can route a rank both to the main&upper speakers. We can do some virtual vertical positioning of the pipes by adjusting the balance between main&upper, just like we can adjust the horizontal positioning (left & right, front and back) by adjusting the stereo balance and the balance between front&rear.


Adding vertical dimensionality to the sound--in addition to horizontal--is a great idea. I've already been doing this for years using Hauptwerk 4.2. I use preamps and other devices to create multiple outputs from each Hauptwerk channel. Some of these outputs are routed to amplifiers feeding speakers mounted up on rear and side walls to create the vertical sound dimension. Of course, it should be mentioned that more than just the volume of these vertical outputs needs to be adjusted. In a large church or cathedral the sound waves bouncing off high walls are delayed before reaching the listener and have different reverberation from the more direct sound waves. Therefore, I manipulate all the outputs to the vertical speakers adjusting their delay and reverb individually to custom tailor the sound of each organ for maximum realism. I do not own Hauptwerk 5, so I don't know if such parameters as reverb depth and delay can be adjusted separately for each individual output path. Hopefully so. However, I have not yet heard any sample of Hauptwerk 5's internal convolution reverb that did not sound somewhat artificial to my ears compared to my own system using outboard processors. YMMV.
Offline

josq

Member

  • Posts: 870
  • Joined: Tue Jan 24, 2012 12:11 pm

Re: Why no vertical channels in samplesets?

PostThu Jul 30, 2020 1:35 am

kasterling wrote:Of course, it should be mentioned that more than just the volume of these vertical outputs needs to be adjusted. In a large church or cathedral the sound waves bouncing off high walls are delayed before reaching the listener and have different reverberation from the more direct sound waves. Therefore, I manipulate all the outputs to the vertical speakers adjusting their delay and reverb individually to custom tailor the sound of each organ for maximum realism.


Thanks. I think we are discussing two different concerns. I had the vertical dimension of the organ case in mind, for example Bovenwerk versus Rugwerk in Haarlem, Kampen, etc. The pipes in these divisions are a few meters apart vertically.

You are right to mention the vertical dimension of the reverb: sound reflecting from the high walls/roof. Good to hear that you are able to achieve satisfying results using multiple speakers and convolution reverb. I guess the reverb/delay needs to be optimized per organ?
Offline
User avatar

mdyde

Moderator

  • Posts: 13601
  • Joined: Fri Mar 14, 2003 1:19 pm
  • Location: UK

Re: Why no vertical channels in samplesets?

PostThu Jul 30, 2020 3:39 am

Hello kasterling,

kasterling wrote:I do not own Hauptwerk 5, so I don't know if such parameters as reverb depth and delay can be adjusted separately for each individual output path. Hopefully so. However, I have not yet heard any sample of Hauptwerk 5's internal convolution reverb that did not sound somewhat artificial to my ears compared to my own system using outboard processors.


Just for the record: yes -- on v5's audio mixer you can apply different reverbs to different buses (audio outputs/speakers), and adjust their depths and delays ('pre-delay') separately. Hauptwerk's convolver doesn't have any 'sound' of its own (it's a mathematically correct true stereo 4-channel 32-bit floating point convolver, with no compromises), so how it sounds purely depends upon the reverbs (impulse responses) that you load into it.
Best regards, Martin.
Hauptwerk software designer/developer, Milan Digital Audio.
Offline
User avatar

Jan Loosman

Member

  • Posts: 357
  • Joined: Sat Dec 20, 2008 4:33 pm
  • Location: The Hague, Netherlands

Re: Why no vertical channels in samplesets?

PostThu Jul 30, 2020 8:24 am

In my younger years I owned a Yamaha dsp e300 sound field processor. In the back of the room there were rear presence speakers just as a Hauptwerk setup but above my main speakers on the ceiling there were also front presence speakers.
I think Yamaha as a pioneer in sound field processing understood what it takes to recreate soundfields.
I remember this worked very convincing, creating sound fields like jazz clubs, Munster cathedral, concert venues etc.
All very realistic, giving a sense of being there.
But only by using 4 surround speakers In the corners you could recreate genuine soundfields. Just using main speakers and rears was less convincing.
I hope sampleset producers will read this topic and consider giving the option in future samplesets to add true recorded height channels.
I would happily give up one of the front perspectives (three is really overkill in my opinion,also introducing to many phase issues!) and use this to power these channels by adding height speakers.

Regards Jan
Offline
User avatar

engrssc

Member

  • Posts: 7283
  • Joined: Mon Aug 22, 2005 10:12 pm
  • Location: Roscoe, IL, USA

Re: Why no vertical channels in samplesets?

PostThu Jul 30, 2020 9:43 pm

A slight further thought, I set up a system where the high corner speakers faced into the corners. That added an additional dispersion feature that gave the listener the illusion that sound was coming from those basic areas but blended between the 2 walls and the ceiling that then gave the "image" of it being a bigger space than it actually was. These blended very nicely with the lower speakers as well. A wall of sound if you will. I should mention, these corners were hard surfaces, no sound absorbing items nearby.

We had to mount these corner speakers with what we used to call piano wire.Two wires from the ceiling at approx a 120 degree angle with the bottom wire fastened into the corner. They were smaller mid range and painted the same color as the walls aesthetically not too bad to look at.. 8)

Rgds,
Ed
Offline

josq

Member

  • Posts: 870
  • Joined: Tue Jan 24, 2012 12:11 pm

Re: Why no vertical channels in samplesets?

PostFri Jul 31, 2020 12:56 am

engrssc wrote:A slight further thought, I set up a system where the high corner speakers faced into the corners. That added an additional dispersion feature that gave the listener the illusion that sound was coming from those basic areas but blended between the 2 walls and the ceiling that then gave the "image" of it being a bigger space than it actually was. These blended very nicely with the lower speakers as well. A wall of sound if you will. I should mention, these corners were hard surfaces, no sound absorbing items nearby.



Yes, often people report that room reflections are beneficial to achieve a more enveloping sound experience. Wall-facing speakers are one possibility, omnidirectional or bipole or dipole speakers also can be used, or other speaker types with very wide directivity.

How good such approaches work may differ among sample sets. My theory is that it could work well with relatively dry sample sets. In (very) wet sample sets like Utrecht, the recording contains plenty of reflections from the large cathedral itself. An excess of additional reflections from the listening room will blur the original acoustics, and the experience of standing in the cathedral may be destroyed.
Offline
User avatar

Jan Loosman

Member

  • Posts: 357
  • Joined: Sat Dec 20, 2008 4:33 pm
  • Location: The Hague, Netherlands

Re: Why no vertical channels in samplesets?

PostFri Jul 31, 2020 5:53 am

josq wrote:Since HW5 we can route 4 "output perspectives": front&rear main, and front&rear upper. So it looks like Martin has nicely anticipated the demand for adding a vertical dimension to the sound.

I doubt however if extra recording positions are required. Rather, we can route a rank both to the main&upper speakers. We can do some virtual vertical positioning of the pipes by adjusting the balance between main&upper, just like we can adjust the horizontal positioning (left & right, front and back) by adjusting the stereo balance and the balance between front&rear.


Hi Jos

I just doubt if you wil have the same effect by just simulating height with only vertical positioning off the pipes.
Mics to record the organ samples have directionality mostly in the horizontal plane and will miss a lot off the vertical acoustics.
Sampleset recordings already record many channels so why not make a genuine copy of the front height acoustics. They record all these channels why not add two more and so make a more comprehensive copy off the organ including a lot more off the original vertical acoustics now.
I expect that the synergy of combining al these channels, now adding also a real vertical dimension will further enhance the sensation off playing the (almost) real thing.

Regards Jan
Offline

Atn52

Member

  • Posts: 26
  • Joined: Thu Aug 10, 2017 6:21 am

Re: Why no vertical channels in samplesets?

PostFri Jul 31, 2020 11:08 am

In Germany there are so-called SACD, which take up front, back and top. I simulated this by palpating the speakers at the front and back and placing 2 above the front speakers. with a good attitude I could output the room signals there. see my comments about 3D Surround in this forum
best regards
Antonius
Offline
User avatar

tf11972

Member

  • Posts: 281
  • Joined: Tue Jan 14, 2014 4:33 pm
  • Location: Bavaria, Germany

Re: Why no vertical channels in samplesets?

PostFri Jul 31, 2020 1:18 pm

@Jan
Every additional channel needs more time for postprocessing of the samples, rises the demand of RAM and results in higher costs for samplesets, computers and audio-equipment. Would you pay for all that?

@Antonius
It‘s not right that a SACD contains height channels, there are only 5 surround- and one subwoofer-channel. What you mean is Dolby Atmos or Auro 3D, but these sound formats are only on BluRays or netstreams.
Best regards
Thomas

Forestpipes - Virtual Pipe Organs
http://www.forestpipes.de
Offline
User avatar

Jan Loosman

Member

  • Posts: 357
  • Joined: Sat Dec 20, 2008 4:33 pm
  • Location: The Hague, Netherlands

Re: Why no vertical channels in samplesets?

PostFri Jul 31, 2020 3:34 pm

tf11972 wrote:@Jan
Every additional channel needs more time for postprocessing of the samples, rises the demand of RAM and results in higher costs for samplesets, computers and audio-equipment. Would you pay for all that?


Hello Thomas.

Recent sample sets especially from Sonus Paradisi are extending their front channels already with 3 perspectives.
So the demands for computer power etc. are already increasing.
But my personal preference would be oldschool 4 channel recording with a added vertical presence channel.
It would be nice if sample set producers would also offer this configuration.
Yes I would ad a extra height channel and buy these samplesets.
Sample set producers are innovating by adding channels, Hauptwerk is evolving with version 5 ,why not go vertical in the evolution of our Hauptwerk setups.? Dolby did this also evolving to Atmos!

Regards Jan
Offline
User avatar

tf11972

Member

  • Posts: 281
  • Joined: Tue Jan 14, 2014 4:33 pm
  • Location: Bavaria, Germany

Re: Why no vertical channels in samplesets?

PostSat Aug 01, 2020 2:09 am

Because of the optimal acoustics, organ recordings are already made at great heights, so one would actually have to add two channels close to the floor to create the impression of the organist or listener sitting on the floor.
Best regards
Thomas

Forestpipes - Virtual Pipe Organs
http://www.forestpipes.de
Offline

münsterorganist

Member

  • Posts: 96
  • Joined: Sat Aug 27, 2016 2:24 am

Re: Why no vertical channels in samplesets?

PostSat Aug 01, 2020 2:26 am

Dear HW User,
it would be more important to me not to keep inflanting the recordings, the sets then become too expensive, but to invest more in an authentic reproduction. If I can no longer hear the difference between a CD recording of the organ and the sample, I would be happier than working with any perspectives in a living room. A excellent wet sample is enough for this.You can still use the excellent perspective mixer to distribute the works( HW,RP,BW..) on the room. That would be my approch.
Next

Return to Hauptwerk instruments

Who is online

Users browsing this forum: No registered users and 7 guests