Building a home audio system + virtual pipe organ. What do I need?

[bradleybj09]

I'm really skeptical of this whole business without throwing a ton of money at it.

By far the weakest link in the virtual organ is the bass, but the rest is not as good as the hype.

Speaker systems in general have a lot of trouble reproducing pipe organs.

The bass is a big issue. In the virtual organ the bass was nothing like a real pipe organ as the Q was too high.

I have only ever heard the bass of a pipe organ realistically reproduced at power by transmission line speakers. These really can get you there when the rest will not.

I'm not certain that this bipolar argument is correct. A pipe is an omni directional radiator, a bipolar speaker is not. Also no speaker encircles like a pipe does.

I think the virtual organ is far to tied to Def Tech speakers because they are bi-polar. Def tech make some of the very worst speakers on the planet and none of them good.

There are a lot of other issues. High on the list is the swell box, which is not a volume control alone, but also reduces the HF radiation as the swell louvres close. I don't know if the OPs software does this or not. The next issue is chiff. Tracker mechanical organs have this, electric action instruments generally do not and pneumatic action organs definitely do not. Chiff is what happens in a tracker organ when you do not entirely press the key or depress and release it more slowly. It is a feature that makes tracker organs especially attractive. I have no idea how you replicate that from sampled sounds.

I have carefully researched all this today. I will post good recordings I have found on YouTube. There is a good video of the Hauptwerk organ imitating the Caen Cavaille - Coll organ. The bass between that and the real organ is night and day. I think I know why. If you are sampling an instrument you have to record it at close range with a highly directional mic, to avoid the environment contaminating it too much. However the more directional a mic the worse its bass response.

I have some form on this as when my son second was in college doing his electrical engineering degree, we did a lot of instrument sampling to try and make better synthesisers. After a lot of work we became aware of the limitations. We did a lot of work on strings which are also omni directional radiators. The above issue was a big problem. The trumpet, which is a directional radiator was an easier prospect. This by the way was the impetus for my first DAW which was in 1994.

So I would recommend TL speakers. In this case for his room a pair of TLs that could reach 16 Hz, a pair optimized for 16 ft pipes and another for the 8 ft pipes.

The next problem with pipe organs is that they have enormous power in the mid and high high frequency range. The enormous power, especially of the eighteen century instruments can easily be death to tweeters. I burnt out a tweeter in the coaxial driver of my center speaker with Bach a few years ago. I made mods to the speaker and so far so good.

I do know my current set up with its huge power available to the mid range can get the job done. So far the the HF units have handled it OK. I'm very careful not to cross my tweeters too low. If you are an organ enthusiast never cross a tweeter below 2.5 K Hz and higher if you can.

In my Grand Forks home, I did have an Allen electric organ console that was two manual and pedals. I got it gratis after it was outed when its speaker and amp system bit the dust. I did some mods and played it though my dual TL speakers that are now my rear backs. I'm not a very good organist and learn pieces slowly. But I did have fun with it. One piece I had fun with was an early Italian echo sonata and I had it so the left and right speakers could answer each other. It would have fooled no one that it was a pipe organ, I enjoyed it. The most realistic aspect was the bass.

Lastly the major reason that I design the speakers I do is because I'm a pipe organ enthusiast, and always have been.

My system here reproduces the pipe organs with a realism that other even very good systems can not approach. It is the realism of the bass and its articulation that sets it apart.

Unfortunately the build of this system cost a lot more then $6000 and that was 13 years ago now.

However this post has caught my eye so to speak. If the OP is interested I would be prepared to work with him to design a custom system. He would need to visit here to see just how accurately this TL based system can reproduce pipe organs.

I'm going to make another post of my researches today of the organs he is particularly interested in and three others. I have only selected good performances and excellent recordings. This will enable members to see the scale of the engineering problem.

Your thoroughness is very appreciated. I would of course love to talk to you more, but am also a little apprehensive of the scope of the endeavor. While the thought is exciting, I'm not a caliber of player or so much of an obsessive to need the *singular best* virtual organ in the world.
I'd be happy to have some of your time if you're giving it, but to some extent I think I'd be much more comfortable spending 3-6 grand and having a reasonably convincing instrument and 95% of the way to perfect than tripling (or more) that amount to close the last few percentage points.

When you're referring to "the" virtual organ, you're referring to Cameron's, right? I'm pretty sold on nothing DT at this point

As far as software imitations - I've generally heard that the chiff is pretty faithfully reproduced from sampled organs. Of course, that action does not exist in a midi keyboard, but is generally included as part of the sound sample (which are generally made up of 3 parts - attack, sustain, and release. Chiff is faithfully reproduced in the attack). There is a fairly well known personal (in-living-room) Hauptwerk instrument on youtube:

He's reproduced the essence of the St Sulpice console.

 

[TLS Guy]

Your thoroughness is very appreciated. I would of course love to talk to you more, but am also a little apprehensive of the scope of the endeavor. While the thought is exciting, I'm not a caliber of player or so much of an obsessive to need the *singular best* virtual organ in the world.
I'd be happy to have some of your time if you're giving it, but to some extent I think I'd be much more comfortable spending 3-6 grand and having a reasonably convincing instrument and 95% of the way to perfect than tripling (or more) that amount to close the last few percentage points.

When you're referring to "the" virtual organ, you're referring to Cameron's, right? I'm pretty sold on nothing DT at this point

As far as software imitations - I've generally heard that the chiff is pretty faithfully reproduced from sampled organs. Of course, that action does not exist in a midi keyboard, but is generally included as part of the sound sample (which are generally made up of 3 parts - attack, sustain, and release. Chiff is faithfully reproduced in the attack). There is a fairly well known personal (in-living-room) Hauptwerk instrument on youtube:

He's reproduced the essence of the St Sulpice console.

I fail to see how chiff is faithfully reproduced, because it varies as to how you press and release the key and the electronics have no means of sensing that.

I played that video on my big rig. I suspect that it was recorded direct from the electronics bypassing the speakers, as I heard no ambiance. It was very sterile. However it is a much better effort than fully electronically generated sounds.

This performance from Brooklyn by Olivier Latry shows the importance of building interacting with instrument.

I think if funds are limited I would just use two or may be three speakers that are really good. Whatever the situation two really good speakers best a bunch of inferior ones. There is never and exception to that what ever the application. I strongly recommend a transmission line design to reproduce the pedal division.

Its an interesting project though.

 

[bradleybj09]

I fail to see how chiff is faithfully reproduced, because it varies as to how you press and release the key and the electronics have no means of sensing that.

I played that video on my big rig. I suspect that it was recorded direct from the electronics bypassing the speakers, as I heard no ambiance. It was very sterile. However it is a much better effort than fully electronically generated sounds.

This performance from Brooklyn by Olivier Latry shows the importance of building interacting with instrument.

I think if funds are limited I would just use two or may be three speakers that are really good. Whatever the situation two really good speakers best a bunch of inferior ones. There is never and exception to that what ever the application. I strongly recommend a transmission line design to reproduce the pedal division.

Its an interesting project though.

I don't know anything about the transmission line designs, but I'd be interested in learning.

You're correct that the chiff cannot be reproduced directly by touch on the sampled instrument. Rather, I believe they make an effort to have a consistent chiff on all of the sampled pipes.