Forum Top Banner Ad


Ebay Classic organs



No announcement yet.

Improving organ with graphic equalizer

This topic is closed.
  • Filter
  • Time
  • Show
Clear All
new posts

  • Improving organ with graphic equalizer

    Old-timers here may remember my stories about swapping out organs at my previous church. Over a period of 17 years playing at that church we had 8 different organs, all of which I obtained for the church because -- being in the organ service business -- I was lucky enough to be involved whenever some church was giving up an older organ for a new one. Started with a discarded tiny Wurlitzer 4300 chapel spinet, moved up though two different Conn models (800 and 720), a big old 3m Baldwin 601, two Rodgers analogs (660 and 890), a smallish Allen digital (ADC4000), finally obtaining a magnificent Allen Renaissance R281. Then I had to leave that church behind over a grave disagreement with certain social and political stances being taken by the church and its denomination (SBC).

    I reported back in October that my wife and I are now on the staff of a wonderful Disciples church. We're becoming more accustomed to the new church and its unique worship ways and loving it. The organ I have to play there is just OK, though, somewhat less satisfying than the Allen I left behind! It's a Galanti Praeludium II installed in the late 80's or early 90's. It had become a little run down and was having an intermittent keying problem when I started at the church, but I've solved that problem by replacing a chip on one of the boards. So the organ has been basically working well lately, but the sound still leaves something to be desired.

    The organ was installed with the internal speakers disabled and the signal sent to a nice two-channel Carver commercial amp installed inside the console. I think the amp is rated at 100 watts RMS per channel, not huge but adequate for a 200 seat church. There was a pair of big Walker "credenza" style organ speakers in the chambers, which produced a fairly nice tone. The Great sounds through one channel, the Swell through the other, and the Pedal sounds equally from the two.

    I wasn't satisfied with the available volume level and laid it to the fact that the Walker cabinets were too large for the smallish window-like openings in the chambers. Much of the sound was being lost in the chambers, and turning up the amps seemed to border on clipping. It was not an option to enlarge the openings, so I decided to abandon the Walker cabinets for a split speaker arrangement.

    I placed a pair of Rodgers SW7.5 woofer cabinets on the floor at the rear of the choir and a pair of old Rodgers M10 cabinets in the chambers above, where they are just right to fit the small openings. The outputs of the Carver amp go into a passive crossover that sends the range below about 200 Hz to the subs and above 200 Hz to the M10's. I added an additional horn tweeter to each M10 cabinet and aimed it at the wall behind the choir.

    (The M10's themselves have to face straight out each opening, which points more or less at the opposite wall of the choir loft, not directly out toward the nave. This works OK because the room is very lively with a high wooden ceiling and smooth surfaces elsewhere.) From where I sit at the console at one side of the choir loft, I hear both divisions quite equally, as the swell speaker is over my head and bounces back to me off the opposite wall, while the great is directly across from me aimed in my direction.

    One day this past week I took a graphic equalizer to the church and inserted it between the organ's output and the amplifier's input. It's a rather ordinary graphic equalizer, cost about $100 at RadioShack a few years back. Has about a dozen bands per channel. The lowest band is centered at 24 Hz and the highest at 16KHz.

    I began by boosting the lowest bands, below 100 Hz. This dramatically improved the pedal, giving me the chest-thumping bottom end that I'd been missing. I tinkered with it quite a while and settled on a boost of around 8 or 9 decibels centered about 50 Hz, judging from the looks of the sliders.

    Next, I listened to full organ chords while manipulating each slider and discovered that certain bands between about 500 Hz and 5000 Hz are best attenuated just a tad because they contribute to a "honkiness" in the tone that seemed unpleasant to me. I found that doing too much of that damaged the beauty of individual stops, though, so I settled for just a moderate depression in the frequency response in the middle, made the sliders look like a "smile" of sorts.

    Finally, I listened to the top bands centered above 10KHz. These bands seem to contain little material at first (or maybe my hearing in that range is not what it used to be!) but the more I listened, the more I decided that they are crucial to an open, airy, engaging pipe-like sound. It seemed to benefit the organ to boost them a tad, and I used the "smile" of the middle range sliders as a pattern and boosted the upper bands a little more as I went up the scale.

    Was this little exercise in "voicing" worthwhile? I think so. The organ seemed more pleasing to me this morning, and my wife, a pianist and vocalist with little organ knowledge, said it sounded nice. I think it probably made more of an improvement than about anything else than can be done to an old organ like this one, which lacks any kind of internal voicing capability.

    Next step is to replace the 20+ year old Carver amp with a pair of new bi-amp-capable 900-watt Peaveys that I have in the shop. The bi-amping will do away with the passive crossover networks and give me even more control over the bass to treble tonal balances. Will report on the results of that.

    Final step -- get another organ! I'm casting about all the time looking for a premium brand organ that the church can get affordably. Will report on that too!
    Last edited by jbird604; 01-15-2012, 02:56 PM. Reason: sp
    *** Please post your questions about technical service or repair matters ON THE FORUM. Do not send your questions to me or another member by private message. Information shared is for the benefit of the entire organ community, but other folks will not be helped by information we exchange in private messages!

  • #2
    Very interesting, John. On another thread a few months back I opined that a 1/3 octave equalizer would do wonders for many digital installations, particularly in a small room with high-Q resonances. I did not receive much support for this viewpoint and ended up having a discussion about the theory and practice of linear systems and whether too much equalization would destroy the "phase response" of the system. (I believe, to the contrary, that equalizing for proper amplitude response will improve the phase response as well.) Your experience seems to bear out my viewpoint! As I assemble my ADC 7000 system, I do intend to use equalizers on all of the main channels; experiments with my (coincidentally) $100 Radio Shack equalizer have already convinced me that a big improvement can be had over relying on the very coarse finishing controls on the TG cards.



    • #3
      Well stated, Don. I suspect that narrow-band resonances are quite common in smaller churches and especially in home settings. Applying EQ should do a world of good in many cases. With this old Galanti the tonal shortcomings may have more to do with the speakers I'm using (especially the old M10 Rodgers cabinets made up of 6x9 car radio speakers and inexpensive piezo tweeters) and with the inherently poor sampling and rendering used in these fairly low-end organs. Whatever the cause, applying EQ really breathed some new life into this old thing.

      That ADC7000 has about a dozen channels, doesn't it? That calls for a lot of separate EQ units, but at $100 for two-channels it isn't prohibitive even for a big organ like that one. A more typical organ in a small to medium size church with maybe 4 channels, or perhaps even 6 could be supplied with EQ for only a few hundred dollars, a very reasonable price to pay for vastly improved sound.

      Keep up the good work. I look forward to hearing more about this project.
      *** Please post your questions about technical service or repair matters ON THE FORUM. Do not send your questions to me or another member by private message. Information shared is for the benefit of the entire organ community, but other folks will not be helped by information we exchange in private messages!


      • #4
        I think I remember saying the first time around "if it sounds good, it is good." That still holds true, so if whatever you do makes the thing sound better, it's a win.

        That said, I think I do want to weigh in on the "room resonances" issue. In full disclosure, I own a company that works in pro audio systems design and integration, mostly in houses of worship, and we play with some reasonably high-end solutions and measurement tools for this stuff every day. The average "room resonance" in a smaller room is a rather high-order modal event with corresponding events happening at each harmonic. A 31-band graphic EQ is not a very high-Q filter array, and especially in cheap ones, there's a lot of activity happening outside the passband in both the time and frequency domain. So not a very effective tool for correcting room response compared with a well-implemented parametric with variable filter widths. I would also say that there is no similar "room correction" equivalent in the pipe organ other than scaling and voicing, or in any other instrument either. The best corrective approaches are at the source and in the destination- using very accurate speakers and by working with the room acoustics where possible. Filters are often a necessary middle part of this process in PA work, but the result always sounds better when filter use is minimal and the speakers, signal chain, and environment are optimized.

        The graphic CAN be used to smooth out speaker response or to compensate for the limitations of organ mixing circuits or DAC paths with pretty good success. One of the great mysteries for me in the electronic organ market is the fact that so much resource is invested in tone production by various companies, and then so little into the signal chain following that tone production, including a majority of speakers designed by folks who have a far better grasp of woodworking than they do of acoustical physics. But that's another story.

        I have used graphics on Galantis and Allens over the years. The trick is, knowing that any filter circuit is introducing other factors besides frequency stuff in the passband, to be sparing and precise so that the side effects of the cure don't outweigh the magnitude of the disease, so to speak. My work with organs usually consisted of one or two cuts, and maybe a high end boost, at the most. People in my field just getting into working with equalizers almost invariably overdo things by trying to correct every little anomaly and end up introducing a whole new set of issues.

        But again, if it sounds good, it is good! And it's your organ.


        • #5
          Thanks, Michael. Interesting perspective you have from the pro sound business. You're surely correct in saying that a good parametric EQ is what one would need to correct room resonances precisely. So true that even pipe organ voicing can't fully compensate for such resonances (nor can note-by-note leveling on digital organs) since it is specific frequencies that need attenuating, not just particular notes of the scale. Effective EQ is our best ally when it comes right down to it.

          I suppose what I did with the Galanti was more along the lines of modifying the tonal spectrum to give a more pleasant bass-mid-treble balance than what was coming straight out of the DACs and filters in the console. And I was going on my own preferences rather than any objective measurements, of course, lacking any equipment to do that. As you said, if it sounds good, it's good.

          Wimpy bass doesn't do the job an organ is supposed to do, so copious output in that range is required if any digital organ is going to approach the power and glory of a real organ, and EQ is going to be necessary to coax that much bass out of the typical speaker.

          But I agree wholeheartedly that overuse of EQ is inappropriate and may introduce some unintended sonic quirks. In my experimentation, cutting the 1 KHz band seemed to make the full-organ sound more pleasant, even cutting it a lot. But then, when I listened to individual stops I could tell that a drastic cut in that band made some stops sound really strange. So I settled for some very moderate cuts in the mid-range. Likewise, I found that a considerable boost way up high, around 12 KHz IIRC, produced a noticeably crisper sound, but I imagine that too much boost in that region could conceivably damage tweeters and/or amps, and most people can't hear that high anyway.

          My business partner plays on a big old MOS2 Allen at his church, and I still think in some ways that organ sounds more pipe-like than many of the newer organs we service and install. There is an "edge" to the sound that is possibly just an artifact of the sampling and rendering of MOS, but it sure reminds me of what I hear from real pipes. This is testimony to me that purely accurate reproduction is not as important as other less-understood qualities that make organ sound interesting and convincing.

          I know that in the MOS and ADC days we would quite often crank up the bass pot on the DACs to near max on the pedal channels because it was understood that the bass response of the speakers rolled off way too steeply for convincing pedal sound. This wasn't so necessary in organs with separate subwoofer channels, where the amp could be turned up as needed to boost the bass.

          Great observation about the organ companies not putting enough emphasis on the audio chain, although you have to admit they've done some experimenting with speaker designs anyway. Witness the great variety of speakers Allen has used over the years, the enormous bass cabinets Rodgers used at one time, the speakers in cardboard tubes promoted by Saville and others. At this point in history, both Allen and Rodgers seem to have adopted pretty nice near-audiophile quality speakers.

          Organ speakers serve a little different function from audiophile speakers (or even pro PA speakers) in that they represent "musical instruments" rather than mere reproducers. They are the "pipes" of the pipeless organ. In order to "be" a 32' organ pipe, a subwoofer must be able to pump a lot of air, so big sub cabinets like the Allen SR-1 may not be so much accurate as just muscular.

          Perhaps we will see more EQ capabilities in upcoming digital models as the use of computers to control the audio continues to dominate. No reason why the voicing software can't have multi-band or even parametric EQ built in.
          Last edited by jbird604; 01-16-2012, 06:28 AM. Reason: add
          *** Please post your questions about technical service or repair matters ON THE FORUM. Do not send your questions to me or another member by private message. Information shared is for the benefit of the entire organ community, but other folks will not be helped by information we exchange in private messages!


          • #6
            Originally posted by michaelhoddy View Post
            I would also say that there is no similar "room correction" equivalent in the pipe organ other than scaling and voicing, or in any other instrument either.
            True. The stop-by-stop and note-by-note voicing in modern organs will do an imperfect job of correcting for lousy acoustics since it has limited capability to apply a non-flat response curve to each note; mostly it is used just to raise or lower the amplitude of all harmonics in a note by equal amounts.

            Originally posted by michaelhoddy View Post
            . . . One of the great mysteries for me in the electronic organ market is the fact that so much resource is invested in tone production by various companies, and then so little into the signal chain following that tone production. . .
            And as evidence, see my review posted yesterday of Allen's Opus VII. The largest digital organ maker in the world, given a virtually unlimited budget and unlimited engineering and production resources, appparently cannot make one of their finest instruments sound good in an actual church setting. Yet as John agreed, sound clips recorded under factory conditions are wonderful.

            Would it be fair to say in general that modern digitals are so good in most cases that 90% of the variability (and 90% of the problems) with their sound arise from details of the audio installation?



            • #7

              Good to see you back here participating in discussions here.

              I suspect that some issues here relate to organist and instrument being somewhat at odds with each other. I take it you prefer a more massive, American type sound, whereas the Galanti PII sounds best as a European style organ. Also, the fact that the organ is not voiceable must frustrate you to some degree.

              I would suggest your usage of Rodgers M10 speakers not helpful. These were speakers for Rodgers analogs, where they were part of the voicing of the organ. In other words they were the final filter in the audio chain. They had no bass really, and they typically had many peaks and valleys in the frequency response. The Carver amp. sounds to me like just adequate, the way it is used.

              Here is what I do, when confronted by a PII, and I look after quite a number of them. The folks who peddled them, and installed them generally used a big amp, put in an equalizer, some big bass boom boxes, and some open back speakers with a number of 6 X 9 speakers with a domed tweeter. The sound always sounded a bit odd in places, no matter what was done with volume, equalizer settings, etc. I suspect if one did a frequency and phase analysis of the speaker setup, that the result would be all over the map. Since a lot of the speaker are falling apart after 20 years, I sometimes am asked to re-do the speaker system. First thing I get rid of the equalizer. I find a lot of them are a source of service calls, such as distortion in one channel, one channel going dead, the sliders going bad. Usually that cleans up the sound somewhat, but can also make the organ sound flat or uninteresting.

              When asked to redo the audio system, I generally only retain the amplifier, which was of good quality and approx. 200 watts per channel into 8 ohms. I then put on high quality speakers of reasonable efficiency and good power handling, that go down to say 50 hz. Between, the amps and speakers I put in a powered subwoofer, usually from SVS, which has high signal inputs and outputs. I set the volume as high as needed, and also check out the phase switch to see which works best. Usually, at this point the organ already sounds better. Since it is a sampled type organ, you want the tone to be as natural as possible. From then on, I work on the tone generator cards in the organ. To do the work here, you will need the schematics, and understand the location of the stops in the TG boards, and also the signal coming off the board going to the mixer. Here is where I play around with resistor values to set the signal level coming off the board. You have to understand that usually there are 2 manual stops combined on the M-114 chip, so changing for one stop affects the other. Also there are H (high) and L (low) outputs. By carefully changing resistor values ( the final resistor in the audio chain on the TG boards), I change the balance of the treble/bass relationship and also between the various stops on their respective divisions. When I am done usually, the organ sounds very, very different. Less strident in the treble, better bass balance.

              On the PII, I don't change the values all that much, as it was a fairly well sorted instrument. On the PI, it was a different story, as it was voiced for really poor internal speakers, and when externals were added, the organ was generally very harsh and bright sounding.

              Anyways, that is my way of dealing with the sound/tonal issues with the PII.

              Also, typically in smaller rooms, I do not find room resonances to be a major issue, unless it is very live acoustically. Typically most are of moderate acoustics, and by the time it is full of people quite dead.

              I don't think the sampling quality of the organ is really a problem, so much as that you can't do a whole lot with it. The audio put on these organs is usually more of a problem. In the end, I would say from your description, that your tastes are not this particular type of organ.



              • #8
                When we did Galanti installations in the northeast US, we were usually using two of the ubiquitous Walker WM128/1308 three-sided speakers, which had good component quality, but which have some serious issues in especially the time domain, looking back now. So better than what Arie describes in his travels, but still not optimal. The PII had some nice individual sounds for the time, but really did lose a good bit going full-tilt.

                Would it be fair to say in general that modern digitals are so good in most cases that 90% of the variability (and 90% of the problems) with their sound arise from details of the audio installation?

                I think most of us agree that at this point in development, tone generation is no longer the weak link. The real issue is twofold: certainly the varied quality and implementation of the signal chain downstream of the TG (including DAC's, mixing, amps, speakers, etc.), but more fundamentally the fact that a speaker, even a good one, presents sound to a room completely differently than an organ pipe.

                A pipe is a relatively non-directional "low-Q" point source, with the exception perhaps of reed pipes. But even they are projecting resonant frequencies in many directions from the resonator. Pipes are louder at the mouth and at the end, but the sound is not coming uniformly from a singular direction. And this overall behavior does not change a whole lot with pitch or frequency.

                A speaker is directional, but this behavior changes with frequency. All speakers are far less directional at low frequencies, and far more directional at higher frequencies. Attempts to make speakers less directional and more "pipe like" by adding midrange or high-frequency drivers faced in different directions (as with the 3-sided Walkers or the Phoenix speakers) have the effect of creating multiple point sources, which has profoundly detrimental effects in the phase domain at anything approaching even half a wavelength of distance. And THIS phase behavior also changes with frequency, just to make things more complex. Add more notes or more stops to the speaker, and the soup gets even murkier.

                So even if you have one good-quality speaker enclosure for every note of every stop, you are still looking at a fundamentally different acoustical animal than a pipe organ. And the typical electronic organ has far less in both quantity and quality than this, introducing the other issues above. EQ has very limited capacity to even begin to address this. The best that can be done is to attempt to flatten out the speaker response and compensate for shortcomings.

                It could be argued that organ speakers with their attendant anomalies are really part of the "instrument" and not subject to the same rules as sound reinforcement loudspeakers. This was definitely true with analog organs. But I would propose that with sample-playback organs, the goal should be to make the signal chain downstream as transparent and flat to the original sampling as possible.

                Certainly all organs, pipe or electronic, are detrimentally affected by poor room placement or acoustics or room timbral response. Any organ sounds better in a good space. But I would make the generalization that electronic organs are more degraded by these kinds of environments than a typical pipe organ.
                Last edited by michaelhoddy; 01-16-2012, 11:48 AM.