Quilcom NEWERCHORD: Living in the past?

[Spogg]

Hi everyone!

This synth was made in honour of Hammond’s 1939 Novachord keyboard which is the very first polyphonic synthesiser and used 163 valves! It’s not going to be seen as a go-to synth by any means, but it might be interesting to see what you can get out of it and replay a bit of history…

Download:

https://www.dropbox.com/scl/fi/er7pop3z ... k52p3&dl=0


Video:

https://youtu.be/h-CE36zFduI

[k brown]

You just realized one of my fantasy projects that I failed miserably at.

If anyone could do it, it would be ye!

[Spogg]

You just realized one of my fantasy projects that I failed miserably at.

If anyone could do it, it would be ye!

Thanks Kevin!

I can’t help wondering why you “failed” though, considering your considerable talent…

[k brown]

I only had 162 valves handy.

Here's my naff version from a few years ago. I was more pedantic about sticking to 'rotary switch' controls, resonators don't do a helluvalot, etc., etc. Can't remember now why I chose that name.

As typical with me, it looks prettier than it works!

If you're interested, I'll see if my uncertain immediate time sched will give me time to take a shot at adapting this GUI to Newerchord. If so, I was thinking along the lines of the same look for the top row of controls (which coincide closely with the Novachord's), and a more utilitarian "under the hood" type control panel for your second row of more 'custom' controls, and FX, separated by a strip of wood panel.

Extravox.fsm

(6.66 MiB) Downloaded 736 times

ExtraVox copy.jpg (20.62 KiB) Viewed 15224 times

P.S.: I'm curious to see if the text I chose displays properly on other computers (was made on Win XP).

[Spogg]

I don’t know how you can say that was a complete failure! You’ve captured the essence of the instrument really well in my view.
On my Windows 11 PC the text looks fine.

Here’s a few points:

The FULL TONE isn’t correct. When the 5 filters are OFF and FULL TONE is OFF, there should be no sound. The FULL TONE knob should bring up the dry unfiltered signal into the mix.

When I switched to WHEEL I expected the mod wheel to act as a swell but it doesn’t, but maybe I missed something.

The oscillator waveform looks like it’s based on the published one in the service manual, but you never get to hear that “raw” signal in the real instrument. Each generator has a high pass filter so you get a more pulse-like “saw” waveform feeding the filters and the FULL TONE mixer. You could add a tracking HPF to simulate that.
The waveform shown in the service manual is to scope the output of the dividers directly (I think), so you could minimise the half-way trigger glitch from the previous divider/master osc when re-calibrating.

The WOODWIND switch isn’t on the original but I can understand you wanting to provide alternative timbres.


If you would like to make a new GUI for the NEWERCHORD at some point, that would be nice and what you suggest sounds really good. But no rush, and only if you want to do it.

[josevo]

P.S.: I'm curious to see if the text I chose displays properly on other computers (was made on Win XP).

Hi K. Brown,
Nice UI, as I'm used to see.

In my case, the font readability can be improved a bit through "clear type" rendering setting, which is only accessible through ruby, as far as I know:

Code: Select all

def draw(v)         
   b = Brush.new(@iColor)
   v.setTextRenderingHint('clearTypeGridFit') # if (@iClearType)
   v.drawString(@iText, @iFont, @iAlignment, [0,0, v.width, v.height], b)
end


BTW: I'm not sure this feature works in Windows XP.

[k brown]

I don’t know how you can say that was a complete failure! You’ve captured the essence of the instrument really well in my view.
On my Windows 11 PC the text looks fine.

Here’s a few points:

The FULL TONE isn’t correct. When the 5 filters are OFF and FULL TONE is OFF, there should be no sound. The FULL TONE knob should bring up the dry unfiltered signal into the mix.

When I switched to WHEEL I expected the mod wheel to act as a swell but it doesn’t, but maybe I missed something.

The oscillator waveform looks like it’s based on the published one in the service manual, but you never get to hear that “raw” signal in the real instrument. Each generator has a high pass filter so you get a more pulse-like “saw” waveform feeding the filters and the FULL TONE mixer. You could add a tracking HPF to simulate that.
The waveform shown in the service manual is to scope the output of the dividers directly (I think), so you could minimise the half-way trigger glitch from the previous divider/master osc when re-calibrating.

Interesting points - I haven't looked inside that schem for years - hopefully what you say will make sense when I do! I think I might be more interested in putting that GUI to your better take on it.

[k brown]

P.S.: I'm curious to see if the text I chose displays properly on other computers (was made on Win XP).

Hi K. Brown,
Nice UI, as I'm used to see.

In my case, the font readability can be improved a bit through "clear type" rendering setting, which is only accessible through ruby, as far as I know:

Code: Select all

def draw(v)         
   b = Brush.new(@iColor)
   v.setTextRenderingHint('clearTypeGridFit') # if (@iClearType)
   v.drawString(@iText, @iFont, @iAlignment, [0,0, v.width, v.height], b)
end


BTW: I'm not sure this feature works in Windows XP.

Thanks for the tip!

[k brown]

A chunk of time opened up that I wasn't expecting, so here's the emulative GUI based on my plug.

I made the panel of additional controls 'hide-able' - clicking on the 'Quilcom' logo reveals it. If you'd prefer it, another option would be a toggle switch on the far right of the main panel, where the Novachord had it's power switch (it would mean crowding the knobs a bit). Another thought I had, was to make this reveal switch saveable in presets, so one would always know when a non-emulative control had been used or not.

I realized halfway into it that the Preset Manager was unlocked, so some of the presets may have been disturbed; the Factory Bank should be re-installed.

>> EDIT!: I was just re-reading your user guide and saw that you supplied the Resonator freq adjust controls with double-click to set to those of the hardware; hopefully that can still be done with the knobs I used.

Don't be shy about any labeling or other changes you'd like!


I did have one functional suggestion - the attack rate of the shortest soft-attack preset EG shape seems too slow; takes quite awhile for full vol to develop. I would think the shortest soft-attack setting should not take so long. I don't know if the actual EG times on the Novachord were obtainable.

GUI 2 copy.jpg (25.39 KiB) Viewed 14994 times

GUI 2b copy.jpg (26.48 KiB) Viewed 14982 times

For some reason I'm again unable to post the schem, or even a Zip of it (?), so you have to get it from Sync.

https://ln5.sync.com/dl/f957c5c80/iqgba ... m-3xhkmvsr

[Spogg]

Kevin, you never cease to amaze me mate! The GUI is lovely and you did it SO quickly!

I moved the label for the CC# for SWELL. If you enter 1 for example, a label appears saying MODWHEEL, but it was cut off.

I do think it would be better to have some sort of specific switch for unhiding the lower panel, because I reckon clicking on the Quilcom label isn’t very intuitive. I had the idea it could be a text switch which changed from NOVA to NEWER but that’s not a request, just an idea. I also like the idea of the switch having a preset parameter (but not automatable), then I can update the preset bank to show when the newer stuff is in use. I’ll leave the manifestation of the switch entirely up to you though.

I’m not so bothered about double-clicking, because control-click sets the defaults.

I got the envelope timings from the Cherry Audio demo. I only found out later that it’s modelled, not sampled. However, they say they referred extensively to the fully restored version. The attack has a “charging capacitor” exponential shape, so it starts off fast then slows down. I was able to modify Martin’s envelope by modulating the attack time from the envelope’s output. However, unlike the Mini Moog, the exponential attack goes up to maximum so the rate of volume increase slows down noticeably as the attack phase progresses. The Mini Moog envelope attack stops at about 0.7 so sounds different. I agree that the original isn’t ideal, but I provided the ADSR settings so it can be tweaked.

Here’s the zipped schematic:

https://www.dropbox.com/scl/fi/6spmrunv ... b0mwj&dl=0

Cheers!