Someone asked for some details about how I made the tracks in my album and I thought it would be fun to go over some of the things I think stand out and that I can remember. All of them started as modular jams. The patches weren’t really technical or ‘clever’, just a lot of analog oscillators and filters which have lots of inputs and outputs for audio and CV, and lots of modulation options, which I used extensively. When I first did these jams, about 8-9 years ago, my eurorack setup was mostly modules by a maker called Blue Lantern. He was selling complete systems at a pretty steep discount so that was my entry point into the format.

Blue Lantern modules can be very cool because, at least for those earlier ones I have, he would have a lot more ins and outs for modulations and weird controls that you wouldn’t see on other modules that were using the same circuit as the source for their design. They’re almost like ‘circuit bent’ versions of these common circuits. One of the oscs I used is the Transistor Pyramid Core VCO which has two types of synch, which sound different, exponential and linear FM, which sound different, many output shapes and PWM, and it just sounds great (based on a Buchla design). I had that and another BL osc with similar features, and a lot of the grungier sounds I got on the album, the growly, gurgling bass sequences that also have a lot of higher freq stuff going on, were achieved by cross-modulating those two oscs in various ways.

For most tracks on the album I wasn’t super concerned with having elaborate melodies or sequences, I just used a Blue Lantern arpeggiator module to generate some arpeggios, and I would turn the knobs on that live to get different notes brought into the sequences it made, which I would be using to control the oscs. The oscs would be sent to various Blue Lantern filters which had the same design philosophy as the oscs- built on designs that have been around for decades and used and reused by countless makers, but adding extra features everywhere possible for more sound design capability. This really resonates with me. So a lot of the tracks on the album were based around me just patching these arpeggiated oscs so they are mangling each other, and then going into these filters which were further mangling them, until I got a nice soupy wall of sound.

I would do filter sweeps and other adjustments live. Another key module for at least some of the tracks is the Synthrotek Eko, a very cheap and simple PT2399 based delay. It can do some simple echoes but if you modulate the delay time you can have everything collapse into a black hole and explode again in pure chaos. It’s pretty crazy and fun, and for a few tracks I would ‘rehearse’ a sort of compositional idea where I would use it to collapse everything into chaos at just the right time. In some cases I had the major elements of the track going into it and I would use adjusting the delay time to change the tempo and rhythm of everything.

There were a lot of other modules being used, mostly other BL ones, and I would just keep patching them together, using everything in my rack (maybe 25 modules would be patched in a big spaghetti mess, to make each of these tracks), to create these patterns that could keep going on their own, and sounded pretty nice to me, then I would record myself tweaking knobs live for a few hours. A few years later I began editing out some of the best bits and turning them into tracks. The original recordings were usually MONO mixdowns, and probably driven a bit too hot somewhere in the chain because I wanted a crunchy sound. I also usually avoided having amplitude envelopes on the oscs, opting for a continuous wall of free-running oscillator sound. The end recording obviously was pretty lacking in amplitude dynamics.

When I started editing, I wanted to introduce some dynamics back in to the recordings, but keep the spirit of the jams intact. I used a lot of EQing, with various plugins, including some Acquas, and even some of my own Nebula programs. This brought some of the frequencies out and enhanced some of the rhythmic qualities. I would edit some bits out that went on for too long. I used various methods to introduce some stereo image to the recordings. In most cases what you hear is still basically the modular jams but highly polished up and enhanced. Tracks 1 and 7 are exceptions. For track 1, I took the soupier bassy gurgling part of the track and processed it very heavily with Image Line’s Harmor, which is an image resynthesis plugin. I used it to harmonize the modular gurgles and create the entire ‘chord progression’ that you hear in the track, by using different harmonizer settings for each harmony. I used multiband processing to leave some of the content of the original recording unmolested by Harmor, and just panned it around instead. Multiband processing was heavily used during the editing/mixing, often to pan higher frequencies around or to add other effects only in certain frequency ranges.

Track 7 was created after most of the other tracks and I used a module called the Grendel, which is a vocal formant filter. The original recording was too monotonous so I sliced it into phrases and pitched some of them up and down to bring in some dynamics to the pitch. I did the pitching mostly just by using Reaper’s built-in pitch shifting ability for audio clips. I experimented with the different available pitch shifting algos for each slice, discovering that the nature of the synth recording (being kind of squelchy), really worked well with some of the more lo-fi pitch shifting methods to get really nice gritty sounds. The end of that track was processed with my Sooper Time Tube stereo program.

I did a few overdubs, like the barely audible chords in track 7 that happen in a few places, but especially in track 8. The intro to track 8 was filtered with different filters from my ‘mixing filters 3 hycor’ Nebula filter pack. Before that it was just a constant bassy gurgle but the different filters and applying a fade in made a nice intro. I think this this track goes on maybe a little too long, but it might have my favorite part of the album starting at 7:58 (those background sounds are overdubs). The squeeky sound that echoes in track 10 used my Nolard tape delay programs, various settings for each echo. The part starting at 1:41 in track 2, uses an unreleased program I made from my orban stereo synthesizer (a pseudo stereo comb filter unit). The effect gets more and more intense and I think it makes everything sound like it’s getting bigger and bigger and coming from outside out of the boundaries of the speakers (or headphones).

Track 3 also used Harmor. Each section of the gurgly synth is processed differently with Harmor, then they were processed through my modular again (mainly using a module that clones the Korg PS-3300 filter bank), and I came up with a ‘drum’ overdub, created with a modular patch and sequenced in a way where it seems to be related to the synth patterns, but in a drunken way. Track 4 is the only one that had different elements recorded separately, so I just panned those around to get a stereo mix for that one. Tracks 5 and 6 are the most ‘pure’ in the sense that what you hear is very close to the original recordings, just a little editing for time, EQing, and some of my Nebula reverbs were applied (like with every track), however the intro and outro of track 5 was filtered and processed a lot (ends with a big wash of my BX20 reverb).

Track 10, the last track, is from one of my first ever modular recordings (2014?), and it’s actually two takes of a more planned out composition, edited together. That one used the cheap Eko delay module a lot, but not for echoes, just for texture and rhythmic changes. I really like the bubbly textures I got here, in the first half. The really low pitched bubbly sounds were made by creating a delayed copy of the main bubbly synth recording and pitching it down (and mangling it a bit) with an old spectral FFT plugin called DtBlkFx.

I used many many instances of Nebula, with my own products. Probably every product was used at least once somewhere in the album. My tube radio and cassette programs were used pretty extensively to alter the tones, and my various reverbs were used everywhere too. I used Waves’ F6 dynamic EQ whenever I needed some really surgical EQing (really nice plugin). After the mixes could have been called finished, I decided to record every separate element (everything was recorded to mono originally but now I had many layers I had separated out of the recording in various ways, and some overdubs and effects tracks) to cassette tape and back using my Marantz deck. This is the deck I used to make Cassette Deck 3. I had used my Nebula tape programs but I wanted to go beyond what the programs can do. I recorded every track in each mix, using various tapes (and tape types), in single and double speed mode (this deck can do 2x speed), and redigitized the output. This was done in one pass while monitoring off of the tape, so the deck was being used as an effects loop, and this allowed me to hear exactly what the saturation on each sound sounded like in the context of the rest of the mix.

I think the main benefit of that was just the tape compression. I didn’t really drive the tapes hot enough for distortion usually. But the compression brought the mixes together more than they were previously. Track 9 is one exception where I used various levels of tape distortion towards the end, to make it more dramatic. After everything was ‘tapified’, I did the final mix-down in my more recent modular system, using a really high quality input/output module and several mixer modules. In a few cases I applied some subtle filtering or other processing while doing this. Then I used my and Michael’s M-EQ for mastering, with some subtle C660 compression.

There’s almost surely some things I’m forgetting here, but these are probably all of the major updates.

• new skins allow you to press LED buttons on the skins to switch between the programs. this really improves the experience; especially with the LFO programs where you can select different LFO shapes with the push of a button. there are 3 separate skins. one is for the EQ boost/cut, bandpass, notch, and pass-through programs. another skin has the LFO programs, and the last skin has the envelope follower effects.
• it’s now called Yourei EQ/BP/N v2
• i made the ability to use this release as an EQ a more up-front focus of this release. the peak boost and cut ability is easier to access and use than they were in the old release (you could use the programs for these purposes but it was really finicky to set up for the boost). the original hardware didn’t offer these (it only has bandpass and notches), i made them by combining the bp and notch samples with the pass through samples, which allows us to have some really nice peak and cut bands for EQing purposes. these are probably the most commonly useful things you can do with the programs in this release so they have been made a focus. It’s a really nice sounding boost or cut that can be placed anywhere in the frequency response.
• the low end ripple that happened as a result of truncated impulses has been removed. there is some subtle ripple that appears if you have a sharp filter at low freqs, but that is seen on the hardware too, so it’s accurate.
• 48 and 88.2khz sample rates were added.
• many of the programs combined the 3 filter types (notch, bandpass, peak) into one program, allowing access to all 3 in the one program. that sounds good but it drastically increased load times (due to having to load the samples for modes you might not even need), and the knob to switch between the modes worked in a confusing and not very good way. now the notch, bandpass, and peak all get their own separate programs, which allows some programs load faster. not having them all in one program is no loss because the new skin has buttons that you can use to switch to whatever mode you want. this is actually a really big improvement in terms of organization and usability.
• all programs load at least a little faster than before. some load a lot faster. the main notch program still takes several seconds to load on my older PC, but it has almost 3,000 samples and there’s no way to make it load faster.
• LFO/envelope programs have smoother motion because i’m using a faster program rate. this isn’t as noticeable at slower modulation speeds but it sounds more authentic to analog modulation (like you’d get in a modular synth for example) at faster rates, especially with the 96khz programs. this does require slightly more CPU compared to the original programs, but i put those out over 10 years ago! the notch at its widest setting sounds like a nice and smooth phaser.
• the LFO width control works in a better way now. the % you set it to will always cause the LFO to travel that % of the way, between where you set the base frequency, and the maximum frequency. this means the LFO will never ‘peg out’ at the highest frequencies, like it could in the past with some settings. kind of hard to explain but it’s just better this way.
• the envelope follower programs are more useful now. in the past there was a bug (this was actually due to a bug in nebula itself), where the ‘attack’ would usually always be instant. you wouldn’t actually be able to get the filter to smoothly slide up (or down) during the attack, before returning for the release. it would just kind of jump to the highest (or lowest) point, then slide back for the release. now you can actually get slower attacks. this really opens these up as nice little auto-wah effects. these also got a threshold control now which gives you an easy way to control how much ‘travel’ the filter does.
• evened out the levels of the bandpass/peak filter across the range. they’re still not exactly the same at all settings but neither was the actual gear. You probably never noticed the levels being uneven in the past because it was usually subtle, but it is improved now.
• removed the external sidechain envelope follower programs because for whatever reason, years ago, Acustica removed the ability to introduce an external sidechain into Nebula.
• feedback is only found in BP modulation (or the for automation) programs, because it’s just not good in the notch and peak programs. the reason is because those programs have the entire frequency response in them, and all of it gets fed back, which just gives you a comb filtering effect. with the bandpass programs only the audio around the cutoff gets fed back.
• the tone programs have been updated to match the updates i gave them for the ‘Cuptones Vol 1’ release, which includes cleaner harmonics and reduced artifacts.
• new manual.

I will be sending out links to the newly updated files by email soon. Here are all of the updates to Tube Radios Direct-

*added new ‘hot’ programs which allow you to get much louder distortion than ever before. i’ll stake my reputation that few if any Nebula programs out there can do anything like this level of distortion, without becoming nasty. these programs do not have the infamous ‘chirp’ that has plagued Nebula for ages, when driven too hot. of course, whether these distortions sound ‘good’ will depend on what you’re processing, as with any distortion. but they absolutely can be made to do heavy distortion and still sound great. this is the biggest and most important update. when i originally sampled these radios (about 13 years ago) i got a range of dynamics sampled, from queiter levels without much distortion, to much higher levels with lots of harmonics. those higher levels could never be reproduced until now.

*the tube compression aspect of the sound (caused by tube saturation when the radios were driven hot) was stripped out of the main tone programs. the tone programs still react dynamically- the frequency response usually slightly changes as a program is driven hotter, and the harmonics also respond dynamically. the actual compression has now been moved over to a separate batch of programs (with their own skin for selecting them!) that only do the compression- your tone is completely unaffected by them. doing it this way is better for many reasons-
1) having both the tone and compression combined in one program doesn’t allow instant reaction due to limitations with nebula. so we couldn’t have the realistic instant squishing effect a radio would do.
2) this also allows you to use the tube compression without a tone program, opening up new possible uses. on default loading they provide instant squishing like the radios would, but i added an attack and release control (modeled from one of my compressor releases), which when adjusted allow you to use these programs in any situation where you need compression. each of the many compression programs has a different knee, so they all react differently.
3) you could also do the opposite and use the tone programs for radio filtering or distortion, but without the compression. or if you do also use compression, you can adjust exactly how much you have, without it being tied to the amount of distortion you get from the tone program (as it was in the past).

*if you want the hot distortion, you probably will need to use a compressor instance before the tone instance, to reduce the dynamic range at least a little, and to catch any loud peaks. you can experiment though and do it however you like! sometimes the compressor sounds best after the tone instance, to my ears. if you’re going for distortion though it usually is best for the input to not have lots of dynamics. you’ll get a feel for exactly how much is ‘too much’ on your own.

*added a new control to the tone programs-

f1 mode – switches between 3 modes for the fundamental element of the effect-
1) the default setting = ‘1’ and gives you the normal result, which is the sampled radio tone. this is the effect on the frequency response and it does react dynamically. it’s usually various levels of filtering to the low and high frequencies, but a few programs are almost flat, while still doing ‘something’ to your sound.
2) a setting of ‘2’ bypasses the sampled fundamental and just passes your dry input to the output. this is actually a really awesome option and i’m amazed i hadn’t thought of it years ago! what makes this useful is that you turned off the radio filtering, but still get the saturation/distortion. this can be used with the normal programs for subtle harmonic enhancement, or with the hot ones for loud overdrive distortion, without the filtering!
3) setting = ‘0’ mutes the fundamental altogether, meaning all you get is the distortion! the main part of the sound has been removed! this is usually pretty weird sounding and would rarely be useful, but it’s neat so i included it as an option.

*added another new control to the tone programs-

THD mode – this switches between different modes for the harmonic distortion portion of the effect.
1) default setting of ‘1’ again gives the normal sampled harmonics, which are lower in level and provide a subtle tube enhancement of the signal.
2) settings of 2 through 5 are the four different ‘overdrive’ modes. they give you heavy distortion, and each of them usually gives a different harmonic profile, and different sounding results. these settings are only available in the ‘hot’ programs. try all of them and just pick the one you like best!
3) setting = ‘0’ this completely mutes/removes the harmonic distortion. use this if you want a clean sound. not available in the ‘hot’ programs, only normal.

*those two new controls, combined with the now separate compressor programs, allow you to use any combination of the 3 elements of the ‘radio effect’, in any ratio you want. this includes the 1) radio filtering/tone, 2) the distortion, 3) the compression. use any or all of them, however you like!

*programs were further cleaned up of aliasing and other artifacts (like the infamous low freq ripple, which wasn’t very noticeable in most of these already). in some cases (usually higher input drive) the artifacts could be audible in the previous versions, but not here. now they’re almost always super low, well beyond inaudible.

*Admiral DO3 program greatly improved. i think somehow i damaged the samples for this one when i updated it last time. i don’t know how that got past me because the frequency response was obviously wrong. now it’s how it should be- very accurate. the frequency response readout in the skin was also updated to reflect this.

*peak hold indicator has now been added. i came up with this idea and have been adding it to my most recent releases. it’s just a readout value that tells you what your loudest recent input signal was at, in dBFS. handy for making sure you don’t go over 0dB. the skin was updated to include this peak value readout.

*added an alt version of the Admiral DO4 program. the original was accidentally recorded with some delayed feedback in the signal chain, but I liked it so i kept it in the collection. this new alt version removes the delays, offering a program more like the others. the harmonics still have the delay, so the hot programs can sound very interesting.

*a lot of the programs previously inverted the input signal. now they don’t.

*some of the impulses were improperly aligned, now they’re perfectly aligned. this is kind of technical and i’m not going to go into detail, but in some cases this fix did subtly affect the sound for the better.

*feedback control has been improved. you can now usually travel all the way to 100% or close before getting runaway feedback. this allows you to dial it in more precisely, if you ever use it.

*these changes make this updated version of the release the best sounding it has ever been.

*last and definitely not least, i got Max to modify his skin and make a new one for the compressor instance. the modifications were needed to accommodate all of my updates.