Cherry Audio Releases Mercury-4 Compuphonic Synthesizer!

[TheHelioSequence]

Yeah, normalizing isn't the way to go. Matching by RMS or LUFS would be much better.

[wwjd]

Yeah, normalizing isn't the way to go. Matching by RMS or LUFS would be much better.

yes, that, totally

[Friendly Noise]

Here is the second part. I focused on matching levels. That was sometimes possible, sometimes not. Details about it in the video:

https://youtu.be/LmsQrfHPmG4

[wwjd]

Friendly Noise, great job on matching levels on this. I feel this comparo is much closer than your previous. was super cool to listen to this! Thank you for taking the time to compare to the real gear!

On the PIANO patch, a couple quick tweaks, and I was able to almost perfectly match your JP4 piano with my Merc4. (or sounding nothing like it, not even close, depending on perspective)
Sounds like Cherry Audio had a different tuned Jup4 than yours?

[Friendly Noise]

Friendly Noise, great job on matching levels on this. I feel this comparo is much closer than your previous. was super cool to listen to this! Thank you for taking the time to compare to the real gear!

On the PIANO patch, a couple quick tweaks, and I was able to almost perfectly match your JP4 piano with my Merc4. (or sounding nothing like it, not even close, depending on perspective)
Sounds like Cherry Audio had a different tuned Jup4 than yours?

Thanks for your comments. I will take a second look at the audio files of the first video and try to get the levels closer. So maybe there will be a second “level matched” version of Part One.

Regarding the presets: the factory presets in the Cherry Audio are so dark compared to the presets in my Jupiter 4… Not only filter settings, but also other parameters seems to be different.

The Jupiter 4 came already with the not editable factory presets. I think they could be globally affected by tweaking the trimmers in the voice cards, but that means -in the case of the Mercury 4- that the overall sound of the unit would be equally darker. But this not the case, as the overall brightness of the oscillators seems to be very similar to the hardware.

A little mystery here…

[Friendly Noise]

Regarding the limits of comparisons:

In the third video, I will compare unison sounds on both instruments.

As soon as I switched Unison mode on, it was completely clear that it wasn’t possible to get the same sounds on both instruments. Why?

The Jupiter 4 is said to be a temperamental beast, almost never completely in tune, voices drifting apart…
Well, mine doesn’t really behave that way. When I got my Jupiter 4, the tuning was horrible and the synth almost unusable. I tuned the synth using the trimmers in the voice cards in a way that when playing some sounds like octaves you still get that pleasant phase cancellation sound between VCOs. That’s very noticeable in Unison mode. If someone’s unit is differently tuned (some people go for a slight detuning between voices), then the Unison mode will be accordingly affected. If there is more detuning between voices, then the Unison mode will sound fatter. If A, then B.

It seems that the Mercury 4 is tuned for fatness and chorusing when used in the Unison mode, no matter the position of the drift knob. For that reason, you will hear obvious differences in sound when I post the examples with both instruments in Unison mode.

[wwjd]

not sure it is worth all that effort, but as long as you are having fun!!!

my biggest wonderment is always "can one tweak software well enough to sound like the real hardware", and using your recordings the answer is yes. Sure, the raw tone of one oscillator is not exact as old 1970's electronic hardware parts, but its easily close enough to be mistaken by average listener (or "nothing like it, not even close" as some might say)

[mholloway]

What is the proposed end-goal of this ongoing matching? I don't really get it, tbh. I think we all know Cherry Audio products don't sound "exactly like" the synths they are based on, and we also all know there's no such thing as "exactly matching" an analog synth in the first place, due to their inherit instability / variation between units.

Not criticizing the effort, just very curious about clarification on the purpose / goal of the project. It's such a fine-level of comparison that I struggle to see the meaningfulness of it for any practical purpose. Is it to make people feel good about buying the emulation because it's close? Or good about not buying it because it's not "exact" ? Or....?

[Arashi]

we also all know there's no such thing as "exactly matching" an analog synth in the first place, due to their inherit instability / variation between units.

Depends on the synth. Not all of them are wildly different from one unit to the next (that whole idea is overblown), and even when they differ, they have a distinct character that’s clearly recognizable, and an emulation can either capture that character or not.

Emulating analog synths is really tricky, but people are getting better at it, and there are a few that I’d say have actually nailed it. For example, Korg’s ARP Odyssey plugin really can sound pretty much exactly like the hardware — I know, because I have the hardware too; I’ve copied many patches from the hardware to the plugin and have yet to come across one I can’t match (by which I mean “either of these could be the hardware”, not “they produce literally identical waveforms that zero out”). I say it “can” sound the same because it takes some work to tweak it by ear, and it’s not as simple as just eyeballing the position of the sliders.

So far I haven’t heard a Cherry emulation that really convincingly sounds just like the hardware, though I haven’t had a chance to use the hardware version of half of them (the Polymoog, Memorymoog, or MG-1 might be dead ringers…I wouldn’t know, but based on the others, I doubt it). But that doesn’t stop me from enjoying all of them. They don’t have to be perfect emulations to be good instruments. I’m still interested in comparisons, though, because I think it’s useful to track the state of the art.

I’m with you on the question of defining the goal and having an actual purpose in mind, since otherwise these comparisons can and will descend into zooming in on waveforms and missing the whole point of synths as musical instruments.

[briefcasemanx]

we also all know there's no such thing as "exactly matching" an analog synth in the first place, due to their inherit instability / variation between units.

Depends on the synth. Not all of them are wildly different from one unit to the next (that whole idea is overblown), and even when they differ, they have a distinct character that’s clearly recognizable, and an emulation can either capture that character or not.

Thank you!

[chk071]

What is the proposed end-goal of this ongoing matching? I don't really get it, tbh. I think we all know Cherry Audio products don't sound "exactly like" the synths they are based on, and we also all know there's no such thing as "exactly matching" an analog synth in the first place, due to their inherit instability / variation between units.

Not criticizing the effort, just very curious about clarification on the purpose / goal of the project. It's such a fine-level of comparison that I struggle to see the meaningfulness of it for any practical purpose. Is it to make people feel good about buying the emulation because it's close? Or good about not buying it because it's not "exact" ? Or....?

It's always very interesting for me when people compare an emulation to the real deal. I think you're exaggerating the variation between units a bit.

I miss a comparison of the filters though. The raw waveforms sound very similar on most synths. Although, in the example, you can already hear that the square sounds a bit different (as usual, not a perfect square on the real deal, rather a bit PWM-d sqaure).

[gentleclockdivider]

It's not really quieter, it's just thinner, less low-end, less saturation by the VCA at a similar loudness level. The comparison is fair.

I have this feeling with every cherry audio plugin
They all have this flat ceiling that prevents it from breathing ,if you understand what I mean .
First encountered with all osc sliders to max on ca 2600 , instant boxines

[TheHelioSequence]

Regarding the limits of comparisons:

In the third video, I will compare unison sounds on both instruments.

As soon as I switched Unison mode on, it was completely clear that it wasn’t possible to get the same sounds on both instruments. Why?

The Jupiter 4 is said to be a temperamental beast, almost never completely in tune, voices drifting apart…
Well, mine doesn’t really behave that way. When I got my Jupiter 4, the tuning was horrible and the synth almost unusable. I tuned the synth using the trimmers in the voice cards in a way that when playing some sounds like octaves you still get that pleasant phase cancellation sound between VCOs. That’s very noticeable in Unison mode. If someone’s unit is differently tuned (some people go for a slight detuning between voices), then the Unison mode will be accordingly affected. If there is more detuning between voices, then the Unison mode will sound fatter. If A, then B.

It seems that the Mercury 4 is tuned for fatness and chorusing when used in the Unison mode, no matter the position of the drift knob. For that reason, you will hear obvious differences in sound when I post the examples with both instruments in Unison mode.

Out of curiosity, did you use oversampling on Mercury 4?

[Arashi]

I have this feeling with every cherry audio plugin
They all have this flat ceiling that prevents it from breathing ,if you understand what I mean .
First encountered with all osc sliders to max on ca 2600 , instant boxines

It does seem like the high end is rolled off a bit too aggressively (if that's what you mean). You can immediately hear the difference when you compare Eight Voice and OB-E; the latter has lots of high frequency sparkle and Eight Voice sounds muffled by comparison. Most Arturia plugins have the same problem.

Cherry recently updated all their plugins to add the oversampling feature. Might be worth testing them again at 2x/4x to see if it makes any difference.

[Friendly Noise]

Regarding the limits of comparisons:

In the third video, I will compare unison sounds on both instruments.

As soon as I switched Unison mode on, it was completely clear that it wasn’t possible to get the same sounds on both instruments. Why?

The Jupiter 4 is said to be a temperamental beast, almost never completely in tune, voices drifting apart…
Well, mine doesn’t really behave that way. When I got my Jupiter 4, the tuning was horrible and the synth almost unusable. I tuned the synth using the trimmers in the voice cards in a way that when playing some sounds like octaves you still get that pleasant phase cancellation sound between VCOs. That’s very noticeable in Unison mode. If someone’s unit is differently tuned (some people go for a slight detuning between voices), then the Unison mode will be accordingly affected. If there is more detuning between voices, then the Unison mode will sound fatter. If A, then B.

It seems that the Mercury 4 is tuned for fatness and chorusing when used in the Unison mode, no matter the position of the drift knob. For that reason, you will hear obvious differences in sound when I post the examples with both instruments in Unison mode.

Out of curiosity, did you use oversampling on Mercury 4?

Highest quality was used.