Arturia releases LEXICON LX-24

[jamcat]

Arturia’s doesn’t sound better. It sounds very similar to the UAD 224, but when you A|B them, you start to notice that Arturia put some veneer on it to give it extra sheen. It kinda gives it a wow factor at first, but it keeps it from dissolving into the mix the way a real Lexicon does. The UAD Lexicon 224 blends right.

[kmonkey]

Arturia’s doesn’t sound better. It sounds very similar to the UAD 224, but when you A|B them, you start to notice that Arturia put some veneer on it to give it extra sheen. It kinda gives it a wow factor at first, but it keeps it from dissolving into the mix the way a real Lexicon does. The UAD Lexicon 224 blends right.

Blending is a matter of your own skills. Besides that UA version is questionable...

There were countless examples posted on the UA forum a long time ago and on Gearslutz (at that time) where people with real hardware posted direct comparisons and presets and couldn't make software (UA) sound like hardware. The difference wasn't subtle. I don't remember the link, but I remember clearly that at least three people posted differences, like several presets (FXP or something was then) and hardware parameters, etc.

They clearly pointed out that parameters on the plugin don't correspond to the same position on hardware and asked why they can not even approximate to get it sounding the same. I remember the most obvious difference was in the low-end, which was really clear to hear.

After a long back and forth an information resurfaced that the actual code for 224 was lost, like lost literally. And that the plugin is an approximation, and it was made by a different company, not UA directly!?

At that point, it was questioned why the UA video says "code for code " or if it was something like that, but then someone said it's a perfect approximation, so to speak.

It goes to show you the level of disengenuous information being thrown in marketing videos. It piqued my interest and I researched on my own back then and found information online on who was original developer of 224 algos was (I forgot the name), and there really was information on code being lost forever (as in the code was lost due to machine failure or something tragically silly).

Anyway, I am aware we are talking a lot from my memory, but it was more than a decade ago. If you don't believe me, ask yourself why UA themselves never posted A vs B; if it was code for code, it would all be over their website.

Besides, if you don't believe me (I suggest you don't believe anyone these days), try to research on your own.

[kmonkey]

edit: found some traces (it is a different website then the one I was reading decades ago):

The original developer of the algorithms for the Lexicon 224 digital reverb was Dr. David Griesinger. He was the key figure behind iconic algorithms like Concert Hall and Random Hall. Griesinger was a pioneer in digital reverb and instrumental in Lexicon's dominance during the late '70s and '80s.

As for the source code, it’s widely acknowledged in the audio engineering community that the original code was lost. According to Michael Carnes, a former Lexicon engineer, the code was stored on VAX systems, which were eventually sold off when Lexicon transitioned to SPARC stations. The source code was never migrated. There may have been printed listings, but their existence isn't confirmed.

[wintoid]

In these modern times, you would have thought the ROMs could be dumped and fed into some AI that would generate readable code from them...

[Gamma-UT]

In these modern times, you would have thought the ROMs could be dumped and fed into some AI that would generate readable code from them...

Even if that would work (which it won't give the state of real-world AI vs idealised AI), the only company that can do that legally, or at least admit to doing it publicly, is Lexicon.

Also, the reality is that the "source code" probably isn't even that useful. The MPU is an 8080, which is basically only capable of passing parameters to a hardwired digital signal processor (well, a 16bit ALU) made primarily from a combination of 74-series logic gates. Assuming they didn't sandblast the IDs off the top, this is reasonably straightforward to reverse-engineer in principle as the contents of 74-series logic gates aren't trade secrets.

This also assumes that the minutiae of the logic board is what delivers that final 5% of accuracy compared to the analogue input and output stages and the 12->16bit->12bit dynamic-range compander logic. Unless the ALU design took some odd shortcuts or maybe handled saturation vs overflow in an unexpected way, your reverse-engineering effort really wants to go into that section in all likelihood.

[wintoid]

Also, the reality is that the "source code" probably isn't even that useful. The MPU is an 8080, which basically only capable of passing parameters to a hardwired digital signal processors (well, a 16bit ALU) made primarily from a combination of 74-series logic gates. Assuming they didn't sandblast the IDs off the top, this is reasonably straightforward to reverse-engineer in principle as the contents of 74-series logic gates aren't trade secrets.

I had no idea we were talking about such an old unit. When you said 8080 I went and looked it up and good grief 1978. Lexicon must have been well ahead of the game in those days.

So yes, I agree, the "source code" wouldn't really get you anywhere other than how to send the control signals, and therefore it doesn't really matter that it's lost!

[jamcat]

After a long back and forth an information resurfaced that the actual code for 224 was lost, like lost literally. And that the plugin is an approximation, and it was made by a different company, not UA directly!?

I’ve been involved in discussions here and elsewhere about this very issue several times in the past. Here is the last thing I had to say about it two years ago, based on a pretty deep dive:

So Michael Carnes said on several occasions the 224 code was lost and Dave Greisinger told him he doesn’t have a copy.

Casey Dowdell said, “AFAIK David does still have the 224 code base. It was the 480 code base that was somehow destroyed.”

That’s far from a direct firsthand confirmation from Dave Greisinger himself.

Casey also said:

I need to correct myself.

As I was thinking back over the related conversation, it wasn't that David was thinking about doing a plugin, he was wondering about doing new algorithms for existing 224s.

Sorry for confusing the issue.

So his recollection on the matter may be unreliable.

Michael, on the other hand was currently at Lexicon porting algorithms to plugins at the time when he said it, and seems to have been speaking from fresh firsthand knowledge about what algorithms were available.

So, I would say all of it is inconclusive one way or another.

I initially took the same position as you, but with deeper research it became apparent that the truth on the matter isn’t so clear cut.

While it is inconclusive from the public record whether or not the UAD Lexicon 224 is the original algorithms, it is undisputed that Arturia’s version is not. And none of this changes the fact that Arturia added their own special gloss sauce to it that isn’t present in a real 224.

Arturia also changed the basic function of the bass/mid reverb time controls, while Universal Audio stayed true to them.

[kmonkey]

After a long back and forth an information resurfaced that the actual code for 224 was lost, like lost literally. And that the plugin is an approximation, and it was made by a different company, not UA directly!?

I’ve been involved in discussions here and elsewhere about this very issue several times in the past. Here is the last thing I had to say about it two years ago, based on a pretty deep dive:

So Michael Carnes said on several occasions the 224 code was lost and Dave Greisinger told him he doesn’t have a copy.

Casey Dowdell said, “AFAIK David does still have the 224 code base. It was the 480 code base that was somehow destroyed.”

That’s far from a direct firsthand confirmation from Dave Greisinger himself.

Casey also said:

I need to correct myself.

As I was thinking back over the related conversation, it wasn't that David was thinking about doing a plugin, he was wondering about doing new algorithms for existing 224s.

Sorry for confusing the issue.

So his recollection on the matter may be unreliable.

Michael, on the other hand was currently at Lexicon porting algorithms to plugins at the time when he said it, and seems to have been speaking from fresh firsthand knowledge about what algorithms were available.

So, I would say all of it is inconclusive one way or another.

I initially took the same position as you, but with deeper research it became apparent that the truth on the matter isn’t so clear cut.

While it is inconclusive from the public record whether or not the UAD Lexicon 224 is the original algorithms, it is undisputed that Arturia’s version is not. And none of this changes the fact that Arturia added their own special gloss sauce to it that isn’t present in a real 224.

Arturia also changed the basic function of the bass/mid reverb time controls, while Universal Audio stayed true to them.

You mean Arturia added their own interpretation on bass/mid, while UA added their own interpretation, neither of which sounds like actual hardware.

Your basis about Arturia adding "sheen" is based on the difference when compared to UA while at the same time UA never sounded like that hardware to begin with?

People pointed out that the original 224 code is lost; therefore, it is and was impossible for UA to make code for code.

I can understand your point when you say "it's inconclusive," but it feels like a strawman argument.

Be honest with yourself and let me know what is more likely in the case where:

A - By the community claims and people who worked at Lexicon - The code is lost, therefore, UA was not able to recreate the code line by line
1 . Several individuals, right after the UA release, posted examples, presets, and descriptions of sound differences - all at the same time expressing inability to even approximate sound with UA controls on the plugin
3. The sound examples and comparison to the hardware posted by unrelated individuals weren't subtle

B - UA never had access to the actual code
1. Which results in the plugin sounding different than the hardware
2. UA never disputed claims that the 224 code is lost; they could have done it at least in their own forum
3. UA never posted samples of hardware vs their software (which seems a bit surreal)

And my actual point is that while you may or may not admit that the UA version is not true to the original, your point of which one is better is pretty moot because you are comparing two products that are trying to mimic a specific hardware - out of at least one we know does not sound like the hardware (UA) - while at the same time you don't have access to the real hardware.

How can you actually tell whether the "sheen" is authentic to the hardware or not? Every time you are pointing to the fact that "UA" does not have "sheen", therefore Arturia did something different or inferior, but in fact, everything is pointing to the conclusion that UA isn't faithful as well.

Therefore, my point is: So you are left guessing when it comes to authenticity or "sheen". Hence my wondering - how can you tell?

[Uncle E]

There were countless examples posted on the UA forum a long time ago and on Gearslutz (at that time) where people with real hardware posted direct comparisons and presets and couldn't make software (UA) sound like hardware. The difference wasn't subtle. I don't remember the link, but I remember clearly that at least three people posted differences, like several presets (FXP or something was then) and hardware parameters, etc.

Yes, the real thing sounds more dense and it isn't subtle. But the UA sounds great, anyway, and there were many times where I preferred its cleaner sound. Try using tape and summing plugins after it if you want more of that hardware density.

[kraster]

There were countless examples posted on the UA forum a long time ago and on Gearslutz (at that time) where people with real hardware posted direct comparisons and presets and couldn't make software (UA) sound like hardware. The difference wasn't subtle. I don't remember the link, but I remember clearly that at least three people posted differences, like several presets (FXP or something was then) and hardware parameters, etc.

Yes, the real thing sounds more dense and it isn't subtle. But the UA sounds great, anyway, and there were many times where I preferred its cleaner sound. Try using tape and summing plugins after it if you want more of that hardware density.

I was using an OG recently and I was really surprised at how upfront and full it is.

I've used them before but this was the first time in a long time and was more used to the emulations since I last used one.

The low mids are particularly dense. Almost the kind of thing that you'd be tuning out of a modern reverb but when pulled back into the mix it has a lovely way of enveloping the sound.

As you say it's not subtle.

[OvoTkyo23]

The lexicon does sound pretty good and you can’t beat the cost as well. It can be a little heavier on the CPU side but not to the point it is unusable.