David Griesinger, the man responsible for all Lexicon's reverberation algorithms is alive and well.

lukpio

That's what I thought. They were talking about him on the 960 literature, and not in the past tense. Or maybe he wrote some of that actually.

Nice to see how many myths there are about reverbs.....


Harrie

I don't know who at Lexicon told you that but it is total bullshit. It would be a snap for any decent experienced IT engineer to figure out the 480L algos just by analysing the eeproms or roms in the thing. It's a very basic machine code reverse engineering task. It would take some time to work it out completely but there's is no difficulty at all.

I still think that posting some AB samples of some audio processed through the same algorithms of PCM91/480 L could be much more educative than looking for some persons in this world or nether worlds

Yes, there is. At least between Altiverb and Waves. They both have 480L IRs, but the Wave is much more closer to the Lexicon. Somehow, they are more dense with the Waves.

The thing is convolution reverbs are better sounding with sampled acoustic spaces than with sampled hardware.

Well, it's sometimes nice, but I still use Waves and Altiverb on many more occasions.

I still find the 480L more desirable, but I wouldn't part from Altiverb for a PCM91.

regards

Beside, they must have figured this out when putting the 480 algos in the 960.

That is IF they want to go deep inside the units. It is not evident that they did that. There are a lot of people here loving the good ol 480 but hating the 960 (exagerating). Search the forums..... And recreating the modules is not very easy when changing DSP-platforms... The same thing happened with the Eventides. The 3000 hardware was so different compared to the 4000 that some of the modules where not recreated. So they sound really different beside the converters etc...


Harrie

Does 480 L have AES/EBU I/O ?

Come on.... It's not that big of a deal. Changing DSP-platforms is no big deal either since new DSPs are much more feature-laden than older ones. It's simply porting some machine code. I guess they do want to go inside their own products !! I'm an IT engineer myself besides my musical wanderings and trust me, really it's easy. Theoretical conception of the algo itself is a very big deal. Why it sounds good or not is highly subjective matter so there's some art involved in that. Implementing the thing on a given platform is easy like lego building, if the platform has the processing power. And if the Eventide 4000 serie is different from the 3000, I guess it's only by choice, not because they were forced by hardware, unless the 4000 has less brute force processing power than the 3000 and I don't believe it's the case.

Well I just clocked the PCM 91 via SPDIF to Big Ben through my DAW and no Lexicon conversion, analog etc... Just amazing sounding. If someone has a 480L we can both get a WAV file and test... I bet you can’t tell... or it would be damn close to as good as most anything you've heard. No complaints here.


Go listen to the verb sound on the Rolling Stones Angie guitar intro!!!!… Maybe it’s in the Lexicon somewhere hiding! Its stunning!

The 480L sounds different when heard through the analog converters than through the digital.

It sounds bigger and warmer through the analog converters.


The analog converters adds to the character.


This happens all the time.

The Quantec QRS sounds different than the QRS XL which both sound better than the 2402.

The Dynatron 255 doesn't sound as deep as the EMT250.


Even the new Princeton Digital 2016 sounds slightly different than the Eventide SP2016.


A lot of the older effect processors were designed around the converters so the converters have become part of the sound.

That is very true IMHO. Also, when the 480 was first developed, a lot more of the circuitry was discreet as opposed to being DSP.

So yeah, technically you can port algorithms over from one unit to the other. And if that was all there was to it, then all AD converters could sound the same, no?. They obviously don't. The surrounding circuitry is crucial.

For this reason I'm told by Lexicon that it is impossible to cheaply recreate the 480.

As an aside, I just purchased a Roland R880. (Thanks "Sounds Great.") Anyway, I haven't opened the mainframe unit, but it's a big mother, heavy and appears to have a shit load of discreet electronics. It sounds awesome.

Roland later used some of those algorithms in the SRV-330, which while a great sounding box, can not compare with the "largeness" of the former unit.

OK, but does it have AES EBU or not ?

A good convertor BTW should not change the sound (to better or worse) at all, it should be rather a neutral mirror. If the reverbs sound "better" with Lexicon AD DA conversion, then it must be an AD DA with some "exciter".
I seriously doubt whether the old 480 L convertors can be in pair with today´s top ones (Lavry etc.). Lexicon is basically a purely digital reverb. Any extra AD DA conversion is always generally spoiling the sound a bit. If there is a feeling that with an extra AD DA it "sounds better", it must be rather then some unwanted by-product coming out of some imperfections ...

But a reverb unit is not simply an AD/DA converter. Your argument in this instance is meaningless.

It's about sound, and what sounds good to your ears. How it is achieved is besides the point.

In this case, a lot of people still use the 480. Is that because it has lower bit level converters than a Lavry and therefore sounds bad?

I think not.

My question still remains - does it have digital IO (AES EBU) or not. If not, there is no other way than to use it analogue ...

I believe it was an option. So yes.

I agree on both points...

The 480L has a color that people like. But how much better can it sound with bypassing the "OLD" D/A-A/D converters and "OLD" analog section????

Can the PCM 91 keep up with the 480L used in a digital format, both clocked to Big Ben... ???

I think this is a good test!

Then compare the samples in the analog world.

This would conclude a comprehensive test.

Why not to say also 960 L ? (if we forget about all the extra effects, LARC, surround etc. ?)

I would be really interested to know whether say Large Hall or whatever common preset on PCM 91, 480L pr 960 L would sound the same or not on the same source. According to Lexicon, the basic reverb algorithms are the same on all the units.

If yes, then for a user like me there is absolutely no need to think about 480 or 960. If there IS a reasonable difference in the sound quality of basic reverbs, then it could be a reason for some slutty dreams (and a used 480 L is not at all that astronomically priced as 960 L)

Hey my Eventide 910 delay has a compressor and two expanders in it I believe. It sounds... well it has character, and I don't think that's a bad thing. I have two of them and they sound pretty different from each other in fact (probably because they are in need of proper calibration, but i like the distinct sound they produce now). I love running my sidstation through a 910.

I know that they are 'delay/pitch' units, not reverb, but they are the predecessors of these units we are talking about.

I would be curious of such a test. I compared side by side a 480L and a 960, and I was a bit disapointed with the 480L algos in the 960.

But I concluded it could have been the converters deceiving me.

anyone having access to the three of them ?

regards

I don't remember the guys name, but he was head of sales and had been with Lexicon for a very long time. So if they CAN do it (according to you), then they don't WANT to do it (because they haven't)....and for the life of me I don't know why because it would sell quite well.

Thrill is exactly right on converters in reverb processors.

The weird thing is that the converters for a reverb processor have contrary requirements than the converters for general music. The latter need to be fast, precise, transparent etc. whereas reverb's sound gains from "bad" converters, that have a lot of smear (to blunt transients), phase distortion that comes from complex analog anti-aliasing filters, etc. because reverberation is ALL ABOUT phase distortion.

The early digital processors (like EMT250 or AMS RMX-16) had A/D converters of 12-bit mantissa plus 2-bit scaler, what is equivalent to 16-bit system (not 14-bit!) and complex analog stages. It really has little in common with contemporary oversampling sigma-delta ADC's with FIR filters that every maker today use.

So a lot of character of the early digital processors comes directly from converters and this is the cause why new Eventide 2016 sounds different than its predecessor from 1983, Dynatron 255 with modern external converters sounds different than EMT250, and Yardstick sounds different than the original QRS (which BTW had brick-wall filter at 8kHz!). I strongly belive that reverberation algorithms themselves in these recreations are more or less the same as in originals.

Regarding Eventide H3000 vs. DSP4000 dispute I agree with Jack. I cannot imagine a single DSP function that can't be implemented in another DSP-platform. H3000's character comes (again) from converters, the dual-slope Sony A/D converter (CX20018), 9-order analog anti-aliasing filters and the D/A converter that any designer today would like to get rid of.

Lukpio

Huh ? Meaningless. DSP is DSP is DSP is digital. We are not in the 50s anymore. A 480L is not the ENIAC tube computer. The algo part of it can't be discrete. Analog output drivers can be discrete.

Flawed logic here. Algos and AD converter have nothing to do with each other. It is 2 different sections in the unit. You can definitely port an algo from a machine to another. Surrounding circuitry has nothing to do with the algos.

Maybe but it has nothing to do with the algos themselves.

The AD converters may sound different obviously. But they are not in question here.

In advanced tech biz, the sales managers are often a bit lost technically. It's the same where I work. But they know what they do on the marketing side. I don't think a recreation of the 480L, even perfect, would sell that well. You know it wouldn't sound the same anyway. You would find hundreds of "vintage 480L" owners who would swear all over the net and elsewhere that the new unit doesn't sound as good because it's not point-to-to-point wired or whatever reason. Then they'll advise any newbie wanting to buy a new one : "No don't !! It's more expensive than the vintage models on ebay, it's not worth it etc. etc."

You're chasing your own tail here.

I realize that A/D converters do not use algorithms. The point I was trying to make (and it obviously flew over your head) was that certain parts in electronics can be similar or the same. Like chips in AD converters. But since devices that use similar chips sound different (and likely have cost differential) then it stands to reason that other factors come into play.

Wish all you want, but you will not find a cheap 480. Until the market puts them there.

But since they sound so good, that time is still a long way off. In the meantime, can I recommend the Roland R880? Not as good, but still up there sonically and reasonably priced (by the market.)

There is a card you can buy that is AES Digital I/O (2 channels in/4 channels out) I don't know if you can actually find it for sale anymore, but you can find 480's with the card built in.

I've been checking out the Lexicon PCM-70, 80, 90 series, but I think I'll wait for some reviews on the Bricasti M7.

http://www.bricasti.com/

Apparently it's made by the guy who wrote the algos for the 480L? He's probably one of the guys who got laid off when Harmon closed the Lexicon R&D in MA and moved operations over to Salt Lake. It will reportedly sell for about $3k.

I've also heard that the PCM-91 has algos from the 480L, but perhaps there's more to it than that.

Yes and No.

Its an optional board that was purchased for it.


Again a 480L,EMT250,Roland R-880,AMS RMX16,Quantec QRS and XL were all designed as a system.



This means that all the parts were specifically chosen to work together.



This was before the computer technology became so cheap and accessible.


These were the flagships for these businesses.

When a studio purchased one of the systems it wasn't like picking upa $2K effect processor...we are talking about $10K(an EMT 250 was $15K).

So yeah you made sure it not only worked but it had an incredible sound.

And remember there was no DAW mixing.


Everything was mixed on analog(i miss those days).


These units were tested with analog consoles.


The analog sections for these converters were all designed in house to work with these units only.


So yeah they contributed alot to the sound.


Its not like today where the analog converters are an afterthought.

Whoa. This thread got a little heated while I was sleeping!

That Bricasti verb looks to be promising. I had seen a thread entitled "What's Up Bricasti?" or something like that. I thought it was a shoutout to someone. Had no idea it was a new reverb. Great thread title!

Nowadays when building a reverb, they just choose an ADA chipset off the shelf. Back in the day, they designed entire systems pretty much from the ground up. I've heard that TC went through pains with the converters on their 6000 system, but even that's not quite the same.

Unless someone builds a verb with converters that use an R2R resistor ladder network or another old technique, that old sound will remain elusive, IMO. [edit] And even then, unless they use old parts instead of modern high bandwidth components, it'll probably sound too sterile and bright, like most reverbs these days.

I'm not sure why it's so hard to believe that a 480L sounds different that a PCM90. The PCM90/91 is a single rack space unit built to a cost. The 480L is a no-holds-barred design from a different era. Back then true innovation was allowed! Now it's all about the company's bottom line.

Hoping that Bricasti verb rocks, but I wouldn't expect it to sound just like a 480L either.