The neverending story of the JRC4558D!

[R.G.]

I'm sorry I am not an electrical engineer. As an audio restoration engineer I restore old recordings. So I am not able to do extensive testing on the chips. But it would probably not show any signifficant difference from a technical standpoint. There is written a lot about that already.

Some people say you can't hear it, and some people say they can.

More to the point, I have seen and read about blind tests where people who, before the test, swore that they could hear subtle differences in this or that. After a blind test where they could not do better than random guessing about which was what, let alone which was better and how, they STILL maintained that they could hear the tiny differences that no one else could, in spite of the test results. They commonly held that the test setup was defective or that they were tricked or lied to about the results. The end result in the hifi field is that the Golden-Ear Subjectivist will NEVER submit to a test where they could be shown not to have the vast powers of perception they had convinced themselves they possessed. Some will even rant about the differences being too subtle to test. Once someone honestly believes they can hear delicate differences in this, that, or the other, it is nearly impossible to convince them otherwise. One really nasty thing to do to someone who is wholeheartedly convinced of their powers of discrimination would be to feed them the *same* audio sections, separated by a mild click and interspersed with one notably distorted "bad" control sample, but telling them that there are three different signals. I suspect they would almost always pick and choose between the identical ones as different. They'll almost never say they're identical. Just a guess.

People can perform amazing feats of digging signal patterns out of incomplete and inconsistent sensory data. They can and do also perceive subtle patterns in noise where there is no pattern.

So if someone tells you they hear something, you must believe them. It's real to them, whether anyone else can hear it or not.

[Der Groovemeister]

I lost a substantial amount of sleep over this.
So what about nos tubes then? Or nos caps? Also bogus? What about if in my work, restoring old recordings, for example i hear a 50 Hz hum (in the Netherlands we have 50 Hz cycle current as opposed to your 60 Hz) do i really hear that hum or do i just think i hear it? Because i expect it to be there?  Or tape noise or vinyl clicks? That would make my job a lot easier!

[anchovie]

What about if in my work, restoring old recordings, for example i hear a 50 Hz hum (in the Netherlands we have 50 Hz cycle current as opposed to your 60 Hz) do i really hear that hum or do i just think i hear it? Because i expect it to be there?  Or tape noise or vinyl clicks? That would make my job a lot easier!

I'm guessing that you digitize the old recordings and clean them up in software. Spectral analysis and the visual representation of the audio will show you the presence of unwanted content, so you have confirmation there.

You don't need to do any electrical testing of the chips, but perhaps you could play the same recording of clean guitar through the pedal and perform a spectral analysis to see what differences might exist. The only variable would be the IC.

[aflynt]

A metaphysics discussion centering around the JRC4558D. Awesome!

-Aaron

[roseblood11]

Could the age itself play a role here? The mentioned IC´s have been used for decades, but do they age? Is it possible that certain parameters change their values?

[R.G.]

I lost a substantial amount of sleep over this.

There is some initial discomfort over the concept. That should pass as you integrate the information.

So what about nos tubes then? Or nos caps? Also bogus?

It depends. Does "NOS" mean "New Old Stock" or "Not Old Stock"?    Seriously, think about what you get if you get "NOS" tubes or caps or whatever. Even if you really do get something that was put on a shelf decades ago, and waited for you to come enjoy its sonic perfection, you have to wonder:
- If it was so good, how come it hasn't been used by now?
- WAS it ever good? How good? In what way? Oldness itself is not a virtue, even in single malt whiskys and wine. The basic stuff has to not only be good, but something that improves with age, or at least doesn't get worse. Aging electrolytic caps is a sure way to make them *worse*.
- there had to be some real difference between the modern replacement and the NOS part in the first place, and it had to be in the correct direction. And in my opinion, that something has to be at least possible to measure. A meter has no way to tell you about stuff that is NOT there. It may not see everything that is there and real, but it will never make stuff up that isn't there. (*Note to the proofreaders: yes, I know there is a whole class of meter errors and subtle problems reading them. I'm illustrating a point here.*)

Some NOS parts are fakes. Some never were good. All are sold on the same kind of promises that get ladies to buy new cosmetics.

What about if in my work, restoring old recordings, for example i hear a 50 Hz hum (in the Netherlands we have 50 Hz cycle current as opposed to your 60 Hz) do i really hear that hum or do i just think i hear it? Because i expect it to be there?

If you hear it, it may be there. Run the recording through one of the very widely available programs which will show you the spectrum. If there's a peak at 50Hz, then your ears told you the truth. The meter/program can't expect it to be there.

There is a widely recognized issue with some recordings being equalized to sound very good with the specific studio monitor speakers. What it sounds like with other speakers gets a little foggy.

Or tape noise or vinyl clicks? That would make my job a lot easier!

Trust, but verify. Think you hear tape noise? Go look. Clicks are easy. Is there a blip in the wave file or not? Use your ears for what they're really great for - finding suspicious places. Then use the tools to help you verify and dig out the truth. Note that your ears are not, and can not be calibrated. What does a head cold do to your hearing? A loud concert the night before? What does that do to your audio restoration?  If you're not using some kind of tools, but just making it sound good, you're painting over the original. It may sound good, but it's a pastiche of old and new.

Could the age itself play a role here? The mentioned IC´s have been used for decades, but do they age? Is it possible that certain parameters change their values?

If either stored or used within their proper operating conditions, no, they will not age in your lifetime, or your children's. There is an aging process in diffused silicon that over centuries and millenia will "relax" the diffusion and the thing will quit working like it did. But this is a truly long time scale. If the ICs in question have been used in abusive manner, they will degrade: See "When Good Opamps Go Bad" at GEOFEX. However, the abusive aging is like arthritis or failing eyesight, not like a single malt whiskey maturing.

I kind of wish you hadn't brought that up. I've always NOT said stuff about this, because now the hucksters will start advertising "pre-aged vintage JRC4558Ds, carefully broken in on aging setups using only tropical fish caps, carbon comp resistors and oxygen free electrolytic copper, fed electrons from batteries over 10 years old. Only electrons were used to power the circuits, no other charged particles contaminated its power feed during the aging process. Packed in conductive foam laid back during the 1968 to 1978 period; special requests for packing foam will incur an extra charge."

[Joe Hart]

So if someone tells you they hear something, you must believe them. It's real to them, whether anyone else can hear it or not.

I agree. Also, I have had guitar necks made using CNC machines and with everything being "identical" and sometimes they feel vastly different. I think people can observe (feel, hear, whatever) very tiny differences. Also, years ago I tried out a handful of different IC's and heard differences between a lot of them. They were 4558's, 4559's, TL072's, TL082's, etc. and I did not research anything beforehand. I had no expectations on how they would sound, but after the fact, I did do some research and my findings did coincide with what others have observed. For whatever that's worth!

But I think the real bottom line is that you can "feel" the differences more than hear them (much of the time), and that is pretty hard to back up with scientific data. And if you "feel" a difference and it makes you play a certain way, that's all that really matters. If you think wearing your "lucky socks" makes you play awesome, then rock on with your socks on!
-Joe Hart

[Joe Hart]

Could the age itself play a role here? The mentioned IC´s have been used for decades, but do they age? Is it possible that certain parameters change their values?

Age itself, or different materials used in the manufacturing process? I mean, is every step of the manufacturing process of IC's exactly identical today as it was 30+ years ago?
-Joe Hart

[Paul Marossy]

Some people say you can't hear it, and some people say they can.

The main difference I've come across between vintage components and their more modern counterparts seems to be consistency!

I agree, but more so in transistors than in opamps. At least that has been MY experience.

[R.G.]

Age itself, or different materials used in the manufacturing process? I mean, is every step of the manufacturing process of IC's exactly identical today as it was 30+ years ago?

**NO** part of the IC manufacturing process is the same as it was 30+ years ago. A semiconductor fab is too expensive to hold static. You improve materials, processes and process control as a continuous business process or the next hot-shot will outrun you. Slow down on improving your semiconductor fabs and you die.

It's like learning to start a fire without matches, skin an animal for food and clothing, and make improvised shelter; it's hard to find the information on how to do things that crudely. In semiconductors, it takes extra effort to make the same crudenesses, the same process mistakes, and so on that marked chips from decades ago. It's impractical to make them that badly. Not impossible, just impractical.

[Galego]

Age itself, or different materials used in the manufacturing process? I mean, is every step of the manufacturing process of IC's exactly identical today as it was 30+ years ago?

**NO** part of the IC manufacturing process is the same as it was 30+ years ago. A semiconductor fab is too expensive to hold static. You improve materials, processes and process control as a continuous business process or the next hot-shot will outrun you. Slow down on improving your semiconductor fabs and you die.

It's like learning to start a fire without matches, skin an animal for food and clothing, and make improvised shelter; it's hard to find the information on how to do things that crudely. In semiconductors, it takes extra effort to make the same crudenesses, the same process mistakes, and so on that marked chips from decades ago. It's impractical to make them that badly. Not impossible, just impractical.

But, wouldn't that also mean that they shouldn't be making 4558's today? Aren't there much "better" opamp alternatives today?
There's a video on youtube of a shootout between several chips on a TS, and i think the guy asked someone from the NJM factory and they said they were still made exactly the same way.

[wavley]

I just always think of this... http://www.rane.com/pi14.html  and I'm glad that a google search turned it up and I didn't have to go find my old photocopy of a photocopy.

[Electron Tornado]

But, wouldn't that also mean that they shouldn't be making 4558's today? Aren't there much "better" opamp alternatives today?
There's a video on youtube of a shootout between several chips on a TS, and i think the guy asked someone from the NJM factory and they said they were still made exactly the same way.

"Better" op amps? But how are they better? If a particular part meets the desired specs, provides the desired performance in the intended application, each part will reliably meet those specs and continue to perform over time and under a wide range of conditions, it's available and cheaper than others, well....why change?

Another way to answer the question is that the 4558 continues to be made for some of the same reasons we still have airplanes that use propellers and piston engines.    

[R.G.]

But, wouldn't that also mean that they shouldn't be making 4558's today? Aren't there much "better" opamp alternatives today?

I don't understand your reasoning. At the bottom of this is the very real question: What is a JRC4558? Really?

The only possible answer is "it's a semiconductor chip which has the pinout and electrical specifications as listed in the datasheet". That datasheet does not list things which are attributed to the purported magic of the tube screamer opamp. All of the things that are supposed to make it magic for screamers are NOT listed.

It is entirely possible to make chips which more than meet the specifications in the datasheet on different/modified processes.

Are "better" opamps available today? Yep, absolutely, just as soon as you write down what "better" means. Faster? Yep. More voltage? Yep. etc. etc. etc. Howerver "same as the original JRC4558s" is not something you can test to (except by listening in a properly constructed blind test) or write a datasheet to.

There's a video on youtube of a shootout between several chips on a TS, and i think the guy asked someone from the NJM factory and they said they were still made exactly the same way.

I was there at the shooting of that video, by the way. 

[Der Groovemeister]

OK, I think it is time now to hook up this great sounding overdrive pedal between my guitar and amp, krank it up and PLAY!

[Mark Hammer]

1) I salute Groovemeister for his audio restoration skills and efforts.  A noble calling if you ask me.

2) One of the elephants in the room is that any device that has been installed via soldering, then unsoldered, and re-installed somewhere, is NOT the same (or at least potentially not the same) as a device that is fresh off the line and never been soldered.  (And please don't start any folk-legends about op-amps being improved by overheating!   )

3) One of the most common sources of misattribution in these sorts of comparisons is that people will find one or two examples of somethng, compare it to one or two examples of another thing, and then mistakenly assume that variation between the two things they have examples of is greater than the variation within each type of thing.  In other words, Groovemeister may have heard precisely what he believes he heard, but the device he heard it from did not reflect the full range of variation in devices that share the same nominal characteristic.  The analogy I like to use is that of thinking you know what the difference is between men and women by comparing a male hairdresser with a female Olympic weightlifter.  They are different, certainly, but do they each fully reflect the entire range of their sex?  Would you be accurate in saying that the woman has a thicker neck than the man?  Probably.  Would you be accurate in generalizing from those two to all men and women?

This latter dilemma faces us all the time.  It is rare that anyone has either the opportunity or takes the time (should the opportunity present) to try out many many instances of the Thing A they wish to compare to the Thing B, whether it is chips, pedals, pickups, or whatever.  That's not a personal failing, but I think it IS the basis of a lot of conjecture that, in the end, results in unnecessarily heated debates and unproductive rumour.

[FlyingZ]

It's funny, I much prefer the older Malaysia RC4558P chips in Orange Squeezers but in a Tube Screamer the current production chip is better then JRC4558D old or new.

[digi2t]

OK, I think it is time now to hook up this great sounding overdrive pedal between my guitar and amp, krank it up and PLAY!

Amen to THAT!!! I still can't figure out the difference between Coke and Pepsi!! Old, new, borrowed, or blue, just go with what sound good to you. I have an old Korg SDD2000 board, which had 5 JRC4558 chips on it. One "DD", and 4 "DV"'s. I've swapped them all around in different pedals, and I never really could tell any difference in the sound between them and newer 4558 chips. Then again, playing guitar at insane sound levels for years has effectively left my kids repeating themselves often at times, so I doubt I would pick up on the subtleties 

Cheers,
Dino

[FlyingZ]

It's extremely common to compare the wrong chips; you need the Malaysia stamped chips only. You still might not hear a difference.

[brett]

Hi
As Mark notes, comparing one chip with another seems to have legitimacy in this forum, but nobody would seriously compare one ford with one mazda and try to conclude which makes the "best" cars.  Or one old ford with one new ford.  Yep, it makes no sense at all.

Analysing the evidence :
1. In anecdotes, old is often said to be better than new. 
2. In unbiased tests, the differences are often insignificant (ie either absent, or present but undectable, or detectable but inconsistent).
3. The inconsistency between 1 and 2 shows (irrefutably) that the differences are due to the test method (unbiased vs anecdote).
So, in the case of old vs new, anecdotes are strongly biased towards old.  They offer a very poor representation of reality (but are widely believed for some reason - which I won't explore).

Of course, those who believe in mojo will claim that I have made and error or that they have been tricked, or that there's a whole subtle "thingy-ness" involving Kirilian maths and parallel universes that I am too dumb to understand.  That's ok.  Nobody said this was going to be a reasonable discussion.
cheers