^ All true
I lost track of when RN left and it became AMS-Neve. I do know for a fact that he was a consultant during the design of the 5534 and 5532, and that there is plenty of evidence (as shown by G Tanner in above link) to suggest that trained engineers from the likes of the BBC were just as happy, if not happier, with 5534s in their gear than anything that had come before.
My point isn't to get into Neve history (as I said before, there are plenty of people who worked at the firm who can write an accurate account), it's more to combat any dogma that you often see regarding discrete circuits having advantages over ICs. Even the firm who made the most famous discrete designs of all was quick to embrace ICs. I don't believe there is any evidence to prove discrete better, and I regard companies such as Burson as taking advantage of 'low information voters'.
NB - the AMS-Neve resissues didn't come out until numerous boutique firms such as BA had started to release clones, as well as rack issues of OG modules. AMS-Neve themselves were pretty slow on the uptake. Once the reissues were released, they have since done everything to consolidate their market position, including trademarking 'Marinair' (a firm that disappeared in the 70s I believe) and stamping on the transformers. If you want to give credit to the firm for using 2N3055s, just remember that they are driving these trademarked transformers, which could be made in China for all we know...(they were made by Carnhill in the UK, but now??)
Discrete is necessary if you can't find an IC to do the job in hand. It can be incredibly quiet (if you are prepared to parallel a lot of devices and go to elaborate lengths). It has its place. If anyone believes it to be inherently superior, then I suggest they stockpile 2n3055s and trademarked transformers.
I lost track of when RN left and it became AMS-Neve. I do know for a fact that he was a consultant during the design of the 5534 and 5532, and that there is plenty of evidence (as shown by G Tanner in above link) to suggest that trained engineers from the likes of the BBC were just as happy, if not happier, with 5534s in their gear than anything that had come before.
My point isn't to get into Neve history (as I said before, there are plenty of people who worked at the firm who can write an accurate account), it's more to combat any dogma that you often see regarding discrete circuits having advantages over ICs. Even the firm who made the most famous discrete designs of all was quick to embrace ICs. I don't believe there is any evidence to prove discrete better, and I regard companies such as Burson as taking advantage of 'low information voters'.
NB - the AMS-Neve resissues didn't come out until numerous boutique firms such as BA had started to release clones, as well as rack issues of OG modules. AMS-Neve themselves were pretty slow on the uptake. Once the reissues were released, they have since done everything to consolidate their market position, including trademarking 'Marinair' (a firm that disappeared in the 70s I believe) and stamping on the transformers. If you want to give credit to the firm for using 2N3055s, just remember that they are driving these trademarked transformers, which could be made in China for all we know...(they were made by Carnhill in the UK, but now??)
Discrete is necessary if you can't find an IC to do the job in hand. It can be incredibly quiet (if you are prepared to parallel a lot of devices and go to elaborate lengths). It has its place. If anyone believes it to be inherently superior, then I suggest they stockpile 2n3055s and trademarked transformers.
I agree. There is no 'better' to me, just different sound.
Neve was amused to see how the vintage units are revered. He wasn't into nostalgia at all and pursued a goal of 'transparent sound'. I wonder if he didn't used discrete aop for some RND consoles ( because of high voltage rails, like SPL did).
But standard products were full of 'regular' opamps without detriment to the sound.
Neve was amused to see how the vintage units are revered. He wasn't into nostalgia at all and pursued a goal of 'transparent sound'. I wonder if he didn't used discrete aop for some RND consoles ( because of high voltage rails, like SPL did).
But standard products were full of 'regular' opamps without detriment to the sound.
Last edited:
Hey Anthoney,
Curious about your opamp configuration as I’ve experienced some scratchyness ( is that even a word, haha) in sound quality. IIRC you used the same adapter (Sparkfun) for an soic opamp. I couldn’t find an lm6171 in a dip package at the time. Of course it’s available now so I thought I’d give it a go except that does mean de soldering the sockets from the dac board which I’m a bit leary about. To easy to mess that pcb up. I’ll just have to cut the existing sockets out and just de solder the pins. Then I’m held to opamps in dip package only, hmmm
Any suggestions?
Thanks
Hope you’re doing ok
With regards to the extensive discussion of Neve console circuitry, most of the Neve flavor comes from the interstage coupling, input and output ransformers, as well as the EQ topology.
The Neve channel strip EQs had exceptional headroom and really could make the input source sound different in tonal balance without adding typical bad EQ artifacts or distortions. Its a bit like hearing the difference between higher end condenser mics - each has its own built in EQ curve but can't just be replicated by applying post EQ, otherwise it would be pointless to spend the money to buy some of these mics in question.
MCcartney's Tug of War album was recorded on Neve consol at Air studio in Monserrat and is a great example of this mentioned EQ magic I'm talking about. Its a very transparent recording with heavy EQ used everywhere. On a great audio system you don't really perceive how much EQ was actually used because it still has that transparency and smoothness despite the heavy EQ. Thats the Neve magic and why Martin wanted his consoles made to his specs using the older Neve design. Even the fully digital recorded Dire Straits Brothers in Arms album sounded quite palatable even with tons of top end sizzle cranked in to highlight all the subtle guutar and drum work. That album is very hard to listen to an a bad system, but can sound great on a really good signal chain. If they had used a different recording signal chain on that record, it would have never sounded that good with that much EQ applied.
The Neve channel strip EQs had exceptional headroom and really could make the input source sound different in tonal balance without adding typical bad EQ artifacts or distortions. Its a bit like hearing the difference between higher end condenser mics - each has its own built in EQ curve but can't just be replicated by applying post EQ, otherwise it would be pointless to spend the money to buy some of these mics in question.
MCcartney's Tug of War album was recorded on Neve consol at Air studio in Monserrat and is a great example of this mentioned EQ magic I'm talking about. Its a very transparent recording with heavy EQ used everywhere. On a great audio system you don't really perceive how much EQ was actually used because it still has that transparency and smoothness despite the heavy EQ. Thats the Neve magic and why Martin wanted his consoles made to his specs using the older Neve design. Even the fully digital recorded Dire Straits Brothers in Arms album sounded quite palatable even with tons of top end sizzle cranked in to highlight all the subtle guutar and drum work. That album is very hard to listen to an a bad system, but can sound great on a really good signal chain. If they had used a different recording signal chain on that record, it would have never sounded that good with that much EQ applied.
Hey sir, sorry for my late reply. I moved last month into a century home that's been vacant for a few years so lots of home renovation work to be done.
Have to apologize, offhand I do not recall the suffix. If I understand correctly, the different suffixes for the opa's usually pertain to specs like operating temperature and voltage offsets (or in other words you'd probably be fine with any suffix)
hope that helps
I agree. When people started using op amps in professional consoles, it became harder to differentiate a console from competitors just by the sound quality of the signal path. EQ has been the most important selling point for everyone, but in addition to that, Neve used transformer trying to retain the original "Neve" signature while SSL kept gaining a popularity with DBX VCA's coloration.
Hi Plasnu,
Well i disagree in op amp desks sounding the same: to had taking care of multipleAms/ Neve serie V( r and x) and SSL 4000 G and E they do sound different:
The SSL are more 'midrangey' and metallic than the V which have more lowend and are more mellower sounding overall. It probably comes down to mic input transformers... i'm not really sure.
I think the real main difference soundwize is into the eq: Ams/Neve's being constant Q ( as well as the G serie eq) and E are constant gain design.
This really does change the feel of it.
Ams/Neve used Vca too ( into compressor/limiter section). Not within the chanel path though, you are right.
Well i disagree in op amp desks sounding the same: to had taking care of multipleAms/ Neve serie V( r and x) and SSL 4000 G and E they do sound different:
The SSL are more 'midrangey' and metallic than the V which have more lowend and are more mellower sounding overall. It probably comes down to mic input transformers... i'm not really sure.
I think the real main difference soundwize is into the eq: Ams/Neve's being constant Q ( as well as the G serie eq) and E are constant gain design.
This really does change the feel of it.
Ams/Neve used Vca too ( into compressor/limiter section). Not within the chanel path though, you are right.
krivium, yep, this is exactly what I meant to say. Without some kind of coloration devices (transformer, VCA), it had become harder to make a difference because the opamp circuit itself is very similar. It's true that Ultimation SSL still sounds like SSL even without VCA, so I guess there should be some secret... I don't really know about Neve VR, but they are also built around 5534, right?
You know right away that it's not a fake. You can immediately hear the difference compared to the typical plane TI NE5532. There’s no trace of 'digititis' (the harshness or unnaturalness sometimes associated with digital sound), and there’s no fatigue even after long listening sessions. It brings to mind Muses, Sparkos, and Burson, but I’d say all of them resemble the original Signetics NE5532 more than the other way around.
This op-amp reproduces the sound of each instrument in a song with precision and clarity. You should be able to distinctly hear and differentiate the sounds of a guitar, piano, flute, drums, and so on—none overpowering or suppressing the others.
It feels like voodoo magic, and it gives me butterflies in my stomach. The warmth and vibrancy of the sound are instantly noticeable, especially when you test the Signetics version with vocal tracks. It brings voices to life in an incredible way, highlighting their warmth and richness. Percussions, shakers, trumpets, violins—all come through with a natural, warm, and well-balanced character from the first note.
I’ve tried this op-amp on a Whammy with headphones and on the Fosi V3 amp, and it outperforms all modern op-amps in every conceivable way. It’s a benchmark in sound quality, showcasing a liveliness and depth that modern designs often lack.
I will definitely be testing the mono/single version on the Miro DAC as well.😎
Yup. Signetics and raytheon 5532 5534 are the bees knees. The only ones i found superior in certain applications were 797 and 1651
Blind test would be blindly replacing everything with 4558's and then lying they were Signetic magic.
Then proceed to listen to all the romantic holographic spatial imaging nonsense stories as if a AI just recapped every audio myth
from the last 40 years.
I removed old production lines in the 90's and replaced with state of the art machines.
And seen armed guards for water and gas systems for the most advanced wafer production facilities.
Pretty much comical thinking 80's wafer production even touches modern
Then proceed to listen to all the romantic holographic spatial imaging nonsense stories as if a AI just recapped every audio myth
from the last 40 years.
I removed old production lines in the 90's and replaced with state of the art machines.
And seen armed guards for water and gas systems for the most advanced wafer production facilities.
Pretty much comical thinking 80's wafer production even touches modern
Sure, swapping everything out for 4558s and pretending they're the Signetics Mona Lisas of sound engineering sounds like a great plot for Silicon Valley: The Audiophile Edition. But let's not forget, while you're reminiscing about U.S.-made "superior processes," the chips in question were more about surviving the cassette tape hiss wars than summoning sonic angels.
By the 90s, wafer production wasn’t exactly romanticized, it was armed guards and industrial espionage. If someone had told those guards, “Hey, this is all for audiophile dreams,” I’m pretty sure they’d laugh while guarding a literal silicon gold mine.
As for moving production overseas, it’s not betrayal, it’s capitalism meeting cost efficiency. The Philippines and Malaysia didn’t get into the game to make the next audiophile utopia. They were just making chips people would plug into boom boxes or amps without needing to sign an NDA about the audio hologram secrets.
So, here’s to modern production lines making chips that probably outperform those 80s ones while costing less than a used LP. And if the audiophiles still hear magic, well, maybe that’s proof the real holographic spatial imaging was inside us all along.
By the 90s, wafer production wasn’t exactly romanticized, it was armed guards and industrial espionage. If someone had told those guards, “Hey, this is all for audiophile dreams,” I’m pretty sure they’d laugh while guarding a literal silicon gold mine.
As for moving production overseas, it’s not betrayal, it’s capitalism meeting cost efficiency. The Philippines and Malaysia didn’t get into the game to make the next audiophile utopia. They were just making chips people would plug into boom boxes or amps without needing to sign an NDA about the audio hologram secrets.
So, here’s to modern production lines making chips that probably outperform those 80s ones while costing less than a used LP. And if the audiophiles still hear magic, well, maybe that’s proof the real holographic spatial imaging was inside us all along.
IIRC different xx5532 datasheets would claim slightly different GBWPs, varying from 9MHz to 13MHz or so. So the requirements for stability may be sufficiently different, that the circuit and PCB layout that was fine for one, oscillates with the other.
If I had to guess, the internal compensation may be different, especially between the DIP and SMD versions, as SMD allows tighter component placement for decoupling caps.
If I had to guess, the internal compensation may be different, especially between the DIP and SMD versions, as SMD allows tighter component placement for decoupling caps.
I remember 45 years ago when Harris sold excellent but expensive opamps with decent PNP devices, back when everyone else had dreadful lateral PNPs. The NE5532/5534 was one of the first to get decent performance without ion implantation tricks by clever design.
I'm sure that even from the same vendor, detail performance must have varied between different fab processes as wafers got bigger
I'm sure that even from the same vendor, detail performance must have varied between different fab processes as wafers got bigger
Well yeah
Signetics is American and everybody knows the Sunnyvale location well.
Nobody invested in audio, it was funded by logic. Aka digital communications and computers.
But Fairchild and National already dominated the logic market.
TDA1034 was rehashed Philips stuff by then.
Not really Signetics by then just a name.
Their legacy was still logic, better known as missles
The guards wouldnt laugh they would just shoot.
Dont matter we cleaned out most that old junk and warehouses by 1989/90
Went to Chicago for scrap. Until 91 when India opened.
Whatever wasnt melted went on a ship
Pretty good legacy but 4562 at 55 MHz was around 2006 maybe 07
Plenty of non public stuff way above 100 MHz but voltage noise not really suitable for audio.
Better idea just slap in some 4562 blind tests
Whatever " magical" mids get past the CD players aliasing filters
Or countless junk anti skip chips
Signetics is American and everybody knows the Sunnyvale location well.
Nobody invested in audio, it was funded by logic. Aka digital communications and computers.
But Fairchild and National already dominated the logic market.
TDA1034 was rehashed Philips stuff by then.
Not really Signetics by then just a name.
Their legacy was still logic, better known as missles
The guards wouldnt laugh they would just shoot.
Dont matter we cleaned out most that old junk and warehouses by 1989/90
Went to Chicago for scrap. Until 91 when India opened.
Whatever wasnt melted went on a ship
Pretty good legacy but 4562 at 55 MHz was around 2006 maybe 07
Plenty of non public stuff way above 100 MHz but voltage noise not really suitable for audio.
Better idea just slap in some 4562 blind tests
Whatever " magical" mids get past the CD players aliasing filters
Or countless junk anti skip chips
Last edited:
Great example of why people without proper training and knowledge of perceptual testing should not be thinking they are qualified to do it.
Blind tests are supposed to measure something, and they are also supposed to have controls.
It appears what you want to measure is how persuasive you can be if you try to tell a convincing lie?
I'll be very open and honest about the whole op amp thing when it comes to no compromise audio. I try to avoid any of these crunchy little electronic spiders when I can, which means using discrete fet based stages wherever possible.
In typical production audio gear, where it can't easily be avoided using op amps, I opt for NE5532/34 for bipolars and AD8066/797 for fet input. Now obviously it matters alot how they're implemented as well as the supporting peripheral circuitry around them.
The signetics 5532/34 are unicorns, which means they don't sound at all the same as other 5532/34 based chips. They do pull more current than the other chips in the same family and can tolerate more than +/- 20V on the rails (sometimes up to +/- 22V under some circumstances). While that's not directly related to SQ, it allows for more output voltage swing depending on load.
The TI 5532s are decent all around substitutes but don't have the low end guts or top end finesse of the signetic chips. Alot of guys like the LM4562s, but I think they sound very midrange forward with lower impedance loads. When i say midrange forward, I mean it isn't a good kind of midrange like in some tube based stage. They also need more care when decoupling the rails as well as compensating for unity gain stability.
For fet based op amps, the AD8066 approaches discrete fet circuitry sonics. On paper it doesn't look as good, but listening to music there's a difference. The main drawback with these chips is the 12V rail limit and they also need very careful decoupling. Their slew rate is in the video range, far faster than anything commonly used for audio. They're also a rail to rail based design, which has its pluses and minuses. They only come in an SMD package. They sound fantastic though.
So to sum this up, I prefer the signetic 5532s over any other bipolar. The TL072s aren't terrible for cheap fet op amps. They just need alot of consideration when designing the surrounding circuitry. For CD player DAC I/V the signetic NE5534 is still the best, even when compared to most of the others. The only thing better is a transformer based output.
What's really weird is how little effect the built in antiparallel diodes in 5532/34s have on the audio. You'd think this would be far more detrimental to SQ in terms of generating distortion at the input.
In typical production audio gear, where it can't easily be avoided using op amps, I opt for NE5532/34 for bipolars and AD8066/797 for fet input. Now obviously it matters alot how they're implemented as well as the supporting peripheral circuitry around them.
The signetics 5532/34 are unicorns, which means they don't sound at all the same as other 5532/34 based chips. They do pull more current than the other chips in the same family and can tolerate more than +/- 20V on the rails (sometimes up to +/- 22V under some circumstances). While that's not directly related to SQ, it allows for more output voltage swing depending on load.
The TI 5532s are decent all around substitutes but don't have the low end guts or top end finesse of the signetic chips. Alot of guys like the LM4562s, but I think they sound very midrange forward with lower impedance loads. When i say midrange forward, I mean it isn't a good kind of midrange like in some tube based stage. They also need more care when decoupling the rails as well as compensating for unity gain stability.
For fet based op amps, the AD8066 approaches discrete fet circuitry sonics. On paper it doesn't look as good, but listening to music there's a difference. The main drawback with these chips is the 12V rail limit and they also need very careful decoupling. Their slew rate is in the video range, far faster than anything commonly used for audio. They're also a rail to rail based design, which has its pluses and minuses. They only come in an SMD package. They sound fantastic though.
So to sum this up, I prefer the signetic 5532s over any other bipolar. The TL072s aren't terrible for cheap fet op amps. They just need alot of consideration when designing the surrounding circuitry. For CD player DAC I/V the signetic NE5534 is still the best, even when compared to most of the others. The only thing better is a transformer based output.
What's really weird is how little effect the built in antiparallel diodes in 5532/34s have on the audio. You'd think this would be far more detrimental to SQ in terms of generating distortion at the input.
I first heard about this specific version from your posts; I didn’t know it existed.
Actually, I’m lying—I saw Signetics in a vintage modem, but I didn’t realize it was associated with high-end audio.
Thanks for shedding light on this! 😎