The best sounding audio integrated opamps

Music: XTC, Calexico, Gillian Welch, EELS, Leonard Cohen, Nick Cave, Tom Waits, etc.

System: Cambridge Audio CD transport > CS4398-based DAC (here is where I've been putting the opamps) > Pioneer integrated amp > Polk Audio bookshelf speakers. Cables are all good quality.


I tried the OPA827 too - too bland and lifeless. The same is true about the LME49722, and others (LM6172 above all). The LM4562 is decent but the LME49723 sounds better. The OPA132UA is the best FET opamp that I know of.


I still want to try (in the DAC I mentioned) the THS4051 and the OPA1611.


The LT1028CS8 is what I use now and it's excellent..
 
Last edited:
I said above about the OPA827... it might be more interesting for I/V conversion.


Yup, the OPA2134 is a fine sounding opamp. Nevertheless, my favorite dual is the newer (and better?) LME49723.


AD opamps are my weak side, since those nice folks don't want to send samples to me. :mad: :p

However, my finding so far with all the AD opamps I've heard would be precisely summarised with "nice, but flawed". The AD845KN is one AD opamp that I like...

The AD797 is quite nice except that the mids aren't quite fluid and grainless.

The ADA4898-1 is interesting but I haven't managed to try it yet. The AD8599 was recently added in that big "opamp measurement shootout" and its distortion was disappointingly high (significantly worse than an OPA2132 for instance).


BTW the LT1028 absolutely blows the OPA827 out of the water for sound quality in simple DAC buffering.
 
Last edited:
Ah, I wanted to mention that I've also tried a few discrete opamps: OPA-Earth, OPA-Sun and OPA-Sun V2, from Audio-GD. The OPA-Earth was tonally dull and midbass lacking, the OPA-Sun was mechanic in the mids, the OPA-Sun V2 was the best sonically but produced a whopping 375 mV of DC offset where all of the chip opamps produced none.
 
I've tried everything or nearly so, and these are the absolute best sounding (voltage feedback) opamps for audio, for my hearing...:

LT1028, OPA132UA, LME49723, LME49720HA


A slightly special mention to the OPA211 too, which isn't bad.
Let me fill in here, if you want to have an opinion here we have two different answers.

1 Best opamp in your particular design.

2 Best opamp on a 100% perfect pcb with everything tuned in for the opamp in mind.

My answer to your question is that there is no answer because the opamp in mind will probably not work optimally.

Is the opamp for a vinyl amp, active filter, tone control, buffer or what? You must choose the opamp with taste and design a circuit with taste also.
 
peranders, whether the circuit is fully optimised or not, every opamp has (and keeps) its own unique sonic flavor. Those I have listed are my favorite flavors, i.e. those which I find to make me enjoy my music the most.


adason, I don't like tubes and I find they even less desirable than opamps. Have you ever tried, say, an LT1028? Things can hardly get more transparent...

Something similar for the LME49720HA, i.e. the TO99 package of it which sounds a lot more interesting than the plastic package. In plastic package I prefer the LME49723 to the LME49720.
 
peranders, whether the circuit is fully optimised or not, every opamp has (and keeps) its own unique sonic flavor. Those I have listed are my favorite flavors, i.e. those which I find to make me enjoy my music the most.


adason, I don't like tubes and I find they even less desirable than opamps. Have you ever tried, say, an LT1028? Things can hardly get more transparent...

Something similar for the LME49720HA, i.e. the TO99 package of it which sounds a lot more interesting than the plastic package. In plastic package I prefer the LME49723 to the LME49720.


have you tried the LM4562 to compare. I've read that its even better in the metal can (LM4562HA/NOPB).
 
peranders, whether the circuit is fully optimised or not, every opamp has (and keeps) its own unique sonic flavor.

Yes, but for example JFET opamps suffers from increased distortion when the inputs see different impedances so if you use a JFET opamp in for example a active filter, you will get more distortion than from a BJT model, unless you are aware about this compensate. Walt Jung has written a good piece about this.
 
AX tech editor
Joined 2002
Paid Member
Yes, but for example JFET opamps suffers from increased distortion when the inputs see different impedances so if you use a JFET opamp in for example a active filter, you will get more distortion than from a BJT model, unless you are aware about this compensate. Walt Jung has written a good piece about this.

It also makes a lot of difference whether you have an inverting (virtual earth) or non-inverting circuit. Some opamps do well in inverting circuits but when faced with higher common mode input in an non-inverting circuit they do much worse. You can't generalize without knowing the application. I mean, you can, but that then means that someone else can have completely different results than you because it is another app.

jd
 
It also makes a lot of difference whether you have an inverting (virtual earth) or non-inverting circuit. Some opamps do well in inverting circuits but when faced with higher common mode input in an non-inverting circuit they do much worse. You can't generalize without knowing the application. I mean, you can, but that then means that someone else can have completely different results than you because it is another app.

jd
I'm aware of these facts, the way FET opamps distort more with higher input impedances, and the other problem of opamps behaving differently in inverting and noninverting mode.

However, for the first, a V-out DAC output buffer application makes this problem nonexistent; as to the second, for me an audio opamp has to sound good mostly in noninverting mode (such as in my DAC), or it's worth looking elsewhere...



Oh, at any rate, I wanted to mention that I find the LME49723 to be more transparent (tonally as well) than the 2x OPA132UA, and more dynamic, notwithstanding its low cost. So... I guess that the good bipolar opamps always tend to outperform the good FET opamps for me and in an even situation (as above).


I have used the LME4562NA but recently I have found the LME49723MA to sound significantly better... I would like to try the LM4562HA too.
 
AX tech editor
Joined 2002
Paid Member
For an analog preamplifier there are some slighty better ones, with lower noise voltage that 1.2 uV (NJM2068 for exemple). I/V stagewoluld be a better application for it since it has slew rate of 20V/us and 1us settling time.

Bafta.

Which one is that that has slew rate of 20V/us and 1us settling time?
Are you saying that this sounds well in an I/V application?

jd
 
Well, you have to, unless you are listening only to the forte passages of a song or to music that is very compressed in dynamics. In other words if you don't need all the 16-24 bits recorded...
PS: If noise is not important, what's next? Distorssion, slew rate, unitary gain freq? A LM741 will suffice, I guess... :p
 
Last edited: