The Opamp thread

[Gustavo1976]

Great review. Would be good a comparison with sx35b or 45b.

 

[Androxylo]

I saw questions about MUSES03, and also comments in general that bothered me, so I decided to post. I last posted in this thread some years ago, so I might be repeating things in this post. But it's how I feel at the moment, so any repeated points just mean they're still valid to me.

To set the stage a bit, what I want from my sound system is involuntary emotional responses. This is generally produced by a combination of dynamics, speed, resolution, tone and general feel. The kinds of emotional reactions I'm looking for depends on the music and material. For example, Burson V6 Vivid are good at producing the feel of energy and power of a live event. For sheer scale and energy, it's one of the best. For example, V6 Vivid is good at making generic "epic" music sound as epic as the composer intended. I'd recommend them for video games. But for vocal focused music, they're mediocre at best, and this is where MUSES comes in. They have their quirks and flaws, which clearly make them polarizing. But MUSES01 has maybe the most beautiful sound I've ever heard from a chip, whereas MUSES03 is in my mind clearly up there with OPA828 as the best chips currently available. If I put OPA2156 in my amplifier I can listen to almost anything and experience an emotional flatline. I get a clean and clear sound, but I can feel my enthusiasm for music draining away every second. But if I install MUSES03 I'm suddenly grabbed and amazed by the power and presence projected by vocalists.

All my evaluation is done over speakers, as I'm swapping op amps in the input stage of a Purifi amp. It's one made by Audiophonics Workshop. It allows you to bypass the input section entirely, so I'm able to use my Phonitor X to drive the ouput stage directly and hear how much detail is removed by initiating the input stage with its op amps. All the op amps remove some detail, with MUSES03, OPA828 and AD797BRZ removing the least. I have two discrete op amps, SS3602 and V6 Vivid. They're close runner ups for the chips, but mess up the sound in different ways that I don't enjoy, so they mostly collect dust these days. I'm looking forward to Burson's V7 to see if they can win my nr 1 spot. The reason I use the input stage at all is because it fixes problems with dynamics and phase. With no input stage active the treble becomes too leaned back and flattened. The sound in general becomes so mellow it puts me to sleep.

The reason MUSES costs so much is because of their novel design. They are all two chip designs, as compared to other integrated chips. MUSES01 achieves -150db crosstalk thanks to this two chip design, whereas MUSES03 being a single channel op amps uses the two chips to separate input and output stages. OPA1612 "only" achieves -130db channel separation. How much does this matter for the audible sound quality? Probably not much, but it shows the MUSES op amps were designed with a high standard in mind, even if it doesn't show in many measured parameters. In the case of MUSES03 the two chip design creates a remarkably powerful output stage at 250mA, which makes all the competitors look weak as they usually only achieve 30mA-100mA.

In general, op amps eat detail, meaning that each one degrades the sound quality slightly. Newer op amps generally preserve detail better than old ones (like OPA828 vs AD797BRZ), and single op amps tend to preserve more detail than duals (like AD827SQ vs AD847SQ). Whether this is a problem or not depends on the price level of gear we're talking about, and I've recently found it also depends on the signal chain as a whole. I've read on DIYAudio forums that a reason dual op amps perform worse is because the crosstalk between channels isn't just electric, but also thermal. Sometimes the loss of using a dual instead of single op amp doesn't matter too much. I've found it makes a big difference if you have 3-6 op amps in a row, like say I/V, LPF and Buffer, or even an op amp based DC Servo in the signal chain. That's where you can gain an amazing amount of extra resolution and sound stage width by swapping duals to singles on adapters, or getting a bunch of discrete op amps. But what I've found is that sometimes the loss of detail can just be in the single % range, so having a dual is no big loss. It all depends on the signal chain as a whole whether it matters.

The question came up recently, why do we swap op amps at all? Why swap op amps when we can simply buy a properly designed device instead? Because that costs money! If we can look at higher end devices for inspiration, modify our own to copy aspects of their design, then doesn't that mean we got 90% of their sound for a fraction of the cost? Probably not, but it feels good when it sounds like we did! Esoteric D1X uses 16 MUSES03 op amps in the filter section. From the Esoteric website: "MUSES03 which was produced through many years of joint development with New Japan Radio Co., Ltd." How much of that $50,000 sound am I getting by blindly putting them in the input stage of a €1100 Class D amp? Probably not a lot. But it sounds good to me, so I'm happy.

Another example, I upgraded my Purifi amplifier with WBT Next Gen Silver speaker terminals, Neotech solid core UP-OCC wiring to the speaker terminals, and also Neotech solid core silver wiring from the XLR inputs to the input board. These are the kinds of upgrades you see in high end Purifi amp builds, but I could do this on my own for just the cost of parts and 4 hours of labor on a weekend. This made a big difference to the sound, so I'm very happy with the result. The chassi for the amplifier is also a bit cheap and had a ringing noise, until I lined it with Dynamat Xtreme. Now it's completely inert, possibly improving the sound like a higher end build would. Then I also put Pangea Sorbothane feet on it. As with op amps, I'm just bolting more expensive parts to this amp in the hope it'll make it sound better. Every tweak seems to work, even if some only change the sound in a miniscule way. The cabling and op amps made by far the biggest difference.

Anyway, back to MUSES03 vs other op amps. MUSES03 have a tone that's feels more or less neutral and correct, with a touch darker tone compared to OPA828 or AD797BRZ. It's only slight, but the difference is there. Their strength is the midrange, but I enjoy their character as a whole. Compared to discrete op amps, MUSES03 can sound startingly dynamic. I said V6 Vivid have a powerful sound, but they're somehow also sluggish and slow compared to chip amps. They have the power, but not the speed. MUSES03 can be much more dynamically varied and impactful than V6 Vivid in a lot of music. MUSES01 and MUSES02 have a much more colored sound than MUSES03, with MUSES01 sounding like everything has been pitched up, and MUSES02 sounding dark and boomy. OPA828 (and other modern OPA models) has a more neutral character, but in a more dull sense. If all you listen to is real life recordings and sound effects, then OPA828 is amazing. With something like the engine start-up sounds in Gran Turismo 7 it can almost like you're in the same room as the car. But for music, they're only satisfying with certain genres. With vocals they're an emotional flatline for me, but with progressive metal or electronica they can be excellent. They have a superb sense of speed and focused impact, so it can feel like musicians are playing faster because each note is being reproduced with more precision and "blackness between notes", or however you describe the effect. Depending on how you listen, what you listen to, they can be the holy grail. I vastly prefer them over OPA2156 and OPA1611, as they are too similar but worse to my ears. I haven't tried OPA1656 unfortunately, so I can't tell how it compares against it. But I've compared OPA828 against OPA627AU, BP and SM versions, and OPA828 blows every version of OPA627 out of the water. Every time I listen to OPA828 I admire them, but I always end up going back to MUSES03. Because in the end they give the most satisfying listening experience across the widest range of material. They're not as epic as V6 Vivid, or as precise as OPA828, but they're close enough. They have their own strengths that makes me come back to their style. But if I get tired of them, I can just spend 5 minutes swapping to a different op amp. I'm probably not hearing anywhere close to the full potential of any of these op amps, as the circuit wasn't strictly designed around them. But I'm still enjoying myself more with these than the mediocre stock op amps, so I'm happy playing amateur designer.

@TheManko super useful. I am thinking about rolling SS2590 into purifi, but it looks that's the one you don't have either. Any educated guess how it may go vs MUSE you like? Rolled nothing yet myself.

 

[TheManko]

I am thinking about rolling SS2590 into purifi, but it looks that's the one you don't have either. Any educated guess how it may go vs MUSE you like? Rolled nothing yet myself.

Yes, as you noted I've never heard the SS2590 myself, but I've read many posts about them and can sort of guess at how they would perform. From everything I've read, bigger is generally better when it comes to discrete op amps, so the SS2590 will have sonic advantages that other smaller op amps just can't match, because they were made with size constraints in mind. SS3602 is already quite good, as it makes my Purifi amplifier sound sort of like an NAD Class A/B amplifier. It's a sound with benign flaws across the whole frequency band, where you notice some added texture in the midband and some phase issues in the treble and bass. The general character is a bit leaned back, which at worst can feel a bit dull. It's an op amp that's easy to recommend because it will sound good with more or less any music you play, unlike V6 Vivid which goes for a more exciting and in your face presentation style. SS2590 improves on everything with a more powerful and exciting sound, but still retains a well judged relatively benign character. But unlike SS3602 it doesn't come across as too leaned back, because the SS2590 is able to do more of everything at the same time. From what I've understood, the bigger op amps like SS2590 are more or less universally praised as delivering sound quality a clear tier above everything else. Therefore if you're able to use them, the options are usually SS2590 or the equivalent Sonic Imagery op amp if you want a sharper sound.

The op amps I use the most these days are MUSES05 and V7 Vivid. I have swapped back and forth between the two dozens of times, and also compared them against MUSES02, MUSES03, OPA828 and OPA1611. MUSES05 and V7 Vivid are much better than everything else I've got. In one of my comparisons I first listened to MUSES02 for two days, then swapped to MUSES03 and was blown away by how much better MUSES03 sounds. It was like adding three extra layers of sound in the depth layer. I then swapped to MUSES05 and was again impressed by how obviously better the sound became, this time in sharpness of focus, the feeling of intensity and microdynamics. The sound also became more 3d in the sense that instead of having a few distinct layers like foreground, middle and background, it became more like a gradient with infinite layers. This is similar to how V7 Vivid sounds, except V7 Vivid expands the sound further in width than what MUSES05 is capable of. But in the depth plane they are similarly satisfying. I think these are the sort of qualities you'll also get from SS2590. From what I've read from posters earlier in the thread, V7 Vivid has many of the good qualities of SS2590, but is maybe not quite as full or powerful sounding as SS2590. But I'm just guessing and inferring from other people's accounts.

 

[Max61MI]

I fully agree with what TheManco said, and I can confirm that the SS2590 are excellent, probably the best I have ever tried, even if to perform at their best they need a long initial break-in period and should be inserted into equipment of an adequate level, to be able to express all their refined potential.

At the beginning I did not like them, starting with a sound too crowded in the mid-high range and that caused me listening fatigue, but then I re-evaluated them, noting that the sound then opens up a lot and becomes more "liquid", starting to flow pleasantly. I confirm that they have performances well beyond those of the SS6302, which in comparison show a slightly "smaller" sound, less expanded and majestic; in every sense, both in terms of bandwidth extension and width and depth of the soundstage.
The V7v have a sound with a slightly more "organic" texture in the mid-high range, which some people prefer with some musical genres, but in my opinion they remain with less bass presence, less extension in the highs and a less wide soundstage.

In addition to a well-present and clean low range and a wide musical scene, the main distinctive feature that I can say I have identified as a character in the SS2590 (and that for some could represent a limit) is to have a fluid, clean and very extended high range but at the same time less incisive in the mid-high range compared to other opamps.

In my opinion, this is a precious distinctive feature that is rare to find in other opamps, which usually tend to oscillate between a situation of less expansion, width and detail vs, on the contrary, a more extroverted and "on your face" sound; the SS2590 are neither of these two opposites, managing to reproduce a wide and natural sound but without ever exceeding in presence effect (clean, refined and never aggressive sound).

I therefore find that they are particularly recommended for "taming" headphones (or DAC or Amps) that tend to a particularly open sound in the medium-high range, smoothing out any roughness and increasing the sense of depth of the scene.

On the contrary, in the case of more "bassey" headphones (and devices) someone might instead feel the need for a greater impact (clarity, "density") in the medium-high range and therefore, only in these cases, it might be more appropriate to choose opamps with a slightly more aggressive character.

With an equally high level of quality, but with a slightly more incisive character, I found the Hypersonic (Virtual HiFi) excellent, but they have two problems; they consume a lot (to the point that some devices cannot power them) and become very hot, so they need adequate ventilation (definitely not good for closed boxes).

Always in case someone is looking for a slightly drier and "forward" sound I can recommend trying the SX35b, which in my opinion offer an excellent quality/price ratio. I have not tried the SX45b, but LoneFurrow says that they have more body in the mid-low range, and therefore I was not interested (whereas the SX35b have "revitalized" the already soft and full-bodied sound of an R2R).


Here are some impressions expressed by LoneFurrow:
https://www.head-fi.org/threads/xduoo-xd05pro.964620/page-228#post-18596708
I also attach a photo of my current chain: SX35b (first position) + SS2590 (second, as output buffer) in the mini R2R DAC, and Hypersonic on the xDuoo 05 Pro (used as an amplifier).

P.s.: Initially with the Denon D9200, which are very open and defined in the mid-high range, on the xDuoo 05 Pro I was using the SS2590 but recently instead, with the FiiO FT1 which have a "softer" caracter, I appreciated the pairing with the Hypersonics a little more.

 

[Androxylo]

Again, as I rolled nothing yet, I would make my educated guess that from all the reviews you need to consider if you have headphones, small or large speakers. With more powerful speakers as mine and quite a beefy amp, but a small room, the paramount feature is bass control, more than anything else. If the wall reflection comes after the direct sound ends, the brain will process it as "air" while if they overlap the brain will merge them into congested sound. If you don't have wall reflections, like in headphones, other features could be more important.

 

[Max61MI]

Again, as I rolled nothing yet, I would make my educated guess that from all the reviews you need to consider if you have headphones, small or large speakers. With more powerful speakers as mine and quite a beefy amp, but a small room, the paramount feature is bass control, more than anything else. If the wall reflection comes after the direct sound ends, the brain will process it as "air" while if they overlap the brain will merge them into congested sound. If you don't have wall reflections, like in headphones, other features could be more important.

In a case like yours then I could certainly advise you to try the SX35b that "dry" the bass thus improving the perception of the texture. They don't cost much and so it could be interesting to try them (always providing adequate running-in to open up the sound).

Although in those cases, acting to improve the acoustics of the room (in addition to choosing the most ideal speaker for that particular situation) probably remains the priority aspect.

 

[TheManko]

If the wall reflection comes after the direct sound ends, the brain will process it as "air" while if they overlap the brain will merge them into congested sound.

If you want to optimize the bass response I recommend downloading frequency test tones. I had to move my speakers closer to the side walls to get rid of a huge null at 80hz-90hz. Rear wall reflections can also cancel out bass deep bass, the frequency of the cancellation changes depending on speaker distance from the rear wall. The op amp can as you noted change the impression of the bass quite a bit, where MUSES02 and 03 give you a thicker bass that hangs around for longer, V7 Vivid is drier and lighter, and Sparkos could probably be described as neutral. What you hear is the result of your entire signal chain, so you can tune the sound by changing cables as well. To save your sanity it's probably best to limit the variables lol. Op amps have a big effect on midrange and treble as well, so if you can fit SS2590 I would get them, tune your speaker placement, and then hopefully be happy.

 

[Androxylo]

Finally tried ss2590. This is not subtle! Comparing with OPA1656 it's not just changing the sound quality, it's changing the sound.
My first reaction was funny, after 3 min of listening I've got scared that the bass is gone and I've got a system that is pushing the upper mids and up. I've use two tracks to verify that this is not the case - Sarabande (feat. Alem) MMMD that the bass can go very low and remain clean, and Longing Yosi Horikawa that the bass is very fast, it this track every hit on the drum is multiple fast reverberations.
My system is desktop and DIY speakers are actually almost touching the back wall, I am using Equalizer APO to make the desired curve, and I have a significant suppression applied in the 100-300 Hz area to compensate for the wall reflection. When the bass is faster, each note ends before the reflection comes 3 ms later. I will have to redo the EQ curves again for the new setup, the bass could be released back. Few posts above, this was exactly my goal.
Instrument separation: my dream was always to crack the Symphony No. 2 in C Minor "Resurrection" to sound as many separate instruments, because I always knew it's amazingly recorded but it always sounded just ok. Now I finally can hear more of instruments here and there. Or Il Trovatore: Act II: Scene 1: Vedi! finally every person in the choir is separate, they don't sound as a group with occasional separates.
Soundstage: Prende La Vela drums, singers, trumpets are spread all over the place, not only wide as before (this is unchanged) but also deeper.
Air and texture: Step Into Paradise even more airy, and the full illusion of the singer to be in front of me. It was great with 1656 but pushing even more.
Dynamics: Tricycle I think it was already good with 1656 so it just remains good. However, remember that I still have to give it more bas..
No particular character in those chips, just everything is better across the range.

 

[TheManko]

Finally tried ss2590. This is not subtle! Comparing with OPA1656 it's not just changing the sound quality, it's changing the sound.

Glad to hear it! You mentioned earlier you're swapping the op amps in a Purifi amp, just like I'm doing. I've noticed after testing many op amps that the Purifi amp module itself seems to have almost no sound character of itself. It just takes what its fed and makes it louder. The op amp feeding the Purifi amp therefore has a profound impact on the resulting sound.

 

[Androxylo]

FWIW SPL curve with everything except opamps being the same. Notice the whole area 300 - 3000 Hz is just flatter. The differences in tighter bass are not related to any change of the frequency response - the previous more sloppy and boomy bass was not actually louder.
Measured at the desktop table with PC monitor and almost touching the wall.

 

[TheManko]

Notice the whole area 300 - 3000 Hz is just flatter.

It's rare to see any kind of in room op amp measurements, nice to see the difference is this clear!

 

[trs1]

FWIW SPL curve with everything except opamps being the same.

Are you absolutely sure the volume was set to the same level? There's wide swaths of trace here that would be neck-and-neck (if not completely overlap) through a significant portion of the graph if that OPA1656 line were boosted 2dB-ish just straight across from 20 to 24k (or if the SS2590 were dropped by 2dB-ish straight across from 20 to 24k). From 20-400Hz, 600-1kHz, and then 3.6kHz out to the end, it really looks like they'd line up more or less bang-on if you just added 2 to every SPL value in the 1656 dataset. About the only place where there wouldn't be significant overlap is that midrange from 800-3.6kHz where the 1656's tendency to get loosey-goosey at full chat is well documented. I'm sorry, now that I see it I can't un-see it :/

I apologize if this all feels like a bit of an attack - it's not. I promise all of this comes with with great appreciation and respect for the time you've devoted to testing these ICs and sharing your findings with all of us, but... maybe double check your test setup and measurements and run it again? "Correlation ≠ Causation," or so goes the old saw, but that doesn't mean tightly correlated measurements like these aren't worth a second look - just to be absolutely sure, know what I mean?

Of course, those discrete op-amps draw twice the input power and are capable of over twice the output power of those 1656s, but I can't see how that might measurably affect the recorded sound pressure levels while also boosting the perceived fullness of sound equally at all frequencies across the audible spectrum... I'm gonna need to pray on this one for a little while, I think

 

[Androxylo]

Are you absolutely sure the volume was set to the same level? There's wide swaths of trace here that would be neck-and-neck (if not completely overlap) through a significant portion of the graph if that OPA1656 line were boosted 2dB-ish just straight across from 20 to 24k (or if the SS2590 were dropped by 2dB-ish straight across from 20 to 24k). From 20-400Hz, 600-1kHz, and then 3.6kHz out to the end, it really looks like they'd line up more or less bang-on if you just added 2 to every SPL value in the 1656 dataset. About the only place where there wouldn't be significant overlap is that midrange from 800-3.6kHz where the 1656's tendency to get loosey-goosey at full chat is well documented. I'm sorry, now that I see it I can't un-see it :/

I apologize if this all feels like a bit of an attack - it's not. I promise all of this comes with with great appreciation and respect for the time you've devoted to testing these ICs and sharing your findings with all of us, but... maybe double check your test setup and measurements and run it again? "Correlation ≠ Causation," or so goes the old saw, but that doesn't mean tightly correlated measurements like these aren't worth a second look - just to be absolutely sure, know what I mean?

Of course, those discrete op-amps draw twice the input power and are capable of over twice the output power of those 1656s, but I can't see how that might measurably affect the recorded sound pressure levels while also boosting the perceived fullness of sound equally at all frequencies across the audible spectrum... I'm gonna need to pray on this one for a little while, I think

No, I had to say that please ignore the volume. There are several days between those measurements so the volume was only set approximately the same. The placement was identical because I measured the exact spot for speakers and the mic. I apologize for the confusion. It's more about that different opamps do not bump any frequency area, only the curve looks smoother. Some of the waves became more flat.
I totally agree the average smoothed shape of the response must be identical in all opamps, and my chart shows this. Please mentally move the red curve 2 Db down and pay attention on the squiggles only.

 

[KaraChaffee]

So I've been spending some time with a new round of opamps:

ISL28127FB: Very low noise floor, a bit more than 4627-1. Extremely clean, smooth, and well nuanced with wide soundstage, yet stereo separation isn’t all that good, I’m not sure how to describe it. Very refined and neutral, though not as neutral as 4627-1B, it is pretty close. It is slightly laid-back with okay articulation and lacks a little timbral transparency, not sure. Slight hint of "yellowish-brown" in its tone.

AD8597AR: Very clean and smooth, kind of artificially analogue-sounding, crazy articulation. The midrange is simply thrilling at first, but then I get bored quickly of the sound. It tries too hard to sound exciting and over-articulates every sound. Bass is too forward.

OPA1611: Very warm and slightly energetic at first, but the color went away with burn-in, now it is more or less neutral. Very good overall with good detail and clarity and what not. Slightly lacked bass compared to the other opamps at first, but then got a bit bassier later. This one has a very wide soundstage like ADA4841-1. Articulation and timbral transparency is very good. I think it does still retain a touch of color, but I can't quite put my finger on it.

THS4081: Very neutral, clean, clear, very detailed, highly articulate without overdoing it like AD8597 and very accurate, kind of natural-sounding and smooth. Crystalline sound with weighty, wicked bass response and tightness. It brings out sounds towards the far left and right like 4627-1. Very beautiful sound overall. Heck, I don't hear a whole lot of difference between this one and 4627-1B. Dang, this one is just more detailed and a touch tighter. Great low bass. This is one of the best opamps I've ever heard.

THS4051: Very neutral, clean, clear, very detailed, highly articulate without overdoing it like AD8597 and very accurate, kind of natural-sounding and smooth. Crystalline sound with weighty, wicked bass response and tightness, but a bit less bassy than 4081. It brings out sounds towards the far left and right like 4627-1. It is like a slightly lesser 4081.

LT1115: Very clean, articulate, neutral. It is very similar to THS4081... Well good grief! This one also sounds a lot like THS4081 and ADA4627-1B. It doesn't seem to be quite as extended at the extremes, though. A slight difference. Overall a touch closer to THS4081 than ADA4627-1B. It didn't like my portable amp for some reason, but has wonderful timbral synergy with the Compass DAC.

ADA4004-1: Very clean, pristine, neutral. A bit bassy, though not as much as 797 or LT1028, and very intimate. Bass and lower mids are slightly smoother than the other opamps. Wonderful sound.

LT1677I: Very clean, pristine, neutral as well with good detail. Slightly lacks bass compared to the other opamps I've tried. Somewhat inarticulate. It sounds slightly greenish-white. It is slightly bright, probably from the slight lack of bass, though I think there may be a slight hump in the treble, around the beginning of the upper treble I think or a little lower, not sure. I'm really liking female vocals on this opamp over the others in this round.

OPA827: Almost neutral, very clean, clear, and fluid. Decent detail, it seems to have a very slight smiley curve to its sound. Reminds me somewhat of 4627-1B, but a pinch aggressive. Wonderful sound, slightly rich and a bit tuneful. Not quite as transparent as some of the others. Excellent articulation. This one is kind of nice with my SR325is, but the lack of resolution is a bit of a bottleneck for the Grado grunge.

I may add to this later.

Hi Max ,
I was very impressed with THS4081 at your suggestion . What a jewel ! Thank you! I am the retired designer for deHavilland and have been monkeying with OP AMPS for 20 years and I would like to tag my observations near your posting and I hope you don't mind . My application is in my own Metronome DAC as a buffer and in my Model 222 tape playback preamp, as a head amp ( gain of 4 to 8 typically) in front of the 12AX7s.

I prefer bipolar (BIP) generally over FET . I value transparency, balance , a warm sound, and delicate and sophisticated highs. I made an "A" list and "B" list and "C" list .

************** A LIST ****************
THS 4081 BIP . Transparent , full bodied , highs sing --voices float like angels up to the ceiling , really faultless. Has replaced AD 8597 in my mind. My personal choice. Tried THS4061 . Sounds rather gray compared to 4081 ....Don't bother.

OPA 827 FET . Dynamic , balanced , transparent , vivid without aggression. Best FET I know of --does what OPA627 does , but more and better. A wonderful device ,but does not sing on top like the THS 4081 . Not a sin of commission , but just not its nature.

ADA4004 BIP . Reminds me of THS4081 , perhaps a bit rounder and a bit slower. But a beautifully balanced and sweet musical amp. I could live with it .

AD8597 BIP Transparent and much like THS4081. I found it to be a bit light in the bass and always wanted more out of it , but it was my top choice. And I selected it over any FET but wanted more . And along came THS 4081.

AD744 FET Has been described as an "electrostatic sound " . I found it to be fast, transparent and sweet . But it put my attention on the highs and edges of the sound much like the 6922 tube . I prefer tubes like 6SN7 . But a very fine device !

************ B LIST ***************
OPA 627 FET Needs no introduction. Is on the B list only because OPA827 does what it does better with no downside.

AD4627-1 FET Nice amp . Maybe not as sophisticated as A list .

LME49990 BIP Very competent and transparent . I think its the best of the LME amps that I listened to. Somehow failed to grab me emotionally. Which is my general feeling about National Semiconductor amps ...
LME49710 , LME 49710HA . A bit bright especially the HA . Failing to grab me also applies.

************ C LIST ******************
BURSON V6 Classic . I tried the V6 briefly and the sound was good , however the pins on the Burson are approx. .025" in diameter where the industry standard for pins is in the range of .019" to .021" . They stretched and loosened my Mill Max sockets permanently and I may have to replace them . This was very disappointing and I sent the V6 units back.

OPA 602 / 604 FET . Big furry chocolate sound . My friend Sy , who hates anything even suggesting bright sound loves them . Easy to listen to , not terribly sophisticated .

OPA 134 FET Maybe the best old school BB below OPA627 . More accurate than OPA602/ 604
OPA 227 FET I don't remember , but probably like 134.

OPA 1611 BIP Failed to grab me . Not overtly bad .

OPA1655 FET Failed to grab me . Not overtly bad.

OPA891 BIP I had high expectations for this one after seeing the specs. but it turned out to be like LME49990. Left me cold especially in comparison to the THS 4081.

AD845 FET Nice chip , a bit crude .

ADA4898 BIP Very vivid and well balanced . I felt there was a bit of steelyness in the mids. Not a bad device.

LM6171 MC 1456 . Too many acid trips ago to recall them clearly , but I believe mechanical sounding and failure to grab.

LInear Technology LT1122 , LT1056 , LT1128 , LT1360 , LT1793 . As a group I thought some were very nice but at the end of the day , the HF character was always a bit unsophisticated.
Well, that is what is in my box of op amps , new and old.
Cheers.
Kara

 

[nemool]

Great review. Would be good a comparison with sx35b or 45b.