Discrete OpAmp Review: Sonic Imagery vs Sparkos

[Genpattonj]

I wish I had the kind of hearing that would allow me to also discern these differences...alas, I could hear no difference between chip op-amps and the free Burson V6's I was sent a while ago...they look cool, but nothing audible for me...

I wonder if it’s the equipment? I’m assuming every piece of equipment differs in the design. All I’ve done is do an amateur double blind test…and four people heard an improvement. I forgot to add we wrote notes down every time the music was played, it was by no means measured by an instrument. Four people came to the same conclusion, one a 20 year old woman with what I would assume has excellent hearing. I’m also a very sceptical person, however I believe there was a difference and it was a positive one. But…..from the pushback I’m getting the impression this is the “wrong” forum for that! ;-O

 

[SIY]

I wonder if it’s the equipment? I’m assuming every piece of equipment differs in the design. All I’ve done is do an amateur double blind test…and four people heard an improvement. I forgot to add we wrote notes down every time the music was played, it was by no means measured by an instrument. Four people came to the same conclusion, one a 20 year old woman with what I would assume has excellent hearing. I’m also a very sceptical person, however I believe there was a difference and it was a positive one. But…..from the pushback I’m getting the impression this is the “wrong” forum for that! ;-O

Extraordinary claims demand extraordinary proof- or at least tight controls. Your description was a classic case of the sort of lack of controls endemic to fashion audio. What you'll find here, though, is a great willingness to help you run better experiments. Besides the resources on this site (and there's plenty of them), here's an introduction to the basics of running valid listening tests that I wrote a while ago (which means that unfortunately it doesn't cover new and very helpful software ways of doing well-controlled evaluations). Nonetheless, I hope you find it useful and entertaining.

 

[Genpattonj]

Extraordinary claims demand extraordinary proof- or at least tight controls. Your description was a classic case of the sort of lack of controls endemic to fashion audio. What you'll find here, though, is a great willingness to help you run better experiments. Besides the resources on this site (and there's plenty of them), here's an introduction to the basics of running valid listening tests that I wrote a while ago (which means that unfortunately it doesn't cover new and very helpful software ways of doing well-controlled evaluations). Nonetheless, I hope you find it useful and entertaining.

Thanks so much! That is very much appreciated, and yes it was my first one and probably amateurish. But this is very helpful! Happy Holidays!

 

[Sam Spade]

It’s interesting that most people in this little area of the internet have…..well not quite a hate on but definitely a disdain for discrete OP Amps. I had no experience nor expectations until I purchased the same DAC my neighbour had, the AUNE X8 Magic DAC. I liked the size and decided I’d try it out to see how an inexpensive DAC sounded. I have several other DAC’s, one a Musician Pegasus R2R solid state and the other the Havana Balanced NOS DAC. I set up AUNE in my system and listened over several weeks. It wasn’t anything special, it was a DAC and it did it’s job. I’m not going to go into specifics as to how much better or poorer it sounded against my other DAC’s, that’s not relevant here.
I had heard a number of reviews where you can swap out the OP amp to the Sparkos SS3602. I ended up emailing Sparkos and the SS3602 was purchased. I replaced the default OP amp with the SS3602 and waited a couple weeks. I’m not a big believer in “burn in” but wanted to get a feel for how it sounded. I did notice after I replaced the OP amp the imaging was way more precise and instruments sounded more real. “I’ll stop the fluttering now!) okay. I don’t trust myself…when it comes to putting in new components..literally. There’s a huge “I just paid and put this gizmo in so it better do something in your mind”. Hey! My neighbour has the same DAC. I called him and we decided we’d do the following on the weekend : set up both DAC’s side by side, made sure they were covered up by a blanket, keep every other component in the chain the same. My Daughter will be switching them back and forth. One has the Sparkos the other has the regular OP amp.
We listened for over a half an hour but the difference was not subtle, it was obvious. Vocals sounded more lifelike, The music had more life to it. The imaging a person experienced opened up. It was the Sparkos. There a lot of descriptions I can use but I’ll boil it down; the Sparkos dramatically improved the sound. It was not just “different”, it was “better”. We listened again repeatedly switching for over another half a hour. Remember all we were doing was unplugging the digital out from my AUDIOLAB 6000 CDT. Then we decided to do another test. We switched the Sparkos into his identical DAC and tried the test again. Again, the results were obvious. One of my Daughters twenty year old friends had dropped by at one point and sat in on it and she listened and picked the Sparkos out as well.
Now this is by no means a “technical measurement”. One could argue many things about this amateur test. However just consider this, our ears have been used for literally many many thousands of years as hunter gatherers. Evolution has refined them to point out objects in your field of view and behind you as well.
Yes, we can measure the discreet Op amps like the Sparkos and I won’t argue that point. However I would encourage any of you to at least consider replacing a solid state OP amp with a discreet OP amp. Sparkos isn’t the only one. I prefer his because he’s a decent young guy who’s started his family business and had their first child. I don’t know him personally but have corresponded with him. I don’t think I’d get that kind of response and help from Texas Instruments! Again, I’m only saying what I experienced with his Sparkos SS3602, there are others. What I will say is from what I've experienced it was well worth the money. To recap: I’m not one for gimmicks, I’m well aware of all the “hype” and blatant misinformation out in the Audio world. This was simply an attempt to let you folks know what I and my Neighbour have experienced.
Happy Holidays to all!

Great post. Love the double blind test.

 

[MaxwellsEq]

I wonder if it’s the equipment? I’m assuming every piece of equipment differs in the design. All I’ve done is do an amateur double blind test…and four people heard an improvement. I forgot to add we wrote notes down every time the music was played, it was by no means measured by an instrument. Four people came to the same conclusion, one a 20 year old woman with what I would assume has excellent hearing. I’m also a very sceptical person, however I believe there was a difference and it was a positive one. But…..from the pushback I’m getting the impression this is the “wrong” forum for that! ;-O

Firstly, when you build a solution with op-amps, you must design the compensation, DC configuration and feedback structure around the SPECIFIC op-amp you have selected. If you decide to change the op-amp to a different product, you must reconfigure the circuit around it.

What happens if you don't reconfigure the circuit when you change op-amps? It depends on your design and the changed parameters of the new op-amp. If you are very lucky, the performance will be only slightly degraded. If you are slightly unlucky the performance will be very degraded, if you are very unlucky, the circuit will oscillate - probably at radio frequencies which can only be spotted with an oscilloscope.

Note: without reconfiguring the circuit you can NOT improve it's performance, you will ALWAYS degrade performance, perhaps tragically.

Secondly, discrete audio op-amps measure worse than single-chip designs.

So if you have a properly designed circuit and you swap the op-amp, you will degrade the performance and a discrete op-amp may increase the degradation relative to a single chip op-amp. If, under controlled tests, you can reliably identify and prefer the sound of the circuit with degraded performance that means you enjoy distortion, noise and inaccurate frequency responses. But don't feel bad, that's quite common!

 

[BDWoody]

Great post. Love the double blind test.

Where did you see that? I must have missed it.

 

[JeremyFife]

I wonder if it’s the equipment? I’m assuming every piece of equipment differs in the design. All I’ve done is do an amateur double blind test…and four people heard an improvement. I forgot to add we wrote notes down every time the music was played, it was by no means measured by an instrument. Four people came to the same conclusion, one a 20 year old woman with what I would assume has excellent hearing. I’m also a very sceptical person, however I believe there was a difference and it was a positive one. But…..from the pushback I’m getting the impression this is the “wrong” forum for that! ;-O

Many people post about the need and the benefits of controlled listening tests (blind) but not many actually try them and share the results - kudos to you for that.
Of course, then you find that there is a rigour needed to actually have adequate controls and that this really matters. If you have the interest and the energy, then this is the time to work out if your controls were good enough and how to improve them - and then retest.

Thank you for sharing what you found, I hope you persevere and add to the body of knowledge on this site.

Appreciated

 

[Sam Spade]

Where did you see that? I must have missed it.

Did you read the post? Daughter switching between identical DACs except for OP amps, hidden under a blanket. That's double blind. When neither the controller/experimenter and the test subjects know what treatment they are being subjected to.

 

[SIY]

Did you read the post?

Yes. Laughable controls even if (big if) reported accurately.

 

[Sam Spade]

Yes. Laughable controls even if (big if) reported accurately.

Laughable why? Normal people in normal living rooms can't do double blind listening tests?

 

[SIY]

Normal people in normal living rooms can't do double blind listening tests?

They can IF they read up on how to set up level-matched DBTs properly and avoid simple tells. I've written an article about this and Amir has several videos.

 

[Sam Spade]

They can IF they read up on how to set up level-matched DBTs properly and avoid simple tells. I've written an article about this and Amir has several videos.

Level matching is important. What simple tells were present in the test we are discussing?

 

[SIY]

Level matching is important.

And yet wasn't mentioned in the description. Nor any description of how tells (like from timing or position) were avoided. Nor was it double blind since one person knew which was which. And on and on. When people skip those pesky details, you already know the controls were sloppy or non-existent. The details just seem to get remembered when prompted...

Don't make people do a Brandolini, just do a proper test informed by people with experience setting them up.

 

[MaxwellsEq]

Level matching is important

No, it's not important, it's absolutely, totally, essential. Frequently people discover that differences vanish when signals are matched to within 0.1V

 

[Genpattonj]

Laughable why? Normal people in normal living rooms can't do double blind listening tests?

What I forgot to mention to mention, the levels were matched. I had decibel meter so it was as good as I could do. However they were identical units, with the meter they pretty much matched. There was my self and my neighbour. My Daughter simply did the switching. So both the same DAC matched and we had no idea with one. Then by chance we had my Daughters friend who had come over, we did the same with her and my Daughter, both who were in their early twenties. It was as good as we could do. No, not a “professional” double blind test whatever that means but it was enough to convince four people, including two young women who had better ears than we had.

 

[Genpattonj]

What I forgot to mention to mention, the levels were matched. I had decibel meter so it was as good as I could do. However they were identical units, with the meter they pretty much matched. There was my self and my neighbour. My Daughter simply did the switching. So both the same DAC matched and we had no idea with one. Then by chance we had my Daughters friend who had come over, we did the same with her and my Daughter, both who were in their early twenties. It was as good as we could do. No, not a “professional” double blind test whatever that means but it was enough to convince four people, including two young women who had better ears than we had.

Also Happy New Year people!

 

[SIY]

when signals are matched to within 0.1V

I think you mean 0.1dB? Which can be accurately done with a voltmeter, and poorly done with an SPL meter.

 

[Beave]

What I forgot to mention to mention, the levels were matched. I had decibel meter so it was as good as I could do. However they were identical units, with the meter they pretty much matched. There was my self and my neighbour. My Daughter simply did the switching. So both the same DAC matched and we had no idea with one. Then by chance we had my Daughters friend who had come over, we did the same with her and my Daughter, both who were in their early twenties. It was as good as we could do. No, not a “professional” double blind test whatever that means but it was enough to convince four people, including two young women who had better ears than we had.

Level matching with a sound meter is not nearly sufficiently precise. It must be done electrically, with electrical measurement gear.

 

[MaxwellsEq]

the levels were matched. I had decibel meter so it was as good as I could do.

That's not level matched. Using a sound level meter was a waste of time. It must be done with a voltmeter.

However they were identical units,

They were not identical units. The gain devices were different in them.

 

[MaxwellsEq]

I think you mean 0.1dB? Which can be accurately done with a voltmeter, and poorly done with an SPL meter.

Oops, yes a typo! I meant 0.01V (or better)