ABX Comparator Event?

Screenshot 2014-05-12 09.07.53Hello everyone,

This is a request for direct feedback from you. Yes, this unit is potentially a game-changer for our industry – but unlike previous events, some of the audio gear we would be testing (power amps) may be more physically challenging to bring with you without hiring a moving company!

So, before scheduling a Power Amp Smackdown, I’m reaching out to ascertain if you would be willing and able to bring your wonderful amplifier to this gathering, or if this is just not realistic.

This event will only work if we have a bunch of different amps in all shapes, sizes and power profiles to demo, so please chime in and let us know if this is compelling enough to you to merit the extra effort. Please post your thoughts by writing a comment to this post (click below).

Looking forward to hearing from you on this.

Cheers,  Alón


  1. says

    I think that the ABX box is potentially the most exciting device featured in a BAAS event in years.

    Why? Because it should be a great tool to advance our understanding of component evaluation.

    If my high expectations for its utility are met, it could be a useful device for BAAS to own — and loan out or rent to members.

    Note that I am not suggestion that DBT is the only — or even best — method of system evalution. Just that it’s potentially very useful.


  2. Drew says

    I’ve got few transportable amps, all humble and “vintage” by most peoples’ standards:
    Audire Crescendo (stereo, 75wpc, solid state, circa 1980); conrad-johnson MV45 (stereo, 45wpc, tube); conrad-johnson MV75 (stereo/unbridged, tube, circa 1980); Fisher x101 (stereo, integrated, circa 1963); Heath(kit) (a couple of tube models I’d need to check for functionality); Dynaco ST-70 (would need to solve audible tube shorting/arcing problem first).

    Beware though, I travel A LOT, so my ability to attend is iffy.

  3. Alón says

    I can bring three amps, from newest to oldest:

    ~Adcom GFA 6000 (5 channel 100WPC)

    ~Electron Kinetics Eagle 2A (my little high current 100WPC monster)

    ~1976 vintage 350WPC Dynaco ST-416 designed by Hafler to drive just about anything.

    One thing for sure – I certainly would not have to go to the gym that day!


  4. says

    I’ve got a real light weight (less than one pound) tripath amp class T (really class D) that I already ABX’ed with a expensive high power amp and found them comparable at moderate volumes. It doesn’t do too well with nominal 4 ohm rated speakers but did very well with my 6 and 8 ohm nominal speakers.

    It is definitely not a “looker” that would appeal to audiophiles, but if the amps are hidden behind a screen everyone might discover something that they would otherwise be dismissive of. So what is the sensitivity and nominal ohm rating of the speakers that would be used?

    I think the ABX box is the best thing that BAAS can do to make our evaluations more objective.


  5. Dan Rubin says

    This is a great idea. I think we should limit ourselves to a manageable number of amps. That said, let me pile on two more candidates:

    1) Sanders Magtech (this is a killer amp and I’d really like to see it included)

    2) DIY Hypex NCore 400


  6. Larry says

    Am very intrigued by this shoot-out as I’m in the market for a new amp (most likely tube). I have a modded Jolida SJ502A with KT120’s. It’s an integrated, but I use it like a power amp with my CJ ET3.

    thanks, Larry

  7. mixpro says

    Well I could bring my DIY Randall III tube monoblocks from this year’s Ed Yang class. (He says it’s his last.) They sound sweet to me. It would depend on the date.


  8. ori says

    For once I’ll play the Devil’s advocate.
    Adding the ABX box in the signal path means adding another component. Unless that component is of the highest quality (a requirement for lab equipment!), it will compromise sound quality to some extent.
    If it’s of much lower quality (I don’t know the ABX, so it’s just “if”) than other components in the system then it might become the “weakest link” in the audio chain and all comparisons become irrelevant.
    My second argument is against comparisons of amplifiers using a single set of speakers. In many cases, the match between speaker load and amplifier output impedance is crucial. You can possibly compare high power solid-state amps with high levels of feedback but it’s hard to come to any conclusions when comparing solid-state to typical tube amp or a current-mode amp. The load impedance requirements of each of these classes is vastly different and the conclusions of such a comparison would be very specific to the set of speakers and cables used for the test.
    For those who think “it’s high impedance so it’s tube-friendly” – don’t. Just check the measured data in Stereophile speaker reviews to get a feel what the single impedance number really means. It’s a “probable guess” at best.

  9. Drew says

    Fine arguments. (They’ve been around since the ’70s – probably before.) Personally, I see the proposed process of ABX comparisons as demonstrative, not definitive. I’m not going to reach conclusions based on one situation – particularly a situation that is unfamiliar. (Room, recordings, ancillary equipment, bunches of people and equipment in the room…) On the positive side, introducing the ABX is a “set” change” it’s there, whatever amps and speakers surround it. Absent gross (worse yet, random) distortions, I don’t feel like it obviates the entire listening comparison.

    Hopefully, crippling amp/speaker mismatches can be avoided or, at least, recognized when they occur.

    For me, this is about learning to listen and to experiment, not shopping for a new hi fi.

  10. Dan Rubin says

    I agree with Ori’s second argument about the desirability of using more than one pair of speakers. I have several pairs of modest but capable stand mounts (Revel, KEF, Harbeth), which are easily transportable, but I’m not sure how useful they would be. Seems like what we’d really like to throw into the mix is something like Maggies or horns or something else that will present a different set of challenges to the amplifiers under test.

  11. says


    Your concerns don’t make sense. If the ABX box, as another component, comprises sound quality to some extent, so what, it should do it equally and not much differently than if you added an extra 10 feet of speaker wire to the set of speakers.

    In your second argument the matching between speakers and amps should only be of concern if one is using very low sensitivity and low ohm speakers to test between all amplifiers. In that case, the solid state high power amps would have the clear advantage. We can eliminate this concern by using speakers appropriate to all amp capabilities.

    As far as adding another component I notice that quite a few audiophiles add extra components to their audio chain that I doubt very much are necessary, such as outboard DACs which pass the signals through more elements than if they would just use the DAC that is already in their device.

    To do a real test we need to test for just one variable at a time. So we must use the same speakers and wires for all testing. Even than the test won’t be perfect in a room full of people possibly moving about and with possible distractions and people making noises and comments. A scientific test would have just one listener in the room and other protocol that would eliminate all sources of bias.

    When I did my AB test with my friend I told him to first listen for sound level matching. Once we adjusted the volume levels on both amps so the levels sounded the same then we went to the next stage to listen for sonic differences. He had no idea which way I was flipping the switch. Neither he nor I heard any real difference between my plus $1,000 receiver and my $20 tripath amplifier at the normal listening levels that we used on my 6 ohm nominal high quality speakers.


  12. Dan Rubin says

    We should match volume levels electrically at the speaker terminals, not by ear or even by SPL meter. I believe this is the recommended (or required) methodology for Stereophile reviewers. Perhaps Jason can obtain JA’s relevant specification.

  13. Alón says

    Vince and Dan: It’s actually even better than that: there are 100 volume steps in increments of 0.1dB… These increments are beyond what my ears can discern. Adjusting the volume for matching is amazingly easy via remote and clearly seen on the screen.

  14. Alón says

    Ori, et al:,

    In my conversations with Frank Van Alstine about the ABX box, he assured me that the “added component in the signal path” issue was addressed in the design of the ABX Comparator – and that sonically, it was completely transparent in their in-house listening tests. I concur. I was listening for it, but didn’t hear it. If it’s present at all, the box’s sonic signature is so small that, to my ears, makes signal degradation a moot point for our purposes.

    And, yes, making it even less problematic is that we would be applying the very same
    “known factor” equally to all components tested.

    That said, now that this conversation has gained some traction, its time to ask you all if you felt that adding another set of speakers of different impedance and/or sensitivity for matching the different amp personalities would be a good idea, or would confuse matters?

    I had two amps and two bookshelf speaker sets while testing the ABX box – and although it was so much fun, I’m leaning towards AB-ing the amps with one high-end speaker pair at a time, maybe a 6-Ohm design with higher sensitivity would be a good compromise for most amps? Anybody have a Raidho pair they can bring?

    It may help if we frame this exciting yet inherently flawed process as an Amp Smackdown, working with one set of speakers – and then do a separate Loudspeaker Smackdown event using one amp and lots of speakers.


  15. says

    I could offer to bring my beautiful marble Nohr 9.0 speakers which are nominal 6 ohm 85.5 sensitivity. They need a solid wide stand like two end tables for the pair. I would need some assistance getting them out of the car and into the event.

  16. Dan Rubin says

    I’d argue for having a second pair of speakers, not for purposes of speaker comparison but for having a second contextual data point for assessing the amps. My preference is that we emphasize current, well-known products for both speakers and amps, but especially for the speakers, to keep the comparisons relatable and relevant.

  17. says

    I forgot to add that adding a second set of speakers would not be enlightening if the goal is to test the results of amplifiers. If we use hard-to-drive speakers than the results would be “unfair” to low power amps. We could test all amplifiers (to reasonable time limits) only if we use speakers that are of average or high efficiency. Even my middle-of-the-road efficiency high-quality Nohr speakers performed very well with my Tripath low power amp which specs distortion in the very low range at up to about seven watts per channel which turned out to be pretty adequate for my 14 x 18 x 9 listening room.

    Speakers on the other hand usually have sonic characteristics, that some may prefer and others dislike. Some speakers sound particularly good with certain instruments, but not as good with other instruments. I’ve also heard high-end speakers that sounded real nice at low volume levels and fell apart at high levels. And I’ve heard high-end speakers where the cabinets resonate at certain frequencies. All speakers have sonic differences so I’m not sure an ABX test would be the way to go, actually I think it would be a waste of time. In my opinion the ABX comparator is better used on things like amps, players, DACs, turntables, and gizmos that purport to improve the sound.

  18. Alón says

    Let’s agree to begin this first Amp Smackdown with a single set of speakers first, and see how we do. We can always have another pair on hand if the testing feels compromised in any way by our chosen speaker. We’ll learn as we go along.

    So, now that we have plenty of amps to audition, we need to address the rest of the set-up for the ABX Box event:

    First, let’s have some nominations on a speaker that you know to be easy to drive, plays nicely with tubes and transistors, is fairly revealing and light enough that you can bring with you to the event.

    Next, we’ll need a decent preamp. I have an Audio Research SP4A Solid state I could bring… Anyone else with a preamp candidate?

    Source. Ideas? CD, Computer files? SACD? Something easy to work with.

    Finally, cables. I cobbled together a motley mix of ICs and speaker cables from what I had laying around and terminated them all with gold bananas. We can go with what I have, unless one of you has a bunch of higher quality banana and RCA cables you could bring… I think we need 8 sets of each or so… thoughts?

  19. Dan Rubin says

    Where are we having this event? If we can do it at Leslie’s, her Vandys can work. And we’re getting familiar with how they sound, which is helpful.

    Since the ABX box has attenuation for level matching, do we even need a preamp at all? The less there is in the system, the better we should be able to isolate the sound of the amps. Nothing chokes a system more than a fair to middling preamp, IMHO.

    FYI, I have three outstanding passive preamps of different technologies. I’m sure some of you think passives are crap, so whatever. I have the Lightspeed (light-dependent resistors), Bent TAP-X (autoformers) and Music First Baby Reference (transformers). The Music First is far and away the best of the bunch, but the others are superb at their modest price points.


  20. Alón says

    Dan, I tried it without a preamp and it was difficult matching SPLs between my amps, so I called Frank. He said a preamp is necessary because the unit is not built to be a preamp and the attenuation is for fine adjustments. I told him there may be a missing element to the diagram in appendix 2 which shows a Simple system config. He agreed, but also gave me the number of the chief designer of the box to double check. I’ve been trying to reach him.

Leave a Reply