[Oz_Audio]

Having travelled 8000km (5000 miles) return to be there, I still cannot believe the evening, the sytems, the music and the people involved.  A great time and also very, very informative.

I have been to Hi Fi shows in melbourne and Malaysia and none where as educational, thanks Ben!!

But the Star of the show was Hugh, the GK1's sounded very similar as did the 55 and 100 AKSA's.  Small differences between the AKSA amps, one could hear they where all from the same family.  

The speakers where as different as they can get.  At one end Malcolms full range single driver 1/4 wave Diotones, AKSonics and a very interesting first time home built Audax based speaker system looking very much like the big Sonus Fabers.

We tried different cables that sounded good with some systems, bad with others, but not the same with each system.

Larry had us sitting down and doing a blind ABX GK1 test with his GK1 and Mals GK1, the only person who could hear the difference 100% was Mals wife!!  

Beer, Pizza and Music combined with friendly, intelegent people, the first time this has happened to me for a long time.

My favorite for the night: Bens talk on digtal audio and probably the first time anywhere in the world where "Jitter" was explained and demonstrated so we could all hear it - amazing stuff.

And Audio Nirvana was achieved at the end of the night!

Thanks to Phil for the Venue, Mal, Hugh, Ben, everyone who brought gear and the others for just being there.

Mark

[Rocket]

hello oz,

glad you had a good time, i'm really envious but unfortunately i live too far away and could not attend .  may'be they can have an aksafest next year in sydney and with a bit more notice i may be able to attend then.

did hugh give you a timeframe for when there new dac will be ready?  i'm looking at upgrading mine shortly.  btw i have my aksa 100 being nirvana'd this weekend, hopefully it will be ready next week sometime.

i would have tried to do this myself but i'm a clux with electronics at this stage.

best wishes

rocket

[Aragorn]

Dear All

I don't know what happened to my last post on this topic but I wish also to applaud Hugh, Ben, Phil and all the other attendees for making a great night.

Hopefully the DVD will be ready in the next week.

[AKSA]

Folks,

The AKSAfest was an unqualified success, and my thanks to Mark (Oz Audio) for his flattering comments (and for bringing along his stunning GK-1 and AKSA 55!!).  About thirty AKSAphiles and prospectives turned up at Phillip Palangas' OpenPlus Software premises at Frenchs Forest on the night of 29th May, and some stunning gear was exhibited by some very passionate and talented AKSAphiles.

First, my sincere and abiding thanks to Phillip who originally had the vision, convinced me, very generously offered up his premises, and whose guidance throughout and knowledge of the industry and business in general has been an inspiration.  Thank you Phillip;  your vision, your kindness, and an evening none of us can ever forget.

Next comes Malcolm Fear, whose kindness with accommodation, food, transport and inimitable good cheer cannot be overstated.  

It was a great pleasure to meet and talk with Malcolm, whose dedication to AKSA has been an inspiration to me, and to others, and whose Diatones are nothing short of a revelation.  As a full range speaker, their performance is just astonishing.  I would like a pair, but tragically these unique drivers are no longer made.......

Larry Ye played an impressive part with his hi-tech AB test rig.  He worked hard to make to have it ready on the night so two devices could be switched for critical AB testing.  There were three GK-1s there;  standard, from Phillip, DACT modded from Larry, and DACT and additional Auricap modded from Mal.  It transpired that the only person able to accurately distinguish them was Mal's wife Toni, and this leads to some interesting conclusions.  It did emerge that interconnects and speaker cables have huge influence on the final presentation, something which caused me some unease, but there it is!!  

On Friday 30th, Larry drove many long miles in heavy traffic to pick us up from Frenchs Forest to dine with him and his wife and son and the evening was wonderful with exquisite food and great company.  His gear is unbelievably well organised, beautifully made, and his junkbox makes my working equipment look dowdy!  I was incredibly impressed with his intellectual rigor and discpline;  this is a man whose organisational and analytic skills are 100% pure engineering and scientific.  Hats off, Larry, and thank you very much for your kindness!

I would like to thank David Collins, who videotaped both presentations with a Sony DV.  During one of the group photos, David, an intensive care medico, cracked us up with a hilarious medical one liner!  (the verb choice is quite deliberate.......)  David, this was very kind of you, and I'm very grateful.  It was a really important link in the chain, and those AKSAphiles unable to be there will be thrilled with the result.

I was bowled over by the incredible AKSAs I saw all around me, and the people behind them.  All had their stamp, a reflection of the priorities and artistic approach of their builders.  All of them sounded fantastic!  It was wonderful to put a face to the people I had known only via email, and the interest this creates in the different personalities and approachs is wonderful.  It really was a privilege to meet these guys, and in the convivial surroundings, with beer, pizza and good cheer, it was great to see so many exchanging emails, addresses, and experiences.  I learned just how important this network stuff was on 29th, and am presently thinking about the next one, when, where, how.  Ultimately I want to do this in the States and Europe, but one step at a time!

Ben's presentation about digital audio was fascinating!  His walk through the world of jitter (and there are at least four different types of jitter problems!!) gave us insights we had never heard before, and when played a CD with all the different jitter categories, we all realized just how terrible the digital distortions really sound.  The clincher was when he played the subtraction signals, the error output, you just knew than if added to the original it would sound muddy, indistinct, and unmusical.  This is particularly malevolent distortion, of a type quite unrelated to the analogue distortion I wrestle with in amplifiers and preamps.  Ben's presentation style was redolent of our best uni lecturers, and superbly delivered.

My presentation had originally been written as a 4,700 word speech.  When I met my audience, I knew they woud not stand for it, and would want something more spontaneous.  So I just used the overhead transparencies and gave a fireside chat on the Aspen Amplifier Company, ad libbing my way across the chronology, development and strategy.  It went across well, though a little too long perhaps, but gave clear insight into the inspiration and drive behind my passion for audio.  I felt that was pretty much what people wanted, and it was a real pleasure to meet people, talk animatedly about audio, and the feedback was fantastic.

There is huge interest in the AKSADAC and the GK-1 remote.  These are essentially digital products, and represent a new page in Aspen's development.  Ben's expertise has been a revelation here, and as we saw at the presentation, his ideas are world class.  I believe these long awaited products, when they are available, will be hugely popular, because there are so many horrific compromises in present day digital technologies.  I am convinced that we can jointly produce a DAC so good it will best almost anything available in the market today, but it will only be possible to deliver this excellence at moderate price with Red Book CD, and I'd like to make this clear.  This DAC will only be suitable for the 16/44 CD format, so-called Red Book, which has huge room for improvement on playback, but it will enable those with large CD collections to fully exploit their collections for decades to come.  In truth, a universal product is certainly a 'nice to have', but the intrinsic compromises are unacceptable to any one standard, and besides, royalites may be due for SACD systems anyway.  This scenario is not the answer to us as audiophiles, and hence our resolve to offer a DAC solely dedicated to the Red Book CD standard.

In light of delays on other products, I'm no longer prepared to give deadlines.  New products are always contingent on company income and R&D constraints, and I'm not willing to release anything unless it's been thoroughly tested and is the best it can possibly be. This policy is a mix of pragmatism, having been caught before, and determination to deliver the best possible at optimal cost.  I apologize for this approach, because I know people are impatient for the AKSADAC, but I think it's a fair tradeoff given the complexity and cost of the development cycle.

Ben, Michael and I really enjoyed the AKSAfest.  We want to do more, company funds permitting.  An AKSAfest creates great networks, is a wonderful support for existing customers, and a good forum for prospective customers.  But it's an expensive exercise, like all marketing endeavours, and to a large extent future AKSAfests, in Australia and abroad, depend on continuing sales.  So I'd ask all AKSAphiles to keep talking about their products, to help each other where possible, and to be patient with new products.

Once again, my heartfelt thanks on behalf of Ben, Michael and myself, for the unqualified support, kindness and graciousness extended to us on this AKSAfest from all of you!  May there be more of them into the future!!  I will get cracking on the CDROM of our presentation, and when it's ready will let you all know both on this forum and on the website.  I will likely get it out the door for around $US15 including postage anywhere in the world.

In closing, I offer my thanks to all interested AKSA customers, prospectives, and a very special thanks to all those who brought along their gear, make a big effort to come along and showed their AKSA products on the night.  It showed me there is huge interest in AKSA products in Australia, some very, very clever and accomplished people, and a huge awareness of the joys of recorded music in the domestic setting.

Thanks guys,

Cheers,

Hugh, Ben, and Michael

[Al Garay]

Congratulations to all of you for pulling this off. The first is the toughest. Perhaps, in the future you can do a web broadcast. I'm sure the technology will be available where we will be able to participate over the web. The listening will be tough.

I will hold off buying a DAC until I hear more about the AKSADAC. I'm hoping it will work with my digital coax output from my multidisc DVD player.

Cheers,

Al Garay

[Jens]

Congratulations, Hugh et al! It really seems that the AKSAfest was a huge success.  

All of us who are obviously too far away (Denmark, Europe) to have been able to make it are extremely envious. However, I take it that some material from the fest will be made available to all? You mention a DVD - please let us know if and when, and where it becomes available.

In the meantime, how about putting a few photos on this forum or on your homepage? Just to whet our appetites, you know ...

I hope to be able to listen to a GK-1 remote at some point, as this would be the next item in my chain that could improve performance further. It would definitely complement my AKSA 55 N beautifully. The DAC also sounds very, very interesting! Finally, any extra improvements to the 55 N that you can come up with would always be welcome. Not that it is not very, very good already - but aren't we always trying to improve?

Finally, didn't the cable differences you heard set the little wheels and gears in your head whirring? Perhaps we shall see AKSAcables in the future?

Good going!

Cheers,

Jens

[Larry]

I think I should say something about the AB test which some people may like to know more details.

First of all, I'd like to thank everyone who had participated in the test and appreciated their enthusiasm and courage. Thank you very much for your contribution to our journey to the audio nirvana.

Methodology

Assessment of subtle differences in HiFi systems is a difficult work that requires careful design of tests in order to obtain reliable results to make conclusions of significance. I am aware of some standards in subjective assessment on audio systems like EBU (European Broadcast Union), ITU (International Telecommunication Union) and some industrial test procedures, most of which are designed to assess the impairment of reproduction of lossy coded signals against an original. However, in HiFi audio, there is no original but two different systems. In my opinion, in order to assess various aspects of HiFi systems, 3 tests are necessary: ABX test, ABA test and Critical Listening test. As we only did the ABX test, I will concentrate on the design of the ABX test here.

The ABX test is double blind, in which not only system in play is unknown to the subjects but also the system is randomly selected in the test. I produced several random sequences and randomly picked one to use to select the systems under test.

An ABX test is worth doing when the difference of two systems is not obvious -- it is difficult to tell  the difference by simply listening to them in normal casual listening sessions. This is our case when we listened to two GK-1s with minor variations like different capacitors. An ABX test can establish the existence of any subtle perceptible differences, if any.

Test Materials

The test materials are very critical which must stress the systems to exhibit their differences. Some systems are more suitable to certain types of music and in casual listening, people tend to demonstrate their favorite music that sounds good on their systems. In this test, I used 10 short pieces of music with a range of characteristics from high, medium to low ranges, simple solo instrument to orchestra, slow and fast etc. Each piece is only about 30 seconds long to draw the attention to the limited elements of sound. To assess subtle differences, we must use our short memory and the brain can only process less than three logical elements at a time. Too many elements in a longer passage will confuse our brain and result in unreliable judgement.

1. Come Away With Me, starting at 0:00: solo female vocal - medium frequency timbre and purity (Noral Jones, Come Away With Me, Blue Note Records, 7243 5 82067 2 2)

2. Time, starting at 0:10: bells - high frequency timbre, imaging, layering (Pink Floyd, Dark Side of the Moon, Mobile Fidelity Sound Lab, UDCD 517)

3. Stranger in My Bed, starting at 0:50: female vocals - imaging, focus; a good system can tell how many singers in the back-up voices (The Power of Seven, Up Front, Sheffreld Lab, 10033-2-F)

4. Dance of The Knight, starting at 0:00: symphonic orchestra - imaging, depth/width, layering (The Leinsdorf Session, Vol.1, Sheffreld Lab, 10043-2-G)

5. Zhouma Of Zhouma, starting at 0:35: new age (vocal + synthetic electric music) - low frequency extension (Dadawa, Sister Drum, UFOCO, 4509 99592-2)

6. Poem of Chinese Drum, starting at 5:00: percussions - low frequency timbre, transients/speed, bass resonance (Yim Hok-man, Master of Chinese Percussion, Marco Polo, 8.225942G)

7. River Seeks The Deep, starting at 1:25: female vocals - resolution, clarity; only a good system can intelligibly render the lyrics (The Absolute Sound, Harts of Space, HS11103-2)

8. Gypsy Serenade, starting at 0:15: violin and piano - acoustic timbre (Arturo Delmoni, Songs My Mother Taught Me, John Marks Records, JMR 1G)

9. Julsang, starting at 1:00: organ and carol - holographic imaging; a good system can tell the size of the church and the arrangement of the choir (Cantate Domino, Proprius, PRCD 7762)

10. O Sole Mio, starting at 2:12: male vocals - midrange timbre (Pavarotti and Friends 2, Decca, 444 460-2)

I have others but the aboves were used in the ABX test.

Test

After the subjects listened to two different systems in random orders (A and B) then an unknown system (X), they were required to identify the X system by giving an answer as:

A for X is A system;
B for X is B system;
X for X is unidentifiable (no difference).

An answer is considered wrong only when A is said to B or vice versa. Answering X is not a wrong answer.

The systems we managed to test were GK-1s built by Malcolm and Larry.

Malcolm's GK-1: DACT attenuator,  3 Auricaps in signal coupling, Teflon CAT5 internal wiring.

Larry's GK-1: Goldpoint attenuator, 1 Auricap in output coupling (stock), PVC CAT5 internal wiring, EMI shielding (transformer and attenuator), mechanical isolation (minimizing microphonic effects)


Results and Data Analysis

All answers are subject to a confidence test before used for induce any conclusion. In this case, the answers with 8 correct answers out of 10 were considered valid and others were excluded. That is, we can, with 95% level of confidence, to say that the subjects with more than 8 correct answers are able to distinguish the differences, in other words, with 95% probability the results are not produced by guessing. (95% level of confidence is a common threshold in statistics. Everyone can guess about 5 correct answers out of 10 without listening! That's why a confidence test is important.)

There were 2 subjects with 8 or more correct answers. (Note: After the test, I was told only one person with 2 ro less incorrect answers. Now when I looked at the answer sheets at home, there is one more person with only 2 incorrect answers. He told me he had 6 wrong answers at the night but in fact he counted the 4 X answers as incorrect answers.)

Correct answer:   B A A B B B B B B A
Subject 1:                   B A A B A B B B B A    (9 correct)
Subject 2:                   B X X B A B X B X B     (8 correct)

So, there are 17 correct answers from 2 subjects. To conclude that there are perceptible differences in these 2 systems with 95% level of confidence, there must be more than 12 A/B answers out of 17 correct answers. In this case, we have 13 A/B answers out of 17 correct answers.

Conclusion

We can confidently say that there are very subtle perceptible differences between these two systems under test.

Discussion 1

The test shows that Subject 1 (Malcolm's wife, Toni) has incredibly sensitive ears and brain to identify very subtle differences. The differences are instead very subtle from the fact that majority of the participants were unable to distinguish them. Subject 2 also has a remarkable skill to give reliable answers - he gives an A/B answer when he is sure he has identified the differences and gives an X answer when he is not sure by not trying to guess the answer.

Discussion 2

What happens to the wrong answers for these two subjects? Does that mean that they were also guessing? Or in general, did other people giving wrong answers guess?

No, not exactly. In statistics, if a sample set does not exceed the 95% level of confidence, it is said to be by guessing, which means the answers are statistically close to guessed answers. But it does not mean the subjects are really guessing the answers, or subjects are not able to perceive the differences. For instance, in a test material, there are 2 sound elements of two drum beats, Beat 1 is low and Beat 2 is very low. System A renders Beat 1 to 60 Hz and Beat 2 to 50 Hz; System B renders Beat 1 to 50 Hz and Beat 2 to 40 Hz. When System X is System A and the subject might use the Beat 2 at 50 Hz in System X to compare to the Beat 1 at 50 Hz of System B and makes his/her judgment that System X is System B, which leads to a wrong answer. In fact, he/she is able to perceive the difference but does not align the sound elements properly in making the judgment. This also explains why some people swear one system is better than the other but other people swear it is not; people may compare the taste of the part at the side facing the sun of an apple to the taste of the part at  the other side of an apple . That is partly why the material in AB tests needs to be short to limit the sound elements for the subjects to focus on.

So, people should not feel too frustrated by giving some incorrect answers. It is wrong to conclude that people giving some wrong answers in ABX tests are not able to hear subtle differences. By participating in more AB tests, one can improve one's skills in identifying differences and make sensible assessment on the performance of systems.


Discussion 3

One person had voluntarily indicated her preferences in the ABX test. Her preferences were:

BBBBBBABBA

As the A/B was randomly selected, the preferences almost evenly distributed to both systems after having associated A/B in different trials to the systems. So, basically there is no preference over any system. Each of them sounds differently but not better than the other.

From the preference answers, we could see that she tended to prefer the last one she heard. This is probably because she could catch more details when she listened to the material the second time and believed that the second system sounded better. When the material is presented by System A the first time, the subject is unprepared and only can observe few details while the material is presented by System B the second time, the attention could be paid to observe more details. That's why the material needs to be presented by System A again to fill in the details missed in the first time to make a fair judgment. An ABA test works as first A is a preview to the material; then B to observe details in B and second A to observe details in A and make judgment based on details from B and from second A.

Unfortunately, we could not manage to do an ABA test to assess preferences. The night was too short! If I had at least 10 sets of answers in ABA tests, I could tell which system sounds better in treble or renders a better soundstage etc by lassociating the categorised preferences to the test materials. But from the ABX test, we could not make such conclusions.

Remarks

1) As the subwoofer was not working, the system could not go as low as it should. So, the test was conducted without adequate bass reproduction.

2) There was strong EM interference over the signals. When the switch switched, there were annoying clicks even when there was no music playing. This was an indication that there were strong electrical activities over signal links which could be a result of strong common mode EM interferences,  possibly from the building or the equipments over the unshielded twisted CAT5 pairs used for interconnects. The switching caused transient producing frequencies in audible range that could not be canceled by the differential input pair. We don't know how the EMI distorted the signals. That's why I prefer to use shielded interconnects.

[mb]

Larry,

Thanks for the comprehensive description. I too was disappointed not to be there. We've had our small AKSA powwows here in Singapore, and have wondered just how much of a 'family sound' AKSAs have, and the variance between units, due to choice of x'formers, internal cabling, etc.

2) There was strong EM interference over the signals. When the switch switched, there was annoying clicks even when there was no music playing. This was an indication that there were strong electrical activities over signal links which could be a result of strong common mode EM interferences, possibly from the building or the equipments over the unshielded twisted CAT5 pairs used for interconnects. The switching caused transient producing frequencies in audible range that could not be canceled by the differential input pair. We don't know how the EMI distorted the signals. That's why I prefer to use shielded interconnects.

Imho, that's sometimes also due to dc offsets. Switch from a source with 0 offset to one with highish offset -- click. Switch back -- click. What switch was used? Make before break, I presume?

Cheers.
mb

[dogears]

Congratulations!  

[Larry]

Imho, that's sometimes also due to dc offsets. Switch from a source with 0 offset to one with highish offset -- click. Switch back -- click. What switch was used? Make before break, I presume?

You are right. If there are DC offsets, we could hear clicks as well.

In the system we had, I understood that there is no input or output cap omitted so that there should not be any DC on the signal lines.

I don't hear any clicks when I switch my GK-1 at home. So it must be enviromental; that's why I pointed my finger to the EMI. Another observation was that if the interconnect was unpluged (open), there were strong hums, which picked up from EMI. My guess is that there was a floating case (unearhted) around.

I use remote controlled relays to do the switching.

[AKSA]

Larry, Mervin,

In my experience any DC offset in excess of around 10mV is clearly audible at the input to a power amplifier as a click on switching.

Damn nuisance, you have to get rid of them........  

Cheers,

Hugh

[Larry]

In my experience any DC offset in excess of around 10mV is clearly audible at the input to a power amplifier as a click on switching.

Yes, I have noticed that as well. So my power amp switch has a delay in switching the amp output and signal input to avoid any possible clicks due to offsets.

In my source switch that I used, I have some delay as well so that there should not be any clicks if there are no strong coupled interferences.

In my room, I have taken care of possible EMI so that I don't have any clicks while switching the GK-1 or AKSA.

Do you mean the GK-1 could have 10mV offset at the output?

[AKSA]

Folks,

brief report and group photo now under NEWS on the AKSA website.

More to come!

Cheers,

Hugh