John Curl's Blowtorch preamplifier part III

Status
Not open for further replies.
Things get much more complicated with DACs. For example, very few take advantage of a clear step up in sound quality that comes from using an external interpolation filter which is free to use more than the stock number of taps, and a more relaxed transition band (possible due to pre-upsampling). It is more than a subtle improvement.
Which proof is that based on?
 
So, tell you what. If you want to see for yourself, then I will pay for your plane ticket to fly here for a visit. That, I could manage.
Nice try, but you are too far away for a plane ticket offer.
Bait & switch. :rolleyes:

Even if you would still buy plane ticket for him or anyone else, what test procedure will they be subjected to? I'm asking for the sake of its worthiness for the visitor's time spent on this adventure. I often see the sound quality claims stemming from casual subjective auditioning sessions with no control for bias or consistency. In such case, why should anyone bother?

Giving the benefit of the doubt, how will you set up the comparison method?
 
@DPH, I am not equipped with a $40,000 top of the line AP analyzer to measure small but audible differences, much less have money to fund professional research into hearing, which could cost much more than the AP.

So, tell you what. If you want to see for yourself, then I will pay for your plane ticket to fly here for a visit. That, I could manage.

I apologise because I didn't make my point clear: I hardly expect anyone to have good hard measurements on diyaudio. We're more lucky when someone does. I would really like folks to make those zeroth and first order assessments based on the measurements they have and lean on the widely available knowledge base available to us all. Otherwise, when someone encounters an experience that flies in the face of established knowledge, it should give pause. What got me riled up about your post is not that you have your own perceptions but that you're projecting them forward and making judgement calls based on extremely shaky ground. I'm sure you come across differently in person than here, and I'm more bugged by the myopic nature of the message. Much in the same way I'm infinitely grouchier here than in person, even though I maintain similar levels of incredulity.

To the end of a visit: if/when I make another road trip up) down I5, I'll try to make it out your way. You guys have some great hiking/trails to boot, or at least what isn't crispy and regrowing. I am sure I'd learn plenty from you and hopefully vice versa, but in terms of getting to a greater truth of audibility versus measurements, I don't see it making a dent. Would it be useful in terms of the joy of building relationships? Absolutely, and ultimately infinitely more important than whatever high end system thingamajigs we get rankled about.

You do not need it at all. And, fancy equipment cannot and does not substitute the necessary knowledge.

Yeah, I was going to make this point, I don't have access to the same test equipment I had at my former university, so I need to tool up a few notch/low pass filters (hi, Alan!) and test boards (similar to Bob Cordell's magnifier) to put my emu0404 to work. And battery power everything. The measurement hardware needed to get to ppm sensitivity isn't too bad, the technique is.
 
Can´t agree on that so easily; actually overall wrt multidimensional evaluation it is quite hard to say, but for single threshold parameters that were in fact examined in large scale experiments, it was common to find a distribution where 1% did better than 20 dB below the mean (of course 20 dB worse as well).
Based on the numbers from experiments done in the USA during the 1940s, for example single tone at ~880 Hz absolute hearing threshold; number of participants 35,589.
Licklider printed the graph in his contribution to the Stevenson Handbook of Psychology in 1951, page 997.

Obviously there are a lot more doing worse than doing better as a lower bound must exist due to physiological reasons.

Hard to swallow, not impossible, which means our perspective is probably closer than expressed. I give a decent berth not only to physiological, but training and natural talent to get them most out of our nature. That said, exceptional cases need careful examination.

PS thanks for adding your reference!
 

The curie temperature marks a transistion threshold for magnetic properties but it is reversible process, so no damage occurs.

Mechanical stress is different but it seems to depend on thickness as a parameter; i haven´t seen numbers but application examples where adhesive mu-metall was used as a winding around wire or wire bundle. Would be of questionable effect if dropping in permeability would be as severe as it is in thicker sheets after mechanical bending.

Afair the Vakuumschmelze Hanaus uses tempering temperatures from 880 to ~1100 degrees (depending on the specific ally formula) to restore the permeability after mechanical forming. Processes consist of a couple of hours at this high temperature followed by additional hours at around 200 - 300 degrees (Celsius) .
 
Chris, you do not have a great system, IMHO. In fact you have a rather mediocre one, at best.

I finally got my lab system up, MET 7's with a 20W T-amp and my one stage phono (documented here in its own thread) powered by Jan's silent switcher and my $279 kickstarter TT. I found a 5A 12V wall wart left over from an old CPAP machine which upgraded the PS for the T-amp. The usual LP's sound awesome, mediocre ? no way. Your opinion of others experiences is just that, your opinion, and it has no objective value at all in any context.
 
Markw4, you too have been through the wall of criticisms of others, like myself, who actually listens, as well as tries to improve audio equipment. Fun huh? But our critics are just having sport with us, so we will carry on as usual. I am interested in what you have found that makes an audible difference in digital. I think you are on a path that is similar to two other separate colleagues, who also have mentioned improving digital to me. They probably will not say much publicly, until they have their patents applied for. '-) Still, I commend you for trying to make some progress here.
 
Hard to swallow, not impossible, which means our perspective is probably closer than expressed. I give a decent berth not only to physiological, but training and natural talent to get them most out of our nature. That said, exceptional cases need careful examination.

PS thanks for adding your reference!

Finally i could extract the graph, see the attachment.
(S.S.Stevens editor; Handbook of Experimental Psychology; John Wiley & sons,1951, 997)

That´s why i wrote, quantification for multidimensional perception is much harder to do, as hard evidence is rare despite the numerous experiments done wrt various formats (compression or DSD vs. PCM ). Having noticed a surprisingly (or maybe not that surprising after all) large spread in our experiments -and astonishing perception abilities on other occasions, i wouldn´t rule thing out without special care too.
 

Attachments

  • Licklider1951_880Hz_Threshold.gif
    Licklider1951_880Hz_Threshold.gif
    34.9 KB · Views: 229
I am interested in what you have found that makes an audible difference in digital.

IMHO, it requires giving attention a lot of little details. But, the specifics which are a bit different that I have been working on include improved clocking schemes, and taking best advantage of available upsampling chips with possible conversion of PCM to DSD. The last piece of the puzzle to figure out for me at this point is getting the interpolation filter out of the dac chip and in into an FPGA where it can be better optimized. A decent DSD modulator in the FPGA might be a big plus or not, have to try it to really know. Hard to say how the dac chip modulator will sound best without listening tests.

EDIT: By the way, I know about notch filters and distortion magnifiers. The latter are of no use for dacs, and single notch filters are not good for multi-tone IMD. I don't assume all distortion is stationary as a function of frequency, so that can make some measurements complicated. What KSTR is doing with block averaging looks pretty interesting.
 
Last edited:
www.hifisonix.com
Joined 2003
Paid Member
Best option for thin strip is to spot weld, a capacitor discharge is probably the simplest.

Years ago I used to work with shielded TV tuners. Provided you have flux that works with nickel alloy, soldering is also a valid option, I used discarded batteries, austenitic cheapo stainless steel or nickel sheet scrap to verify. Regular tin solder melts at around 190C, not so bad but you want to keep both time and area of heat exposure to a minimum, use plenty of flux. Whenever applicable, low temperature solder < 140C is much better. However, you should look up the datasheet of the mu metal and seek advice from people with more expertise for critical application.

Thanks for that.
 
Status
Not open for further replies.