[limpidglitch]

Quote:

 

I'm sure there's gear of which that's true.  The audiophile gear I'm familiar with does a lot better on these sorts of things - e.g., the Wavelength stuff, whose Cosecant was measured by Stereophile as having the best jitter numbers they'd ever seen in a USB DAC; or the Theta I've got, 20 years old but designed by folks who knew jitter was bad and how to minimize it.
 
Regarding how much jitter is audible, I've seen statements all over the lot.  There are some extremely well respected people in the industry (Keith O. Johnson of Spectral is one who comes to mind) who claim differences in the tens of picoseconds are audible.  I don't know about that, but then there's a hell of a lot I don't know about audio compared to Keith O. Johnson.

Have you ever considered the possibility that Johnson and other people in the business may have some interests to protect?
 
Any serious independent study that has been made on the subject show that you need several orders of magnitude more jitter than what Keith O. Johnson 'think' you do, for it to be audible.
A good example is the 1998 AES paper by Benjamin and Gannon showing that under the most optimal conditions jitter under 10 nanoseconds cannot be detected. For music, where you get all sorts of masking effects, none of the subjects could detect it under 20ns.

 

[Willakan]

Based on all the stuff I've seen on jitter, claiming 10ps differences are audible seems like pure fantasy to me - like claiming you can distinguish 0.0005 and 0.0001% THD.
Audible jitter is in about the 10s of nanoseconds IMHO (as above post) which is three orders of magnitude worse. (3!)

 

[skeptic]

Quote:

I'm sure there's gear of which that's true.  The audiophile gear I'm familiar with does a lot better on these sorts of things - e.g., the Wavelength stuff, whose Cosecant was measured by Stereophile as having the best jitter numbers they'd ever seen in a USB DAC; or the Theta I've got, 20 years old but designed by folks who knew jitter was bad and how to minimize it.
 
Regarding how much jitter is audible, I've seen statements all over the lot.  There are some extremely well respected people in the industry (Keith O. Johnson of Spectral is one who comes to mind) who claim differences in the tens of picoseconds are audible.  I don't know about that, but then there's a hell of a lot I don't know about audio compared to Keith O. Johnson.

 
Not sure which version of the Coescant you're thinking of, but stereophile's measurements of the v3 actually revealed some serious flaws.  See http://www.stereophile.com/content/wavelength-cosecant-v3-usb-digitalanalog-converter-measurements and http://www.head-fi.org/t/560122/all-dac-s-sound-the-same/225#post_7574387
 
 

 

[i_djoel2000]

ok, maybe this is a stupid question but someone has to ask it anyway

 
shike, have you tried b22 before? how do you compare this o2 to the amb b22 (2, 3, 4 , 5 channels, whatever)?

 

[Shike]

Quote:

ok, maybe this is a stupid question but someone has to ask it anyway

 
shike, have you tried b22 before? how do you compare this o2 to the amb b22 (2, 3, 4 , 5 channels, whatever)?

Haven't tried a b22 before, sorry!

 

[Shike]

See EDIT 2: pictures and comments on usage with K601 added.  Still VERY impressed to say the least ^_-

 

[khaos974]

I know why you have problems with the flash, the sensor on your camera has a limited dynamic range, even the best DSLR are limited to less than 14 bit of real dynamic range, point and shoot camera (or phone cameras) have sometimes less than 8 bit of dynamic range.

The very high contrast of the scene (black background vs silver amp) tricks the exposure meter and makes the flash brighter than necessary and the sensor simply can't handle this amount of light, creating burned areas in the brightest spots.

 

[Shike]

Quote:

I know why you have problems with the flash, the sensor on your camera has a limited dynamic range, even the best DSLR are limited to less than 14 bit of real dynamic range, point and shoot camera (or phone cameras) have sometimes less than 8 bit of dynamic range.

The very high contrast of the scene (black background vs silver amp) tricks the exposure meter and makes the flash brighter than necessary and the sensor simply can't handle this amount of light, creating burned areas in the brightest spots.

Yeah, it's a Panasonic LS70 which was at best "okay" at the time I got it . . . now it's ancient.  The second it dies I hope to buy a newer and hopefully better "true" compact camera (the bulk on this thing is the second largest drawback for me)

 

[estreeter]

(disclaimer: I dont have a dog in this fight - no AMB products, no CMOY, and no axe to grind with Jude or anyone here)
 
My thoughts after reading nwavguy's incredibly detailed blog entries on this amp:
 
- whatever you think of nwavguy, he makes some good points - when the hell are amp manufacturers going to included detailed specs with every new blurb on their latest 'must have' amp ? Forget expensive measuring gear - I understand they dont all have access to that - some RSA amps dont have even the most basic of specs on the product page for some of their amps ....
 
http://www.raysamuelsaudio.com/products/sr-71b
 
- nwavguy has antagonised a lot of people here, and there is no denying that he can be a thorny character, but offering $500 to anyone who can independently verify their amp tests better in a DBT is, AFAIK, the first time anyone has been willing to do this. I guess some will write it off as a 'stunt', but I dont know of too many charities that wouldnt want $500. Time for someone to put up or shut up.
 
- he has put his design where his mouth is, which is a long way from most of the keyboard warriors in Sound Science. Anyone who wants to measure/DBT/whatever his design and tear it to shreds (or not) is now free to do so - its no longer people arguing over the theoretical benefits of balanced ground or whatever. Even if this doesnt alter the audio landscape irrevocably, kudos to nwavguy for doing it
 
Now to play Devils Advocate. nwavguy has gathered a host of acolytes who lap up everything he says, accept his measurements without question and pretty much bag every other sub-$1000 amp on the planet. I dont have the technical understanding to challenge any of his measurements, or even his design principles, but I struggle with the concept that a single amp (roughly the footprint of the QA350) costing $150 assembled will drive everything from 16-ohm IEMs to planar magnetics flawlessly. Go back and re-read the early posts where he debated the merits of a gain switch, finally conceding that he needed to make it happen : I think that was a good move.
 
The final point I want to make is that he has put himself in an enviable position, assuming the amp lives up to its promise of 'wire with gain' : in any blind test where you tell me I couldnt tell the difference between a $150 amp and one costing $450, which one do you think I will buy ? Even if the aesthetics arent as good, is the additional 'sig cred' of the $450 amp worth it ? He mentions that many will favour the 'name brand' amp in a sighted test, and I'm sure thats true, but anyone paying for an amp with their own money will struggle to follow that with their wallet IF the subjective differences are negligible - at least that would be my attitude.
 
I dont accept everything nwavguy tells us at face value, and I'm far less concerned with graphs than subjective performance, but for $150 assembled, what competition does this amp have ? nwavguy may have taken some paint off my beloved E9 (output impedance too high), but for that kind of money, I may just have to put my ZO V2 plans on hold.
 

 

[maverickronin]

It not impossible that he faked all those measurements but if he did it would have to be some of the world's most elaborate trolling because otherwise its just going to ruin his reputation for no other gain because just about no one (including me) is going to pay attention to anything he says anymore if such a thing turns out to be true.
 
Given all the emphasis he's put on measurements someone is probably going to check it out sooner or later (even if all his talk about finding someone else to verify the measurements is just a smokescreen) and if both sets of data don't match fairly closely there are precious few excuses as to why that might be.
 
Also, that fact that he uses a handle doesn't matter to the fact that no one will pay attention to him anymore.  Its not like he could start another blog under another name testing with the exact same kilobucks worth of testing gear and expect no one to notice it wasn't the same person.

 

[mikeaj]

I agree; a good amount of skepticism here is healthy, but as pointed out above, the whole thing is rather believable. Personally I have a degree in EE and am working on another, but analog electronics is definitely out of my area of specialty.

The questions come down to
1) Are the measurements valid and as advertised? (edit: to be clear, I think it's borderline preposterous to think that they were faked, though 3rd party testing is always welcomed)

Shike tested the crosstalk at least, and it seems to match up. Some other independent validation would be nice though. If you have an O2 and no expensive test gear, it should be still fairly easy with a reasonable sound card to do a line out -> O2 (loaded with resistor or even headphones) -> line in loopback test. Run say the 1 kHz test tone THD test, guesstimate the output level (or measure it with a multimeter) and compare the result with the matching part on one of his graphs. Either you can see if the measurement is similar to his, or you can note that line out -> O2 (loaded) -> line in produces the same THD score as line out -> line in, meaning that the O2 (loaded) added less distortion than your sound card can measure. If it's better while driving headphones than your sound card can measure, that would be a noteworthy data point in of itself.

2) Supposing the measurements are valid, do you believe that a device with that kind of performance will sound good and drive everything as advertised, or do all those extensive industry-standard benchmarks not capture the capability (or sound quality) of the device? Does a device that does really well at reproducing combinations of test tones at a good range of output levels and into different loads, somehow fall flat with real music?

Seems really unlikely but possible. Also, why would those be accepted industry benchmarks and metrics if they were not valuable in some sense? One possible test that could be done would be to play and record some real music (again, line out -> O2 loaded -> line in) and then do a comparison between the original music file and he recorded file. Comparison could be a listening test between the two files or doing looking at the difference in the files sample by sample (and/or do some spectral analysis) to see if some uglies showed up, if anything was clipped, if somehow the treble got peaky, if somehow the bass changed in character and became less authoritative, and so on. Again, any differences could be caused by your D/A and A/D rather than the amp though.


All this in addition to just listening to it normally of course.

 

[LFF]

Quote:

Does a device that does really well at reproducing combinations of test tones at a good range of output levels and into different loads, somehow fall flat with real music?


All this in addition to just listening to it normally of course.

Music is a series of tones and silences.....
 
If something does well at reproducing test tones, it will do well with music....
 

 

[Maxvla]

"Sig cred" is overrated. Once one gets past this, you finally begin to gain from your experience at Head-fi. Many people get caught up in the who's who of Head-fi with gear lists and gold lined signatures, but you have to get past that and get back to the real passion, the music. The only reason I post gear in my signature is for a point of reference for readers of my posts.

If I find the O2 to be the best amp for me, it will find it's spot below my posts just as other gear does. Recreating the music in as transparent a way possible while not spending a lot should be the goal of any person on this site. Not spending a lot is relative to ones budget, but you should never buy anything just because it's the next new expensive shiny toy everyone has crowned champion. This is one of the reasons I highly recommend going to meets. There's no replacement for experience.

 

[LFF]

Quote:

 This is one of the reasons I highly recommend going to meets. There's no replacement for experience.

So true.....
 
That is my first recommendation to any newb that PM's me...
 

 

[LizardKing1]

Quote:

It not impossible that he faked all those measurements but if he did it would have to be some of the world's most elaborate trolling because otherwise its just going to ruin his reputation for no other gain because just about no one (including me) is going to pay attention to anything he says anymore if such a thing turns out to be true.
 
Given all the emphasis he's put on measurements someone is probably going to check it out sooner or later (even if all his talk about finding someone else to verify the measurements is just a smokescreen) and if both sets of data don't match fairly closely there are precious few excuses as to why that might be.
 
Also, that fact that he uses a handle doesn't matter to the fact that no one will pay attention to him anymore.  Its not like he could start another blog under another name testing with the exact same kilobucks worth of testing gear and expect no one to notice it wasn't the same person.

Like you said, why would he do that? Besides the reputation of being the inventor of one of the best DIY amps, what would he gain from faking those measurements? The fact that he doesn't make any money out of the O2, added with the fact that he's still got the 500$ amp-DBT running would make that possibility very slim.