January 19, 200818 yr Done by a 3rd party who happens to own both units. Someone I trust who has no hammer to pound with. I see no reason not to post this, as it should be clearly duplicatable by just about anyone. http://gilmore.chem.northwestern.edu/picovspredator.htm
January 19, 200818 yr Wow. I cant believe how close the 2 units are in terms of frequency response. Almost have to wonder if Ray is using the same DAC
January 19, 200818 yr frequency response is a pretty crappy indicator to use, honestly almost any modern ss source or amp is going to have a pretty damn flat response. the interesting differences here are the crosstalk numbers.
January 19, 200818 yr frequency response is a pretty crappy indicator to use, honestly almost any modern ss source or amp is going to have a pretty damn flat response. the interesting differences here are the crosstalk numbers. Don;t forget the THD numbers as well/
January 19, 200818 yr The Pico's graphs just look much cleaner overall in terms of noise floor, etc. Distortion figures from RMAA should be viewed as relative and not absolute values in my opinion. RMAA has a reputation for producing distortion numbers that aren't really believable.
January 20, 200818 yr the Predator, while obviously outclassed by the Pico, in that regard, has reasonably good distortion numbers, i think. Really? The crosstalk is interesting from the standpoint of rising with lower frequency; contrary to what is often seen (although probably due in part to the very low feedback resistors used). However, I think the distortion graphs are both more interesting and more meaningful from the standpoint of performance. That is a lot of IM distortion, and that's at nearly full scale with a stable waveform. On actual music, it's probably even worse. Distortion products at nearly -80dB with a ton of odd and high order junk along with noise is rather unimpressive. In person, the story wasn't much different (to me, anyhow).
January 20, 200818 yr the interesting differences here are the crosstalk numbers. No kidding, this is the first time I remember seeing crosstalk figures get significantly worse at low frequencies.
January 20, 200818 yr Author No kidding, this is the first time I remember seeing crosstalk figures get significantly worse at low frequencies. Actually this is exactly what you would expect from a tle2426 (or compatible) rail splitter that has to fight a low impedance load. Ray decided to seriously lower the values of all the feedback resistors on the opamp such that at medium and high gain positions each channel puts about 120 ohms of internal load to ground. That is 60 ohms that the rail splitter has to drive before you actually add any headpones. So without any tantalum low impedance caps on the output of the rail splitter where they would actually do some good, the output of the rail splitter is going to sap up some of that seperation.
January 20, 200818 yr Can someone explain to the smaller brained of us how to read these graphs? Pico very good. Predator not so good. Thus ends the extent of my knowledge.
January 20, 200818 yr I'm not sure why the Predator has both 60 and 120Hz residuals. The Predator's graphs don't look that bad on their own (other than the crosstalk one which is really bad). It's when you see the Pico's superimposed that it starts to look bad. I always wanted to know what causes this (for the Predator... this is the IMD graph... all the squiggly crap). Intuition tells me this is the rails being modulated...: [img width=150 height=93 alt=imd1]http://www.head-case.org/gallery2/main.php?g2_view=core.DownloadItem&g2_itemId=1279&g2_serialNumber=2&g2_GALLERYSID=3b85edc3f3e17fc5e8b25ff2ef57eac4
January 20, 200818 yr It's great to see some data to backup what we already knew about Ray's "design skills".
January 21, 200818 yr I always wanted to know what causes this (for the Predator... this is the IMD graph... all the squiggly crap). Intuition tells me this is the rails being modulated...: [img width=150 height=93 alt=imd1]http://www.head-case.org/gallery2/main.php?g2_view=core.DownloadItem&g2_itemId=1279&g2_serialNumber=2&g2_GALLERYSID=3b85edc3f3e17fc5e8b25ff2ef57eac4 Looks like part digital-side stuff to me, and part analog as well. Poor filtering/noise shaping and perhaps a sloppy modulator on the 2704 coupled with the peculiar aspects of the Predator design. You can see the digital-domain stuff in the Alien DAC RMAA as well - http://www.amb.org/rmaa/AlienDAC_16b_44k_battpwr_transit_20070420.htm. However, the Predator's results are still worse.
January 21, 200818 yr How were the amps being powered for these graphs, battery or AC? I assume that both were being powered the same way.
January 21, 200818 yr Author Dac's both powered from the usb port. Both units run from batteries to avoid any possible ground loops. Same intel based computer used for both. I think i know what all the IMD crap is on the predator, but until i get some time to test and verify, i won't say anything. Lets not forget that jude absolutely loves the predator. Can anyone continue to trust jude's recommendations at this point? Slightly biased maybe, no never... I think it is the same problem where the predator makes lots of hiss on some laptops... All i have to say is Induced Harmonic Compensation. Yeah, that's it... Ray is copying Rudistor. Truely classic.
January 21, 200818 yr All i have to say is Induced Harmonic Compensation. Yeah, that's it... Ray is copying Rudistor. Truely classic. The dark forces are colluding against us...
January 21, 200818 yr Ray is copying Rudistor. Truely classic. Remember, "Imitation is the sincerest form of flattery"
January 22, 200818 yr Holy stereo crosstalk! So this test was with both their USB DAC's only as the inputs? How about running the analog input stage only test as well? When I heard the Predator, I heard significantly more noise with the USB input than with the analog input. -Ed
Please sign in to comment
You will be able to leave a comment after signing in
Sign In Now