Add a $2,000 Unicorn Tears A/C cable and call it the "Pro³" ???
Add a $2,000 Unicorn Tears A/C cable and call it the "Pro³" ???
lacking Bluetooth might help push the cost of the device down since besides the cost of the actual chip they won't have to pay the licenses for all those codecsThe only thing that this device lacks to make it perfect is bluetooth![]()
And yet, for the Quant Asylum QA402, they switched to the PCM1794A from 2003ish after the AKM fire. It seems to be able to hold its own, which I would not at all have expected from its data sheet.It's impressive how much DACs have gotten better since 2007 when the WM8741/2 came out.
Heck, I didn't realize this product used such a pricy IC, ouch, I just looked it up, and I've read in this thread that the D30 pro use 4 of em. I wasn't seeing this, price TBD, but are we looking at some higher end, near flagship territory here? by Topping naming convention... 30 normally means cheap, this is an expensive design, unless they've really cut corner at PSU, but the numbers are good too...
I consider AK557x a downgrade especially when recording at single speed, though it is compensated by the ability to record at higher sample rates (384/768kHz) to deal with the aliasing and ripple.AK5394A level was -110 dB THD+N.
https://www.audiosciencereview.com/...-asus-with-a-pcie-sound-card.6104/post-199165What board or ADC model does your PCM4220 live in?
I consider AK557x a downgrade especially when recording at single speed, though it is compensated by the ability to record at higher sample rates (384/768kHz) to deal with the aliasing and ripple.
I think this site is still using a (modded?) E-MU 1820m as the analyzer, which uses AK5394A.
https://reference-audio-analyzer.pro/en/about.php#gsc.tab=0
The best ADC I have at this moment is a PCM4220.
I suppose not the same thing since you are talking about specialized analyzers but mine is a soundcard. Output volume is controlled digitally via the card's CA20k2 DSP/controller. Creative tuned the PCM1794A in a way that it has high DNR, but also higher distortion at higher volume. The Realtek ALC892 for example, has poorer results but at high volume, the degradation added by THD alone is lower than the soundcard's PCM1794A output, like this:You mean loopback test and there is a sweetspot at -10 to -20 dB output? That is very common, happens in the QA-400, 401 and RTX.6001, too.
Don't know, but mansr has a Tascam UH-7000 with PCM4220 and he made a comment like this:Yep, AK5394A outperforms it's THD+N from the datasheet in several implementations.
Do you know how PCM4220/4222 compare to CS5381? They seem to be closely matched. The DNR is better on PCM4220 but CS5381 might have lower distortion near 0 dBFS.
$300/€250 seems reasonable. Especially in a D30Pro+A30Pro deal, €500 both. Perhaps individually a little bit pricier then.Quite an attractive product! Hoping for a price of around 250€, though I am not sure if that's a realistic estimate.
Actually no. The sweet spot of THD+N (SINAD)is usually 0dBFS, especially when the harmonics are low(lower than -120db). It's the THD that has a sweet spot. But that's also often not the case now with these new chips and when implemented well.Well, if you go by Amir's measurements, the D30 Pro sits at -135 dB HD2 and -138 dB HD3 with a noise level of about -155 dB (after averaging), and the D90 is at -132 dB HD2 and HD3 each with about -159 dB noise (assuming the analyzer settings were the same).
I don't think you need the HD performance in audio, and it is questionable if this difference is really there (interaction between source and analyzer nonlinearities), but the D30 Pro seems to come out on top, and just as a measurement source, I would choose the D30 Pro.
Amir, I noticed you always do the 1 kHz FFT at full output. How about testing also at say -6 and -12 dB where many DACs have their sweet spot?
It's just not really true for D30pro. And pretty much all dacs with ESS chip.I can probably drag up a few examples where THD grows out of proportion on the last few dB before full scale, but you are right, this is not generally true. However, S/N is of concern mainly for studio use. For audiophile use, I am way more concerned about HD than I am about whether the noise floor is -130 or -135 dB. Same goes for lab use, optimizing DACs or amps. Pretty much every soundcard or analyzer I have played with (nothing beyond the RTX price wise, though) had a THD sweet spot below full scale.
It could be interesting to do this extra tests, as probably other as well, but Amir can't test everything. Now I believe that the reference should still be 0DBFS, and if THD up there is problematic, I'd rather see it than find the "sweet spot". As many other that have a mix of sources and use preamp and analog volume. I have always left my DACs unattenuated. I have nothing against digital attenuation, you either use it or you don't, but I don't think there are many users that would think it's OK to have to bring it down a bit of two as default because it doesn't perform optimally at full Dynamic range.I can probably drag up a few examples where THD grows out of proportion on the last few dB before full scale, but you are right, this is not generally true. However, S/N is of concern mainly for studio use. For audiophile use, I am way more concerned about HD than I am about whether the noise floor is -130 or -135 dB. Same goes for lab use, optimizing DACs or amps. Pretty much every soundcard or analyzer I have played with (nothing beyond the RTX price wise, though) had a THD sweet spot below full scale.