Except that they are not. It is very easy to get bad measurements from an APx555 if you don’t set everything up correctly. Amir has fallen foul of that on numerous occasions. Some people think he has done so on purpose
1 Like
If his measurements correlate well with another 3rd party measurement or with manufacturer’s own specs, then I use that to say ok, add to short list to demo.
How everyone uses 3rd party measurements varies.
Some use them in a sensible way.
Some put too much blind faith.
I don’t mind the measurements, I usually read them and I believe in them. Its the conclusion that is often drawn from the measurements that I don’t like
1 Like
Yes I enjoy ASR for doing his measurements and simply ignore the conclusions. It’s easy to do ! 
Likewise if I don’t have any other measurements I can compare with, then like anything in life, take it with a grain of salt !
If his measurements correlate with another set, this gives some confidence.
1 Like
One of the mistakes of the measurements there is to limit those to 20 kHz bandwidth. That means that DACs with poor reconstruction filters are not catched, but instead the APx555’s input ADC’s anti-alias filters are used to complete this task. Prime example of case where the measurement equipment affects the results.
This is why I measure things in a different way and DAC designers should too. How would they otherwise know how their reconstruction filters perform?
I sorta do that too, if I happen to find relevant measurements. If something is really bad in test bench on basic things, I look for something else.
2 Likes
By looking at ASR’s measurements beyond 20kHz? See most recent DAC below…
Sure it’s not going into MHz but anyone that argues that they would rather have less data points than what ASR provides is being extremely silly - especially if his particular measurements prove that what the manufacturer says ‘on the tin’ is true.
I completely ignore his subjective impressions and his conclusions when deciding what to demo. I might only read them for a chuckle.
I also won’t demo anything that measures badly if I can’t find another set of measurements to compare with. It doesn’t mean the product is bad ! People seem to jump to extreme conclusions. I’ll just wait till JA at Stereophile measures or HiFi News etc.
https://www.audiosciencereview.com/forum/index.php?threads/loxjie-d50-review-stereo-dac.19528/
For some measurements they have given up a little.
I measure primarily two bandwidths, 100 kHz and 5 MHz.
Here it is not said, but the latter graph is likely using the sharp filter. Reconstruction error of -106 dB means it’s reconstruction accuracy is about 17.5 bits. The results would be much worse with those slow roll-off filters.
I personally prefer digital filters where the error is at least below -192 dB (matching 32-bit PCM resolution). While the ones I primarily use have it below -240 dB.
1 Like
I would like to suggest Lampizator DACs if funds allow. Am stunned by the musicality of their Baltic 3 product.
Why not use iFi Zen DAC? Supports DSD256, Burr Brown 1793, pure USB and the price…
1 Like
… but… The Allo Revolution DAC is much, much better! I got it 2 weeks ago and I’m more then satisfied! Round about 300$. But you need an external power supply 5V.
Anybody try HQPlayer with emm labs dac?
Read a review about Matrix Elements X, and it has some explanation about how it internally up-sampled with ASMD7 for its own custom chip. How would that work with HQPlayer? Will it be similar to how many AK/ESS chips handles this (skips internal up-sampling and uses the HQPlayer up-sampled stream)?
Can you share linky ?
Here is the link to the review:
But I think I mixed things up in my head (it happens ). The Matrix DAC looks very much a “normal” sigma-delta DAC chip implementation (probably a good one though).
Hehe all good, happens to the best of us. That’s why I wanted to see link.
You made my eyebrows raise 
Multiple 3rd party measurements say it’s a well performing DAC and Headamp.
Are you still rockin the RME ADI-2 FS DAC?
Yes, running it on 4 LiFePo4 cells, streamer and opticalModule also on LiFePo4 cells, and then fiber, so all nicely isolated.
1 Like
Nobody else except HQPlayer can do ASDM7, period. Or has someone been stealing my IP?
1 Like
Never mind about that, my brain was on vacation.
Btw the new Holo Spring3 can be purchased with analogue preamp / volume control as a factory installed option.
Something not possible with Spring2 and Spring1.