Probably you're right, tambaqui was designed in 2013, before introduction of sabre PRO series:
https://www.audiosciencereview.com/forum/index.php?threads/mola-mola-tambaqui-dac-and-streamer-review.10770/post-300699
But Is it still the case in PRO sabre series? According to manufacturer it is not:
https://www.diyaudio.com/community/threads/ess-sabre-reference-dac-8-channel.117238/post-6491643
Internals of the x sabre 3. Very similar to x sabre pro, unfortunately clocks for FPGA are no longer top crystek 957s - synchronous mode probably measures worse.
@MatrixAudio,
Do you have plans to release based on the es9039pro x-sabre 4?
Pure dac, with optional external high quality streamer like element s, would be great.
For now it doesn't look that es9039pro is a better DAC, it looks like low-power version of es9038pro. From available measurements it can be noticed that digital filtering is improved compared to es9038pro.
Maybe new modulator in es9039pro somehow makes "sound" better? Time will tell.
Hi Miska,
Do you mean that if I we provide to the ess9038pro externally upsampled data to 44.1kHz x 16/48kHz x 16 we will automatically ommit the first stage of oversampling in the chip?
In well-implemented DAC digital volume control measurements should look like that:
https://www.audiosciencereview.com/forum/index.php?threads/topping-d90se-review-balanced-dac.24235/post-1031608
The harmonics are gone in the noise.
I would say that universal values for -20dB attenuator are: 2 x 2k and 450ohm. Lower impedance than most of potentiometers and high enough not to stress (much) analogue part of the DAC.
Here is the online calculator:
http://www.sengpielaudio.com/calculator-voltagedivider.htm
I mean fully passive attenuator, e.g. -20dB made from low value resistors not to increase noise of the DAC. In theory it should not worsten the SNR/SINAD and solve the issue of gain mismatch in many systems based on DAC connected directly to a power amplifier.
E.g. ready made one, jts ma-123...