At least you get white smoke.
I get nothing. After five days they just return the item. No phone call.
Hi Jens,
I got he kit too a couple of days ago and I just did the mods.
Well it has been a challenge for me as those parts are sooo small so it took me a couple of hours to get them in place !
Anyway, it seems that everything works well on a quick and dirty test (one channel loop, on XLR, 192ks, stepped sinus, switches 1V/1V, -4dB)
JC
I got he kit too a couple of days ago and I just did the mods.
Well it has been a challenge for me as those parts are sooo small so it took me a couple of hours to get them in place !
Anyway, it seems that everything works well on a quick and dirty test (one channel loop, on XLR, 192ks, stepped sinus, switches 1V/1V, -4dB)
JC
I experienced a nasty short distortion when I listened music with the RTX6001. To make it short, it turned out that the piece contains a negative FS sample which causes the problem (with the RTX only).
Attached a loopback recording (upper tracks: out 0dBV, lower tracks: in 10dbV) of the area in question - see timeline 0.3385.
I assume it is some kind of overflow in the digital data processing, most likely due to the fact that I2S uses 2-complement numbers and the computer 1-complement integers.
The most likely culprit, I think, is the RTX USB receiver software, otherwise it would be in the DAC chip itself.
I also attached the .wav sound file clip in the zip.
Attached a loopback recording (upper tracks: out 0dBV, lower tracks: in 10dbV) of the area in question - see timeline 0.3385.
I assume it is some kind of overflow in the digital data processing, most likely due to the fact that I2S uses 2-complement numbers and the computer 1-complement integers.
The most likely culprit, I think, is the RTX USB receiver software, otherwise it would be in the DAC chip itself.
I also attached the .wav sound file clip in the zip.
I did not say one should suppress errors. I said (or wanted to do so) the RTX produces a wrong output (i.e. something that is not in the input data) in this special case.
True, my error, yes the data is also 2-complement - so that guess for the source of the problem was wrong.
However the problem exists.
Other guess - you inverted the polarity - in two complement the magnitude of the maximal negative number is one more than the maximal positive number. Perhaps here something went wrong?
Last edited:
I did. It happens as well with Decibel as with Audacity.
Moreover it does not happen with an other DAC.
Overflow like that happens when the digital filter peaks an incoming waveform. its not common. Usually the system has been adjusted so worst case peaking doesn't cause digital overflow. I have not seen that on an AKM DAC but there is always a first time.
I don't think there is any digital filtering or level adjust in the XMOS interface.
Several things to check.
I don't think there is any digital filtering or level adjust in the XMOS interface.
Several things to check.
Good practice should be saturation/clipping to +-FS in that case. An overflow to around zero, I would consider as a design flaw of the filter implementation.