Any chance the new update will be out this year? Maybe giving us a incredibly Christmas present blush:
Next week according to Paul at the start of this threadā¦
1 Like
There are multiple ways of trying to cancel noise with replicated hardware.
To get rid of uncorrelated noise you can simply average the outputs, each doubling of hardware lowers uncorrelated noise by about 3dB.
To get rid of correlated noise you have, say, two signal sources out of phase and subtract them. e.g. balanced hardware.
The DS Sr uses both of these techniques multiple ways in the analog hardware.
Itās much harder to āfixā FPGAs since you donāt know how much noise is correlated from chip to chip and how much is random uncorrelated noiseā¦
Not quite at the start of the thread but:
Ted - Are you familiar with dithering techniques to reduce noise?
Thanks, Ted. I was only some 287 posts or so outā¦ 
Excitement overtook me.
1 Like
Not so much to reduce noise, but to lose less accuracy when narrowing the bit width. AFAIK you always add noise with dither, but itās (hopefully) noise that isnāt corelated to the signal. Dither can reduce quantization noise however (which is correlated to the signal and hence more noticeable.)
Also, at times, you can control the spectrum of the dither noise and put it in a less noticeable place (e.g. ultrasonic or at least away from 1kHz to 4kHz where our ears are the most sensitive.)
I try to have enough accuracy and precision that I donāt need dither anywhere in the PCM processing (I always keep more bits than the input has.)
You can argue about whether dither in DSD is the same: itās used to increase linearity and nuke birdies, limit cycles and other artifacts of sigma delta modulators.
2 Likes
I must have missed the explanation earlier but anyway, why is I2S input polarity switching desirable?
How would I determine which state to put it in when it is eventually available.
I thought this was just switching right and left channels. I could be wrong
I love the sound of the Direcstream DAC with whatever I feed it. I was curious does it recognize and decode the HDCD disks. They sound great but wondered if it recognized the subcode and processed it?
Thanks
No, it doesnāt decode HDCD.
But you can de-code using a ripping program like dbpoweramp.
When non PS Audio equipment uses I2S with an HDMI cable things may not be in the same positions or of the correct polarity. In particular the PS Audio hardware has always had the data signal and the L/R right signal inverted from the typical HDMI labeling. Thereās also the possibility of left and right channels being reversed on DSD over I2S.
If I were to add the new feature, people who use PS Audio I2S sources (or compatible sources, e.g. Renduā¦) wouldnāt have to use the new feature. People who use other sources may have to use the UI to invert the polarity of some of the I2S signals.
1 Like
I did not think it did. It does not seem to matter as it sounds fine anyway.
That would be kind of self defeating as The DMP source playing the disk undecoded still sounds better than playing it through a computer or server. I have owned DACS in the past and i think my impression was the biggest difference between an HDCD and a CD was the HDCD turned the light on the DAC! Sonically no big deal.
I am late to this party (over 300 posts). So, maybe this was covered earlier and sorry if it was already discussed, but are the updates for Junior and Senior coming out at the same time?
This is really something to look forward to.
Has any work been done on making the updates work reliably for everybody? All these ideas and improvements being put into the new mountain update arenāt worth much if the update process is still broken for some. Iām still on Huron.
I disagree with what you said literally, tho I agree that updates should be more reliable.
I canāt affect updates. I can affect the sound of the product. You are in essence telling me to go work on some non-DS thing and that the work Iāve done on Redcloud and this release āarenāt worth muchā.
Talk to Paul (not the universe) if you think his priorities of fixing the DMP, the P20, etc. and getting updates to work more reliably are wrong. But, for all I know he has someone working madly on getting updates to work better even as we speak.
FWIW Iām always looking for anything thatās flaky in the FPGA, but that doesnāt affect updates AFAIK (tho it might affect some having to do an extra boot after an update to get better sound quality.) Iāve added a few items that should make the FPGA more robust with bad states, but I donāt directly know if they will help the reboot for better SQ issue.
9 Likes
Yes, I know all thatā¦I was not suggesting your code is at the heart of the update difficulties. Paul has said in the past that Matt has been tasked with making the update process more reliable for everybody, and my question regarding work on that aspect was in that context.