The problem is that frequency tracking algorithm would need some sort of (fast) damping to avoid oscillating around a stable frequency. I asked for that several times but up to now Soeren ignored it.
Perhaps it now gets more impact as a professional declares it to be a audible problem.
The announced frequency adjustment in smaller steps will not solve this issue if the algorithm is not changed beyond of that.
Six years after Soren betrayed, that this -for a long time- criticized clock shifting due to the watching streaming movie (which is important for him).
But we are bought several DAC boards for high fidelity audio playback, not cr.p youtube watching.
Are you buy any expensive DAC board, which designed for -only or mostly- streaming video audio reproduction? No.
The correct solution would be to turn it off this "function", which totally unnecessary for audio playback.
The thing is that other, much older solutions manage to reclock and sync perfectly without much lag. In a studio you cannot have high latency, unless it is purely a mastering AD converter (like the Lavry Gold, which is famous for it).
For example, I have used various RME devices for 20 years that use their SteadyClock, which reliably locks unto an incoming clock and provides a stable output (although jitter performance isn't all that great).
Same for my Lynx Aurora16 (and their soundcards, which I used for a time but went back to RME for driver stability). It has their Sync-Lock functionality, which analyzes the incoming signal for 30 seconds or so and then provides a low-jitter permanently aligned output. With Sync-Lock deactivated the converter shows jitter-sidebands with an external clock. After it has settled they disappear.
For example, I have used various RME devices for 20 years that use their SteadyClock, which reliably locks unto an incoming clock and provides a stable output (although jitter performance isn't all that great).
Same for my Lynx Aurora16 (and their soundcards, which I used for a time but went back to RME for driver stability). It has their Sync-Lock functionality, which analyzes the incoming signal for 30 seconds or so and then provides a low-jitter permanently aligned output. With Sync-Lock deactivated the converter shows jitter-sidebands with an external clock. After it has settled they disappear.
Last edited:
As was I. The real issue is that it seems the designer himself never really actually tried a multiboard set up to verify if these boards would behave properly, rather, just said they would and put the burden of proof on the buyer. I bought them on trust, much like you did. Live and learn I suppose.
One problem with the DAM could be on how it measures the incoming frequency. On what I remember the MUC just counts the incoming periods for a certain time. This might just not be accurate enough in short time.
With providing enough computing resources for this task this might be done quite well see:
BG7TBL FA1 and FA2 frequency analyzers - Page 9
Modifications of the idea might be needed as the reference clock of the DAM gets adjusted thus is not stable.
With providing enough computing resources for this task this might be done quite well see:
BG7TBL FA1 and FA2 frequency analyzers - Page 9
Modifications of the idea might be needed as the reference clock of the DAM gets adjusted thus is not stable.
I also wonder how long after an adjust the Si reaches it jitter performance... as it is poked quite often, it is maybe in a stabilising mode for quite some duration...??
I could very well live with a bit reduced SQ in a acquiring phase - especially if the producers of classical recordings could have the good taste of playing back a few seconds of the venues acoustics so that one could get into the mode before the first bar - e.g. like in the real word when the conductor raises the pin - it all gets very quite and exciting - and then...
//
I could very well live with a bit reduced SQ in a acquiring phase - especially if the producers of classical recordings could have the good taste of playing back a few seconds of the venues acoustics so that one could get into the mode before the first bar - e.g. like in the real word when the conductor raises the pin - it all gets very quite and exciting - and then...
//
Here's a comparison of the linearity of the DAM1021 vs. the Lynx Aurora16 (studio multichannel converter utilizing the CS4398 DAC chip, until quite recently one of their flagship offerings and sporting 120dB SNR and -107 db THD+N).
I made these via DeltaWave, utilizing the one recording of the DAM I've got that was actually in sync, using a DCS 904 (discrete ring ADC, now sadly discontinued and beloved by the classical recording community in the 90s and 00s).
It clearly shows the superiority of the DAM1021 wrt linearity.
I made these via DeltaWave, utilizing the one recording of the DAM I've got that was actually in sync, using a DCS 904 (discrete ring ADC, now sadly discontinued and beloved by the classical recording community in the 90s and 00s).
It clearly shows the superiority of the DAM1021 wrt linearity.
Attachments
Last edited:
Well, the graph lacks guides ;-) and you wrote linearity without qualifying wich kind of linearity - I thought it was some funky x-axis frequency scale... no, this style I never saw before.
This from audiosciencereview:
I think the 1021 should not differ a lot...
But yes, this is good. And I cant mess it up with a filter I think... or?
//
This from audiosciencereview:
I think the 1021 should not differ a lot...
But yes, this is good. And I cant mess it up with a filter I think... or?
//
Interesting, thanks!
Mine is the best version availibile in a Rev5 board (dam1021-12, I think it's a mix of 0.01 and 0.02 resistors).
Signal goes from the DAM1021 single ended outputs into my own output amp/buffer/balancing stage and then back into the DCS904.
The sucessor to the Aurora, the Lynx Hilo measures better than the former has in my test, even though it uses the same converter chip:
Lynx Hilo DAC/ADC Review and Measurements (part 1: DAC) | Audio Science Review (ASR) Forum
But my DAM1021 measures far better in my test than any converter's measurements I have seen so far. Looks like it stays within 0.1 dB up to 21 bits (-124 dbB).
The DAM1021 is better at 24 bits than the Aurora16 at 20 bits and abberation at the highest levels is far lower (y-axis has 3 times the scale for the Aurora measurement). That's really something.
Mine is the best version availibile in a Rev5 board (dam1021-12, I think it's a mix of 0.01 and 0.02 resistors).
Signal goes from the DAM1021 single ended outputs into my own output amp/buffer/balancing stage and then back into the DCS904.
The sucessor to the Aurora, the Lynx Hilo measures better than the former has in my test, even though it uses the same converter chip:
Lynx Hilo DAC/ADC Review and Measurements (part 1: DAC) | Audio Science Review (ASR) Forum
But my DAM1021 measures far better in my test than any converter's measurements I have seen so far. Looks like it stays within 0.1 dB up to 21 bits (-124 dbB).
The DAM1021 is better at 24 bits than the Aurora16 at 20 bits and abberation at the highest levels is far lower (y-axis has 3 times the scale for the Aurora measurement). That's really something.