good point, exactly how I "developed" the DDDAC1794. I started with datasheet application. then removed opamp reconstruction filter and used passive I/V (way better) and then switched from normal mode to non digital filter mode ( " NOS " ) - and yep I recognized the beloved NOS sound I knew from my DDDAC1543 series....
so yes, if you keep the DAC constant, you can nicely compare before and after DAC stuff
@ Baffless, @TNT,
You are both correct, of course, that the proper way to conduct our investigation is by ridged adherence to the scientific-method. Which, as you point out, involves changing only one variable at a time during when performing experiments. Leaving the rest of the system under test otherwise identical. However, it seems that we must accept the limitations imposed on us working as a group of internationally dispersed hobbyist. Many of whom do not have access to the necessary instruments, test unit DAC or other resources, to participate in physical experiments. As a result, it seems our conclusions will not be scientifically verified. Which is not the same thing as them necessarily being false. Should we identify a technical 'formula' of sorts, for delivering significantly more satisfying subjective digital playback, I certainly won't dismiss that result simply because it had not yet been scientifically verified.
You are both correct, of course, that the proper way to conduct our investigation is by ridged adherence to the scientific-method. Which, as you point out, involves changing only one variable at a time during when performing experiments. Leaving the rest of the system under test otherwise identical. However, it seems that we must accept the limitations imposed on us working as a group of internationally dispersed hobbyist. Many of whom do not have access to the necessary instruments, test unit DAC or other resources, to participate in physical experiments. As a result, it seems our conclusions will not be scientifically verified. Which is not the same thing as them necessarily being false. Should we identify a technical 'formula' of sorts, for delivering significantly more satisfying subjective digital playback, I certainly won't dismiss that result simply because it had not yet been scientifically verified.
That may provide interesting subjective results, as long as the tweeters are left disconnected for BOTH NOS and OS. In my experience, the NOS sound is still apparent via vocals, and lower range instruments. So, the lack of frequencies over 5KHz may, importantly, still reveal the key NOS/OS sound differences which we've been investigating.
I am also interested in how much and what kind of differences exist between various digital filter implementations. Let's just compare 4xOS. I heard PMD100 is better than SAA7220B/P, and the number of filter coefficients is a differentiator. I imagine there must be filters that give better reconstruction than others.
We will very soon arrive at a discussion on, what effect different implementations of digital interpolation-filters may have on the sound. We aren't there quite yet.
I use the Chord M Scaler filter (build in the Chord Blu transport) and it is amazing !
Only not DIY unfortunately (yet).
Yes, that's the idea! I believe I could have first impressions in a few days. The DAC is PCM1794 i.e. delta-sigma however, regarding frequency response it measures typically NOS.
dddac reports in the affirmative. We will eventually test the conclusions found here against SDM operation as well. dddac’s SDM experience will prove highly valuable.
Aaaah power supply tweaks are so so important even when I have very good stuff that are very well designed, it still amazes me that so much more can be had through tweaking & tuning of PS
Moons ago when this NOS craze started I did try out Sin X by Thorsten can't say I like what I heard, could be components that I used but perhaps Zanden's implementation is the way to go.
Moons ago when this NOS craze started I did try out Sin X by Thorsten can't say I like what I heard, could be components that I used but perhaps Zanden's implementation is the way to go.
some time ago i build a DAC for a friend of mine (DIY'er himself with limited knowledge on electronics) and he stated when using batteries on this thing it would be even better.
So i started tweeking as i do not like batteries ! I came up with this floating PSU and it has now almost equaled the battery driven unit.
So more tweeking is in order...
I tried many time to decribe the sound of it as this is almost impossible as it is just "not there" similar to a live broadcast of an event.
Some people have their TV connected to it and state that the newsreader is now "in the room", it is almost scary... (even with sports like tennis it is almost surround sound out of 2 speakers). I do not know as i have no TV connected to my audio system, but it is nice to get this kind of feedback.
So i started tweeking as i do not like batteries ! I came up with this floating PSU and it has now almost equaled the battery driven unit.
So more tweeking is in order...
I tried many time to decribe the sound of it as this is almost impossible as it is just "not there" similar to a live broadcast of an event.
Some people have their TV connected to it and state that the newsreader is now "in the room", it is almost scary... (even with sports like tennis it is almost surround sound out of 2 speakers). I do not know as i have no TV connected to my audio system, but it is nice to get this kind of feedback.
Last edited:
MagicBus has suggested an relatively easy to perform experiment regarding the presence of image-bands. He is disconnecting the tweeters of his loudspeakers, and listening for whether or not OS and NOS then sound identicle. I think this is a worthwhile experiment, because, in my experience, the NOS sound character is also apparent with sounds well under most tweeter crossover frequencies. If you have OS and NOS capable DAC(s), please try his suggestion and report your findings on the thread.