I believe linear is supposed to reproduce signals under Fs almost perfectly with some well understood artifacts that may be way below hearing limits? Or maybe we’re still treading around that limit
I think more measurements need done on eliminating the series pass resistor and instead using a transistor following the 4v opamp vref output. This is the current mod I have settled upon and have used for the last 4 years, all of the later mods I've done have been inclusive of it. I've mentioned before that added capacitance seems to have an even bigger impact the lower the impedance of the supply. In my case I am using 4 different values of caps between electrolytics and ceramics for each vref including the 330uf very close to each shift register. Has anyone done any simulated transient loads to the vref before and after modding? Also how would output impedance of the ladder (what you are driving) affect the load to the vref? I would say there are a lot of factors that can explain a lot of the differences we are seeing.
Anyhow I have been working on my home lab as of late, acquired some interesting gear. I would like to spend some more time exploring and poking the vref on my setup. One of the main reasons why I added the extra 5 small caps per vref was because I wasn't satisfied with the vref transients and it was most evident at the LSB side of the ladder.
In my signature is also a phono stage with a Bies.de AD24QS with CS5381 ADC with discrete opamps. It has an asynchronous bit perfect USB connection fed back to the PC. This seems like a very good unit for testing the performance of the dam1021 when connected straight via balanced into the AD24QS when comparing various vref mods. Which simulation software do you all recommend for anyone wanting to get started with a simple loopback test?
My goal is to fully compare a stock dam1021v5.
Anyhow I have been working on my home lab as of late, acquired some interesting gear. I would like to spend some more time exploring and poking the vref on my setup. One of the main reasons why I added the extra 5 small caps per vref was because I wasn't satisfied with the vref transients and it was most evident at the LSB side of the ladder.
In my signature is also a phono stage with a Bies.de AD24QS with CS5381 ADC with discrete opamps. It has an asynchronous bit perfect USB connection fed back to the PC. This seems like a very good unit for testing the performance of the dam1021 when connected straight via balanced into the AD24QS when comparing various vref mods. Which simulation software do you all recommend for anyone wanting to get started with a simple loopback test?
My goal is to fully compare a stock dam1021v5.
Me too. Made a substantial difference.
What exactly do you mean? Simulation or measurement?
The signal being compared to has to be simulated in order to be measured, I guess that's what I meant. I use Linux full time so not entirely familiar with RMAA and the variety of windows based tools. The tests have to be well controlled and repeatable. I've seen others here using ltspice and whatnot to make/measure test .wav files.
Guys I’ve been religiously reading this dis idiom but still a bit confused about vref mid for newer v4 boards. I got Soekris point of view and completely agree with him that he designed an amazing finished product. However, I have an itch to try what people do here to decide what my ears tell me. Two questions (regarding v4/5 boards):
1. Is vref mod the one where you add a bunch of capacitors to the 16 shift registers or to the 4 supplies? Or both?
2. I see people say that adding too much will result in too much low end and lose of dynamics. Is it a consensus here? What max C I should add to v4 boards to be safe and try this mod?
Thank you, I know that you probably answered these questions a million times but could you do it 10^6+1
1. Is vref mod the one where you add a bunch of capacitors to the 16 shift registers or to the 4 supplies? Or both?
2. I see people say that adding too much will result in too much low end and lose of dynamics. Is it a consensus here? What max C I should add to v4 boards to be safe and try this mod?
Thank you, I know that you probably answered these questions a million times but could you do it 10^6+1
Guys I’ve been religiously reading this dis idiom but still a bit confused about vref mid for newer v4 boards. I got Soekris point of view and completely agree with him that he designed an amazing finished product. However, I have an itch to try what people do here to decide what my ears tell me. Two questions (regarding v4/5 boards):
1. Is vref mod the one where you add a bunch of capacitors to the 16 shift registers or to the 4 supplies? Or both?
2. I see people say that adding too much will result in too much low end and lose of dynamics. Is it a consensus here? What max C I should add to v4 boards to be safe and try this mod?
Thank you, I know that you probably answered these questions a million times but could you do it 10^6+1
1. They're connected. rev4/5 has vias beside the 100uF caps. Adding them there is the easiest, least risky, and nearly perfect. Spikestabber's implementation is better right at the shift register but it's much harder. No need to do that if you don't feel comfortable with it on a rev4. Get 0.6mm lead caps. I recommend as large as you can find. Try Nichicon for a start. 6.3V is plenty.
2. I don't believe that. Why would you think so?
That's Dimdim's mod for earlier revs. If you have rev.3+ there should be vias in the middle of the board which would make things easier. If you haven't made the purchase, consider dam1121 1921/1941
Btw, I have 3*1500uF UCC caps per rail (don't get UCC though you'd have to trim/thin 48 leads by hand...) and if only you can hear the dynamics right now... I don't know enough about opera but Pavarotti has never sounded better. Pretty much perfect. And I'm never biased.
Last edited:
I can confirm that the Dam is creating serious sleeping disorders, not only because I am tinkering around, but because it just sounds so good now. I didnt expect that just 2 weeks ago, I was thinking of selling it and move on to something more"musical". Not so anymore...
And I also can confirm the difference in software; I downloaded a trial version of Jriver and even though there is not much to see what it does, it sounds again much better than Foobar, in the way Ynmichael describes it above. Again smoother while still detailed and with a BIG soundstage. I would rank the upgrade with the software as quite a bit more significant than the Vref caps.
After adding those I still was not happy with the sound, although it was clearly better. This harshness was just in the way of enjoyment and is, at least in my case, completely gone. Amazing...Now the sound is what I was hoping for when I bought it. And yes, I became aware that I listen to much higher volumes than before, simply because nothing hurts anymore.
Try it when you have a DAM and you are not completely happy with it, even when you are running SPDIF. In the beginning I didnt even upsample and it had already 80% of the effect. And now I can only upsample to 172/192khz because my Waveio is still not up to higher freq.
Trying it cost nothing except some minutes. And some hours at night then for listening
BTW I added a Lifepo4 battery to the clock supply, again less artifacts and more relaxed. Not by much, but enough to keep it.
Again, thanks to Danny I can finally enjoy this fine dac now
So Soeren seems to have done a pretty good job to create a basis for us nerds.
And I also can confirm the difference in software; I downloaded a trial version of Jriver and even though there is not much to see what it does, it sounds again much better than Foobar, in the way Ynmichael describes it above. Again smoother while still detailed and with a BIG soundstage. I would rank the upgrade with the software as quite a bit more significant than the Vref caps.
After adding those I still was not happy with the sound, although it was clearly better. This harshness was just in the way of enjoyment and is, at least in my case, completely gone. Amazing...Now the sound is what I was hoping for when I bought it. And yes, I became aware that I listen to much higher volumes than before, simply because nothing hurts anymore.
Try it when you have a DAM and you are not completely happy with it, even when you are running SPDIF. In the beginning I didnt even upsample and it had already 80% of the effect. And now I can only upsample to 172/192khz because my Waveio is still not up to higher freq.
Trying it cost nothing except some minutes. And some hours at night then for listening
BTW I added a Lifepo4 battery to the clock supply, again less artifacts and more relaxed. Not by much, but enough to keep it.
Again, thanks to Danny I can finally enjoy this fine dac now
So Soeren seems to have done a pretty good job to create a basis for us nerds.
Comforting to know the world gets rediscovered. Again
It is interesting to compare several bit-perfect players in that particular state: when their output is bit perfect. Very, very strange that JRiver, at least imho, scores better than some of the "audiophile" stuff, yet being a general media player. It is important to pay attention to the particular JRiver release as for some reason some are lemons.
I wonder how the signal would have been captured at a quality that - according to you - is so much more enjoyable using the availible AD converters with their little processing power.
Shouldn't there be a way to get comparable quality in real time with little latency and without using a computer?
Shouldn't there be a way to get comparable quality in real time with little latency and without using a computer?
Try it when you have a DAM and you are not completely happy with it, even when you are running SPDIF. In the beginning I didnt even upsample and it had already 80% of the effect. And now I can only upsample to 172/192khz because my Waveio is still not up to higher freq.
Trying it cost nothing except some minutes. And some hours at night then for listening
BTW I added a Lifepo4 battery to the clock supply, again less artifacts and more relaxed. Not by much, but enough to keep it.
Again, thanks to Danny I can finally enjoy this fine dac now
So Soeren seems to have done a pretty good job to create a basis for us nerds.
Nice! But how does it work without upsampling? Thanks for the tip on 192Khz, it makes sense that there should still be an improvement and it’ll make testing more accessible to folks with SPDIF connections only.
Comforting to know the world gets rediscovered. Again
It is interesting to compare several bit-perfect players in that particular state: when their output is bit perfect. Very, very strange that JRiver, at least imho, scores better than some of the "audiophile" stuff, yet being a general media player. It is important to pay attention to the particular JRiver release as for some reason some are lemons.
They’re not bit perfect at all. Upsampling create artifacts. Just look at the THD plot I posted...
Captured? They’re just upsampled using nearly perfect super long filters. Theoretically, if you can have arbitrary long filters the upsampling should be perfect. I read somewhere that 1M taps is considered perfect enough, not sure how they justified it but I believe them. It’s hard to do without a computer because we have LX15 Spartan 6 onboard which limits the computation abilities. In comparison Chord Dave uses LX75 which as its name suggests is 5x more powerful. The top of the line is LX150 which costs more than $200 a piece. Again it would be nice if everything can be handled onboard, but there’s added cost and using too much power onboard (assuming the FPGAs won’t be that much more efficient) may not be worth it. You just need to find good software resamplers...
iZotope
Hi Michael, Yogi,
Nice you get good results with upsampling that resilts in long listening sessions
Now let's try iZotope.
Besides HQplayer and Jriver, there's also a new mediaplayer for windows: Audirvana.
It uses the iZotope 64bit SR
There's a trial download.
Hi Michael, Yogi,
Nice you get good results with upsampling that resilts in long listening sessions
Now let's try iZotope.
Besides HQplayer and Jriver, there's also a new mediaplayer for windows: Audirvana.
It uses the iZotope 64bit SR
There's a trial download.
Last edited: