There is pure uncut insanity in this post.
The spectrum is always near in audio. In every corner, in every section but in some more than others.
Last edited:
Very nice and concise presentation, it will take a while to study everything you included!Here they are, all together:
DDDAC:
https://www.dddac.com/
The full experience is only with the tube I/V, that's what I'll deal with next.
My pet bat would attack your speakers each time it hears that thinking it is a giant moth. What about my new avatar, that is my real fat cat now pushing 22 yrs. It can't even scratch its bum!
You could summarize it that way. There are reasons for it, but the main takeaway is as you say. Also, ABX is usually more difficult than, say, A/B, which can also be done blind. So why insist on ABX instead?
In addition, good A/B software should allow the listener to locate a short segment to loop, without continuously having to push a button for each replay. Maybe only one-button push to toggle between A and B. During training having a simple way to blind or unblind to self-test while practicing. Some little things like that can go a long way to help with reducing fatigue.
Also, some time ago I proposed a sorting test. That's how I sorted differences in opamp distortion by ear (blind) in a non-inverting unity gain amplifier circuit (for PMA's hi res listening test). I found it to be a much more powerful test than ABX for distinguishing very small differences.
Last edited:
That can be if the output stage filter or other parts of system produce a "muddy" sound. In that case some added distortion can help cut though the mud, and thus add some needed (although false) clarity to the sound.
However, that's not the best solution. Its better to find what's causing the muddy sound in the first place and fix it. Some cables can add to it depending on how they interact with the electronics they connect to. Some output stages need reworking to sound less muddy/smeared/congested. Eventually you can get to a system with natural tones and textures, good black space between instruments, and an amazing sound stage. Just removing the output filter instead is more of a shortcut hack.
Last edited:
I forgot to mention that the echo test was actually a preference test, which are supposed to be more difficult than testing for a difference.
The null test that I would suggest is very simple and uses only an oscilloscope if you have one. Set up and match a 1 kHz signal on each DAC using both oscilloscope CH1 and CH2 trigger on CH1 and they should lock unless one of the DACs are drifting in time. When they are absolutely perfectly level matched, select Invert CH2, then the function CH1 - CH2, the result should be a straight line. Now change the input signal to White Noise and if the channels are the same nothing (straight line) then these two DAC are the same in every respect. Any difference will show up.
That is just it. There are many that sound the same. Some different sizes, different appearance and build quality. Even if they sound the same to humans no need to have only one DAC in the world. Not having those doubts is very nice.
So your contention is a broken design makes DACs sound different, and you prefer it. Obviously fidelity and transparency aren't your goals.
931410, you have it wrong, a tweaked DAC, or minimalist DAC. That thing must sound crass like fingernails on a black board.
🙁
🙁
The pre and post echoes that happen at 22,050 hz? And those bother, but all the ultrasonic energy from no filter does not?
No,.the ones that happen all over the audio band, see R. Lagadec and T. G. Stockham, "Dispersive models for A-to-D and D-to-A conversion systems", Audio Engineering Society preprint 2097, presented at the 75th convention, March 1984. If you have no access, you can also read appendix B of https://linearaudio.net/sites/linearaudio.net/files/03 Didden LA V13 mvdg.pdf
I am genuinely curious to learn why ultrasonic images should bother me. Provided of course that downstream electronics are amplifying them without going nuts. Is the main objection related to tweeter overload?
Again, my reasoning is simple: with the dac in question (1862) i can hardly tell the difference between redbook and 176kHz material. If it is upsampled from redbook the sound is a little worse, if native hires - the difference can go either way.
Again, my reasoning is simple: with the dac in question (1862) i can hardly tell the difference between redbook and 176kHz material. If it is upsampled from redbook the sound is a little worse, if native hires - the difference can go either way.
@analog_sa: I don’t think that they should bother you, provided that they don’t provoke misbehavior in the amplification chain which follows. As I believe Marcel had mentioned, the human ear functions as a natural biological reconstruction filter. In addition, as has also already been mentioned, most loudspeaker tweeters provide roll-off above 20kHz, providing further suppression of the image-bands.
Non-scientific (primarily because of the lack of control over the uniformity of the DACs utilized by test participants) listening tests conducted by a few of us awhile back, including Marcel, Hans Polak and myself, pointed to the interpolation filter as among the greatest suspect DAC functional blocks affecting perceptual quality of the reproduction.
Non-scientific (primarily because of the lack of control over the uniformity of the DACs utilized by test participants) listening tests conducted by a few of us awhile back, including Marcel, Hans Polak and myself, pointed to the interpolation filter as among the greatest suspect DAC functional blocks affecting perceptual quality of the reproduction.
Last edited:
The thread was indeed quite unscientific (besides what Ken mentioned, some of the listening tests were sighted, the blind tests had a quite small sample size and in one of the blind test the test leader posted information halfway that he should have kept secret until the test was over), but for what it's worth:
I have no idea. I've built a few of those NOS DACs, compared them to various factory ones, and I like what I'm hearing, so I'm continuing in that direction. This time with a tube I/V as I have seen from others who have done it that it gives a superior sound. I don't know if it's a real sound, and in principle I'm not interested, because it brings me joy, both during production and during listening.
Just don't state that they are filterless when there are transformers involved. I recall having tried a AD1860 based DAC long ago without anything except its internal IV opamp and it was awful. Just out of curiosity and at very low volume. It was already clear that the output stages of then were to be improved hence my experiments. Power supplies and output stage were of high influence on the final result.
Despite their pretty high price the Sharp DX-110H was one of the CD players where many improvements could be done as the outputs stage was mediocre.
Non OS and filterless are 2 separate items that you seem to mix up. And anything with inductive, capacitive and/or resistive properties combined is a filter. Just saying.
Despite their pretty high price the Sharp DX-110H was one of the CD players where many improvements could be done as the outputs stage was mediocre.
Non OS and filterless are 2 separate items that you seem to mix up. And anything with inductive, capacitive and/or resistive properties combined is a filter. Just saying.
Last edited:
It is not broken design. NOS sounds more transparent and detailed than a typical modern 1bit DAC. It sounds better and more natural. We compared a lot, I wouldn't do it if it wasn't superior, believe me. I'm not comparing with 5-10k eur flagship devices, but with an average DAC in the same price category.
At the moment I only have DDDAC with that transformer, the others the others don't have it (AD1862, AD1865 and PCM1702). They have no filter. I'm just finishing the mini AD1862, as you can see there are no others except the mains transformer.
I also tested the AD1860, that sounds great just a little softer than the others, very similar to PCM56. With external OPA and good power supply.
Attachments
Last edited:
- Status
- Not open for further replies.