• WANTED: Happy members who like to discuss audio and other topics related to our interest. Desire to learn and share knowledge of science required. There are many reviews of audio hardware and expert members to help answer your questions. Click here to have your audio equipment measured for free!

A new Dac to improve sound quality from Khadas Tone Board 1

solderdude

Grand Contributor
Joined
Jul 21, 2018
Messages
15,891
Likes
35,912
Location
The Neitherlands
Just for fun play music really loud and then attenuate 80dB.
 

b4nt

Addicted to Fun and Learning
Joined
May 29, 2021
Messages
803
Likes
270
You meant 60 or 50?

If I attenuate 80dB from max 0 (to listen to a silent spinning fan), I may have to poison first birds and a meowing cat outside.

If I atttenuate 45dB from my usual quiet level, I still get a sound out. If I go up 20, it gets loud. That is a range of 65dB. Edit... I may push up for more than 20, but my neighbours will poison me.

The Mahler track I mentionned earlier has big level differences inside. Parts being so at 0,0003%, other less using CD to SACD precision.
 
Last edited:

b4nt

Addicted to Fun and Learning
Joined
May 29, 2021
Messages
803
Likes
270
That is the Mahler Finale track. Some would say "very poor mastering, not so nice in a kitchen or noisy car."

Mahlerfinale.PNG
 

b4nt

Addicted to Fun and Learning
Joined
May 29, 2021
Messages
803
Likes
270
Just for fun play music really loud and then attenuate 80dB.

Don't know exactly what your own expectations and your environment are.

I listened again to that Mahler track, from DSD source converted to PCM 24/384. Sounds great, but different, on both my D50s or Aune setup.

This is my desk PC and for the D50s, PC sitting at 50cm only of my ears. I first went for fanless Akasa Turing FX. But found my CPU and SSD get so hot. Added and tweaked two slow spinning fans. I hear no sound nor any breath from there.

AkasaCoolSilent.jpg


My streamer plus Aune player is 1m away from my ears. There I killed fan noises, killed also HDD noises (heads where there ticking and crackling).

I do not hear any pop not on speakers nor on headphones.

I did a lot so to kill noise sources around me. Over 6 years, I only got pops out of my Aune setup for short mains dropouts. Did the same for the D50s setup, killing all noise sources. Moved my 4G phone away from the D50s. The only remaining noises (ticks, pops, weirds) come here so from track contents.

I beleive I have so a quiet 60 to 80dB audible range as long I close my windows (the birds, the meowing cat outside)

That explains why you would believe swapping op-amps would change the sound signature.

Fit a poor quality op-amp in a recent DAC, signature will change. Or fit any of such 6dB valve buffer in your setup, signature shall change.

I am not sure swapping recent DACs will change anything. Maybe due to slight differences in the output buffers stages? Due to chipsets? Or maybe due to their internal clocks? But all show very low distortion and noise levels according to reviews here. This is why I wondered now what plots for -60dB in/out would look like.

Is the ESS hump noticeable? I can't know, the D50s doesn't feature that: https://www.audiosciencereview.com/forum/index.php?threads/ess-thd-‘hump’-investigation.5752/page-5

On the quad FR was simply extended in both extremes. This is not a FR response issue though.

I did it the other way, shortening my FR range.

On the Aune setup, I have a sub my neighbours don't appreciate. "Makes the walls move!" There I have also an extended detailed treble range, sometime tiring me after hours of listenings. That sounds very different compared to my D50s setup (the sub and treble add a lot there).

My Quad for the D50s is a nice Chineese clone of the 76's era model. Some swap op-amps, I made that choice for a different sound signature. Sounds great compared to the limited NX-50, and not like my Aune setup. Last days, I made some tests with frequencies at 17kHz and above, I ended, because I got headache... I could be sensitive to upper trebles the least.

The Aune, now aged, is driving a Yamaha A-S2100 with more extended and sensitive speakers.

The Aune is of the era where people swapped op-amps or DACs to get out a different signature. I bought the Aune because it was much cheaper and cause I couldn't test nor figure out how exactly a Hegel, a Schiit, a Mytek, Gustar, Audio-GD or a TEAC might sound different (recloking, clocking, output stages).

Aune is AK4495S based. I may at some point order a D90se ES9038Pro based, listen, then return it. But I fear I would get EMI noise with it, like I got some with the D50s before earthing and PSU/USB lines cleanups.

I found also that funy paper, pined by Amir: High Resolution Audio: Does It Matter?
https://www.audiosciencereview.com/forum/index.php?threads/high-resolution-audio-does-it-matter.11/

There is nothing magical about 16/44.1 specification of the CD. It is not like extensive listening tests and research were performed to pick these values. They were selected by Sony and Philips in the creation of CD to balance the recording capacity and fidelity. So at some level the one camp is defending an arbitrary set of numbers.
 

b4nt

Addicted to Fun and Learning
Joined
May 29, 2021
Messages
803
Likes
270
don't have those numbers. Currently enjoying a live Marshall Tucker band album. It's a live album so can expect all kinds of weird things but don't listen for it. Just the music/performance.

Are you an (experienced/practicing) audio electronics designer ?

I don't like lives, for all interuptions plus the many crowd noises you get with.

Marshall Tucker? I found this on Qobuz. Such tracks seem more close to modern pop tracks, with not a lot dynamic range. Adapted to kitchen, car and ear plugs listening.

Do not know how that band would rank compared to Mahler in this loudness war database: https://dr.loudness-war.info/
 

Attachments

  • MarshallTucker.PNG
    MarshallTucker.PNG
    1.3 MB · Views: 76

solderdude

Grand Contributor
Joined
Jul 21, 2018
Messages
15,891
Likes
35,912
Location
The Neitherlands
I don't care. Was just enjoying the music, nothing about recording quality. It was just an album I was listening to at that point in time. Nothing more.
Don't care/mind if it had, noise or other nasties in there. They don't bother me as they are in the recording and can't do anything about it.
I am not a 'numbers' guy nor do I care as long as they are not reaching my personal audible thresholds.
Fidelity is always determined by the recording (and transducers).
Good enough electronics is easy and cheap to make. It's the commerce around it that is the problem and beliefs people have.
 

b4nt

Addicted to Fun and Learning
Joined
May 29, 2021
Messages
803
Likes
270
That is one of test tracks I used, in CD quality, for the many details it contains.

Recent 44.1/16 records and audio can so also be very correct. A nice one, for tests from bass/sub up to trebles.



BazbazDisLe.PNG
 

b4nt

Addicted to Fun and Learning
Joined
May 29, 2021
Messages
803
Likes
270
I don't care. Was just enjoying the music, nothing about recording quality.
....
I am not a 'numbers' guy....
Fidelity is always determined by the recording (and transducers).

There was a missunderstanding. I do also listen to old tracks, of the 70's to 50's and less, some being in mono. Numbers get so important. For mono tracks, I would not need stereo :)

I was surprised listening to that band in this discussions (the FR, the recording quality), that's all.

Basically, I consider:
- my old mono tracks, they have no FR neither dynamics (the audio levels variations)
- stereo tracks up to the 60-70s, as they where recorded, often without FR and with low dynamics
- recent but poorly recorded or mastered tracks, with low dynamics (most the audio is at similar high levels)
- tracks like that of Bazbaz I mentionned as sample, with both detailed low and higher levels informations, fine for good CD quality
- tracks like that of Mahler, with low audio levels, and very high levels, that will render better if encoded in 24 bits (or DSD)
 

b4nt

Addicted to Fun and Learning
Joined
May 29, 2021
Messages
803
Likes
270
That explains why you would believe swapping op-amps would change the sound signature. :)

I have now swapped the Quad 76's era clone for a recent mosfet amp. Not a big difference (at low levels, low DHT). Maybe minimals in trebles, plus in attacks...

I'll now listen so next days, and if ever I get tired by trebles, I'll swap back to that nice and good Quad signature.

I've read this clone isn't an exact copy. Its power transistors could be faster. I've also noticed the PSU comes with a much biger capacitors bank.

The differences you describe could very well be because of the 1.2dB higher output level of the D30.

I know the Quad has a sensitivity of 0,5V. But I had to adapt the level to this.

And I still couldn't figure out how swapping my Aune, its DAC and output stages, could change the sound there.
 

Attachments

  • AQ.jpg
    AQ.jpg
    137.9 KB · Views: 80

b4nt

Addicted to Fun and Learning
Joined
May 29, 2021
Messages
803
Likes
270
I've found more details about why we got stucked in 44.1/16 bits as the best and enough for audio. One may first read the paper pined by Amir about this (commercial and technical considerations, with no objectif proofs, tests and confirmations).

44.1kHz accommodates both PAL and NTSC TV frames and line rates. One could so imagine analog TV broadcasts with digital audio in lines. Or VHS tapes formats with digital audio. In a lot of countries, analog TV and VHS is now old history.

16 bits comes from industrial interests and 80's technologies limitations, like chips, broadcast or storage capabilities. We could be at 14 bits only if Philips would have won that battle (they already had developped a 14 bits DAC and pushed for such a standard). Then came the 44.1/16 74min CD. Soon later, in those early eras, came 48kHz also, for digital studio recordings; I beleive this was still limited by hardware and storages capabilities (especially for studio multi-track, from 8 up to 128).

All other technical details of those vintage or early digital eras can be found via Wikipedia and Google.

20-20kHz is the audible range. But there is no objective reason to remain stucked at 44.1/16 bits, this is 0,22% distortion at -60dB, which is in audible range. Audiophiles will push for 0,01% over the audible and a huge dynamic range, this is probably the main driver for hi-res, where many current DACS and amps are close or could be able to meet this.

The Quad clone is good. A two board set costs approx €40, they shall feature 0,01% distortion up to 50W, where we mostly use amps at 10 to 30W.

In the 80's to early 90's, affordable disk drives where limited to some 20-80 mega bytes (the operating system plus a single audio track). Internet access was at 1200 to 9600 BAUD (then came the cable). Since, we have access to giga or tera bytes of storage space, and many have fiber or ADSL internet access, for streamings at high rates. Chips are since more performant. Why would wenow still keep only or mostly 44.1/16 bits?

One will have to consider if its environment and speakers or head sets will meet this high precision, or what this might cost. For audio in a noisy workshop, hi-res isn't needed. Neither would it be for phone plus 100-16k ear plugs, or for 10% speakers+amp distortion.

One could so still say mp3 or 44.1/16 0,22% distortion at -60dB is way enough for everyone. I found this, of 1981, an expensive audiophile recorder, which was sold for 0,01% distortion with a 89dB dynamic range:

 
Top Bottom