Please note that Mr. Markw4 has a fixation that measurements are only steady-state FFT measurements. My advice about this is that measurements are the only way to objectively assess whether or not the implementation is successful. Listening is important but hearing is not reliable and the expectation bias is very high when judging something self-built. Also there is no real reason why a well-designed and measuring dac would not perform well in any aspect of human hearing.
I totally agree with your point of view, but I'm far from allowing myself a strong opinion since I don't have any experience in developing audio equipment. I don't have a Lab or room for validating, so I just have to believe in the resources I can get. That's why I tend to use something that is 1st well tested and used in this forum and 2nd has decent technical references to allow a novice like me a well build device.
My current plan is to use a 4 layer stackup, most certainly signal-GND-GND-signal. this is needed because the implementation of the DAC inst the only hurdle I will face in designing my USB DAC. I have watched quite a lot of the Altium design guides and did a hardware design course by Phil Salmony. The mixed signal part and USB will be quite challenging. The first board I ever designed has a QFN package and worked first try, not saying that I fully understand why it does... maybe I just got lucky, but I feel confident pushing the boundaries.
I've opted to put USB-to-I2S on a separate board with digital isolators for all I2S, I2C and IO signals. This way dac (and adc) boards do not have to deal with possible USB issues.
This is opening a can of worms as no technical reference can win from opinions
Think of the "it measures superb but sounds bad" syndrome.I think keeping to known design items (like to keep current loops small etc.) and just designing it with a littler hope and a review by an experienced tech and measuring it is an adequate way of jumping in the deep. It may take a revision here and there on the road to a working device. Relying on simulation software is popular but reality is reality.
By all means aim for the highest, your design should outperform most ready made stuff or else it'll be lost time and money.
This is true. By all means learn how to measure first. Just don't get hung up on it like that's all there is or you will chase around tiny amounts of harmonic distortion while not necessarily producing very good sound. Its quite possible to make a pretty poor sounding dac that measures pretty darn well. I've owned a few well known commercial ones, which among other things has been a costly way of learning.
One more thought: there is a thread on 'how we perceive nonlinear distortions.' Many things that affect reproduction system perceptual sound quality have been discussed along the way, some of them not necessarily what we usually measure. Some of the more recent posts have been in relation to dacs, and sigma-delta dac modulators in particular. You may even find some info on measuring certain characteristics that are not included in datasheets. You might also find that we aren't very good at measuring certain other things, and some of the reasons why that is. Enjoy... https://www.diyaudio.com/community/threads/how-we-perceive-non-linear-distortions.379592/
Yes. Even without ever listening to the dac you will have your markw4s claiming that it must sound bad since it measures well. I guess the reason for this is that these naysayers just can't design a dac that measures well.
Can you find a quote of me ever saying the bolded words? Just more of your sarcasm?
https://www.psychologytoday.com/us/blog/think-well/201206/think-sarcasm-is-funny-think-again
alright, so my question regarding collecting some information about different DACs: Has there already been an effort to bundle the knowledge about different ICs? And if not, is there interest in doing so? I only can assume that you guys have some years of forum knowledge and I think it could be a nice community project.
As your name was not in capital letter it does not directly refer to you but others with similar mindset.
These are the types of comments these people make: https://www.diyaudio.com/community/threads/ak4493-dac-board.358778/post-6310375
AFAIK there has not been any such effort. I also have not seen many DIY-designs using modern DACs in this forum. Most seem to be commercial ventures or tweaks to some existing boards.
So you were merely confused as to belief about cause and effect? If so, as I already explained good measurements do not imply great sound. By the same token good measurements don't imply bad sound either. Only a lack of strong correlation between perceptual SQ and measurements. Hopefully that helps to clarify for everyone.
Again I will agree with Mr. Bohrok
There is lot that is known and discussed about dac chip manufacturers and dac board designs, but some of it comes down to some pretty simple take-aways: The two major players today are ESS and AKM. TI hasn't done much new as of recent. Maybe they will jump back into it at some point. For ESS and AKM dacs, they tend to sound different. ESS has mostly gone for measurement numbers. AKM has a golden ear on staff and has mostly gone for SQ over measurements, at least until they had to show they could compete on measurements too. IMHO AK4499 was the first serious effort at that.
Regarding dac board designs, it gets more controversial there. Measurement verses SQ tends to favor one thing over another at least in fine balance, although both approaches can be quite technical and challenging.
Last edited:
That link to sarcasm only counts in the anglosaxon world where words seem to be more important than actions. Where words have to be replaced as there is always someone feeling offended for some reason that is often acknowledged. Fatiguing. It is just words and words can’t hurt. Feeling offended is a choice.
I just love sarcam and ultra black humor but only when there is a smile sold with it. Blunt but honest answers are good feedback to reduce the gain of your self appreciation.
I just love sarcam and ultra black humor but only when there is a smile sold with it. Blunt but honest answers are good feedback to reduce the gain of your self appreciation.
Last edited:
Oh, come on. We have gun violence too. That's action, right? Doesn't mean action is necessarily the most appropriate course.
Yeah guns are appropriate action that doesn’t hurt either Those words however…..
If I would feel offended by words I better leave technical fora. When I get flak big time I sometimes think:”wow that dude sure knows his stuff” and this regardless of the allimportant tone. We call it (translated) that my ears have been washed.
The king of off topic hit again. Sorry let’s go back to DAC chips.
Those words however…..If I would feel offended by words I better leave technical fora. When I get flak big time I sometimes think:”wow that dude sure knows his stuff” and this regardless of the allimportant tone. We call it (translated) that my ears have been washed.
The king of off topic hit again. Sorry let’s go back to DAC chips.
Last edited:
It isn't that words necessarily hurt. Its that some lurking readers may take false words too literally, which is not necessarily in their best interests. If I don't defend my attempts at accuracy and treat someone else's inaccurate words as equally valid, who is that helping in the big picture? The readers? I don't think so. False words are an attempt to discredit accuracy in forum posting. An attempt at Fake News, in other words. We don't want or need a ministry of truth however. How about members just try be accurate as possible rather than intentionally exaggerate? What's wrong with that?