If you wonder why I asked the question you have missed the point about FIFOs, low jitter clocks etc.
Yes, a cd player has microcontroller running at 20Mhz or so - this is the same as a PC or Laptop running at 1.6Ghz is it?
I was wondering whether a very simply micro pc, with an output for a small monitor + USB 3 for the audio would not be preferable to a standard PC or Laptop. (I'm not into RPis, Arduino)
I have a Gigabyte Brix & a 10" lcd monitor lying about so maybe I will try these and compare with my Samsung laptop.
Last edited:
So long as the dac is external to the computer, sound quality can be excellent. The electrical environment inside a computer is a very different matter. It is possible for a very carefully designed dac to do pretty well there, but probably not a state of the art, world class dac.
As far as CD players, none that I am aware of have a great dac, although it could be done. It is just that starting prices for the very best dacs is up around $1,000 per channel, $2,000 for a stereo dac. And that is bargain price for a really good dac. One could spend more and get less if not careful.
Unfortunately, that is where we are right now with dac technology. There are lots of cheap compromise dacs out there which are built to a price price point. It still isn't cheap or easy to make a good one.
That's why I have been trying to help folks mod low cost Chinese dacs to improve sound quality. We aren't up in the $1,000 per channel sound quality range yet, but we keep getting closer. Probably above the $500 per channel sound quality level without too much trouble. The downside is that it takes a lot of diy work to get to that level of sound quality. Most people will never spend the money or do the work, either one, to get a really pretty good dac.
As far as CD players, none that I am aware of have a great dac, although it could be done. It is just that starting prices for the very best dacs is up around $1,000 per channel, $2,000 for a stereo dac. And that is bargain price for a really good dac. One could spend more and get less if not careful.
Unfortunately, that is where we are right now with dac technology. There are lots of cheap compromise dacs out there which are built to a price price point. It still isn't cheap or easy to make a good one.
That's why I have been trying to help folks mod low cost Chinese dacs to improve sound quality. We aren't up in the $1,000 per channel sound quality range yet, but we keep getting closer. Probably above the $500 per channel sound quality level without too much trouble. The downside is that it takes a lot of diy work to get to that level of sound quality. Most people will never spend the money or do the work, either one, to get a really pretty good dac.
Last edited:
Not quite the same, but do you have comparative measures of noise and jitter?
Your argument about low power CPU at lower clock frequency (and with noise suppression to support mobile phone environments) would argue for a RPi. so what is your issue with it?
We start with a minimal $39 Chinese ES9038Q2M dac board. We add an output stage board, an AVCC board, several voltage regulators, a new clock, additional firmware, and an external AK4137 board. A steel case is recommended for EMI/RFI immunity.
In short, very little is kept as original.
It is a lot of work, and it needs to be done in a very particular way, but it works. And it sounds very, very good if all the mods are done as recommended.
Last edited:
Indeed, that is a lot of mods. Have you done "before" and "after" measurements / tests?
Too old too try getting my head around a RPi, otherwise nothing against them at all.
Looks like I need one of IanCanadas FIFO II boards with one I2S input for my CD transport and the other for a USB - I2S converter.
Some, but not equipped to do comprehensive testing. Some of the most useful testing has consisted of listening tests quickly switching between the modded dac and a Benchmark DAC-3 (on the Stereophile A+ list, and state of art measurements). In order to clearly hear the differences, an extremely low distortion power amp is used. While those things may seem offensively expensive to some, they are no more so than a middling AP test set, and for me more of a necessity.
The modded dac does not sound as good as DAC-3, and that would not be expected. DAC-3 HGC is a $2,200 dac (and a bargain according to Stereophile). We do not claim that level of performance. However, we do know how close we are coming to that level of sound quality, and it probably right about where I said, more than $500 per channel sound quality. Maybe more around $700 per channel. Hard to say precisely, but I can assure you it is very good. It easily bettered Allo Katana v1.1. Have to see about Katana 1.2 if they decide to go ahead and send me one for review.
I just looked at the DAC specs and it looks like the spdif and I2s signals are resampled by the DAC (another chip, not the ak4490). If that's the case I'm not sure how much of an impact would Ian's boards have on the quality you get.
I was going to suggest trying the Allo digione which works with an Rpi and gives you low jitter spdif output.
Sighted, blind or double-blind?
As an enginer and designer, I have always found the most important part of design work to be the objective verification of my designs - just like regression testing is one of the most important tools in software development.
All of the above, at different times of course. I strongly believe in double blind testing, although not necessarily the Foobar ABX way.
Also use other listeners as they are available. Good to get multiple and various impressions and perceptions.
It is a very good recommendation, thank you. I will also consider Allo digione signature, which claims has a better design. I think another external spdif receiver could be 502dac.
Hi Ian,
I am very interested in your McFifo and McDualXO for some multichannel applications. However I would like to use external DACs and I need to convert the I2S to SPDIF. You wouldn't happen to be working on a multichannel SPDIF output board? If not are you aware of one that would work with your boards?
Thanks
I am very interested in your McFifo and McDualXO for some multichannel applications. However I would like to use external DACs and I need to convert the I2S to SPDIF. You wouldn't happen to be working on a multichannel SPDIF output board? If not are you aware of one that would work with your boards?
Thanks
Hi Ian,
I was reading through the manual of McDualXO; it supports a wide range of MCLK * Fs values; 128 * Fs, 256 * Fs, ... up to 4096 * Fs.
I am wondering whether it is possible to lock the output of the McDualXO to 128 * Fs?
The use case is sync operation of McDualXO with an ESS Dac so called “128 Fs mode” where DPLL is off and DAC runs fully in sync with the upcoming signal
Thanks!
I was reading through the manual of McDualXO; it supports a wide range of MCLK * Fs values; 128 * Fs, 256 * Fs, ... up to 4096 * Fs.
I am wondering whether it is possible to lock the output of the McDualXO to 128 * Fs?
The use case is sync operation of McDualXO with an ESS Dac so called “128 Fs mode” where DPLL is off and DAC runs fully in sync with the upcoming signal
Thanks!