Ah, an Oppo. IIRC it uses a pretty much textbook implementation for an ESS dac chip, and seems to have a not very good reputation for sound quality at least in its stock form. If that's the one you have, maybe not too surprising if coax and optical sound about the same, as DPLL_BANDWITH may be set too high in the ESS control registers . ESS recommends to set it to the lowest stable value, but the safe thing for dac manufacturers is to avoid in potential stability issues and live with lesser SQ.
The other thing of course is that your coax setup may be lousy and your optical as lucky as could be. Yet another possibility is that the coax implementation in the Oppo allows a ground loop to form through the coax shield, which then degrades performance to optical levels.
In any case, probably best not to generalize too much about optical vs coax from the one experiment.
Last edited:
Thank you very much for your good wishes.
Your post 32 :
I only speak for myself but having measured the jitter of properly implemented DAC using Toslink vs. Coax I have not seen much difference. Both will work.
According to your post above, you state that there are no audible differences between the two types of connections, so I'm not missing anything, despite the "prehistoric" coaxial cable I'm using.
Luckily my setup has no RF or EFM interference, so I'm happy with the coax connection. 👍 😊
Who was spreading such reputation and how did they determine that? Casual subjective auditioning?
They are not? Please shed some light on this.
Oppo knew they had SQ problems with the last CD player. None other than John Curl was called in as a consultant. Oppo chose not to implement any of his suggesting either citing too much cost involved, or else a tradeoff of better SQ at the expense of somewhat less favorable measurements. At least JC got to keep a free Oppo for his trouble.
I have made comparisons between digital and analog, and I perceive important differences. A vinyl played with a decent TT, an appropriate tone arm, coupled to an MC cartridge with its respective RIAA phono pre, a good tube amplifier and good speakers, gives me the "musical nirvana" I want. I do not need more.
It is very unlikely that you can have that experience, because I gather that you are a person inclined to digital technologies, and you probably won't even consider taking the test, either due to lack of opportunities/equipment or because of your training and/or prejudices. Believe me, the sonic difference that is perceived when comparing the same musical theme reproduced on vinyl vs, the same one reproduced by streaming ....... Wow, do you understand what the tonality of the instruments is? Nothing to do with the extension of the audible band, distortion, be it harmonic, due to intermodulation, etc.
It is ANOTHER THING. But those who do not listen to live music with instruments without amplification, can not have a point of reference.
You may have read some of these reviews of the Oppo Sonica, which have points for and against, (as always), both are from the year 2017, just launched the Oppo Sonica on the market.
Unfortunately, Oppo stopped making digital audio players, and turned to making smartphones. Many people considered it a huge loss, and all the owners of an Oppo product, we worried at the time about what would happen to technical support. Well, it's been five years since I've been using it and zero problems, so if any arise, the Oppo will go straight to the bin and I'll buy something new.
In The Absolute Sound review, they talk about the lack of a remote control, Oppo later incorporated it, as you can see the photo in my previous post. With that CR and the one from Prima Luna, I don't get up from my chair for hours, and when I have to do it to enjoy vinyl, the auto stop of my TT gives me the necessary time to decide when to get up and get up, because I have a delicate illness of my spinal column and they are about to operate on me, but that is my problem, I am only mentioning it. It's the only thing I value about streaming, the variety of themes instantly and the immediate access by exploring the application.
Well, I will not stray further from the subject, thank you for informing me of the deficiencies of my Oppo, but since I am not an expert in digital technologies, I take them "with tweezers", since I have read why the virtues of the EssSabre chip, and they seemed very reasonable to me, given the price of the product at the time of its launch.
https://www.theabsolutesound.com/articles/oppo-digital-sonica-network-dac
https://www.whathifi.com/oppo/sonica-dac/review
https://www.audiosciencereview.com/...ppo-sonica-usb-dac-and-streamer-review.11845/
I wonder why people are excited to defenestrate products so happily... 🤔
I wonder why people are excited to defenestrate products so happily... 🤔
Either I don't understand what the DPLL bandwidth is, or your statement is contradictory. If the DPLL bandwidth is the bandwidth of the sample-rate-ratio tracking loop of the ASRC of the DAC, then a larger bandwidth should reduce the jitter suppression and make it more sensitive to jitter differences between optical and coaxial interfaces - but still far less sensitive than a DAC that uses the recovered word clock directly, without an ASRC or very narrow PLL.
DPLL_BANDWIDTH is what you say, DPLL servo bandwidth. If set too tight it may lose lock, unless incoming jitter is low enough audio dropouts will occur. However, the tighter the bandwidth, the better the ASRC output sounds. It is less responsive to noise, so the correct filter coefficients are more likely to be chosen in real time.
As bandwidth is increased, DPLL is able to retain lock on more jittery sources, but SQ suffers. When set wider than needed for either coax or optical it sounds bad simply because it is wide and more susceptible to being agitated by noise. It can sound bad enough to mask otherwise audible differences between coax and optical.
Again, as one starts reducing the bandwidth, SQ improves until lock becomes unstable. Hence, ESS recommendation in the datasheet to set it to the lowest stable value.
As bandwidth is increased, DPLL is able to retain lock on more jittery sources, but SQ suffers. When set wider than needed for either coax or optical it sounds bad simply because it is wide and more susceptible to being agitated by noise. It can sound bad enough to mask otherwise audible differences between coax and optical.
Again, as one starts reducing the bandwidth, SQ improves until lock becomes unstable. Hence, ESS recommendation in the datasheet to set it to the lowest stable value.
What aspect of sound quality? Without the specifics, your post doesn't explain anything.
Its about things that don't show up very well in typical 'figure-of-merit' measurements. Those measurement are of 'steady-state' behavior because otherwise you can't tell what you're looking at very easily on an FFT, there would be way too many frequencies at once to try to make sense out of.
Also, distortion phenomena in a dac can be similar to the model we use for amplifiers, but it can also be much more dynamic in modern dacs due to complexity of delta-sigma modulation (and complexity of subsystems like ASRC). Similar to distortion, noise can be correlated the audio signal (distortion correlation with signal is 100%) Noise can be produced that sounds grainy, ugly, etc. It can sound very unlike the smooth while noise associated with resistors. Some of the above can be measured after a dac is built using special test signals, but measurement equipment may need to be connected somewhere in the middle of the output stage rather than at the user audio output jacks. Other things can only be practically measured by the dac chip manufacturer at this time (such as state-variable settling after transients), although external tests for some artifacts might be developed in the future.
As a result, the only thing that will make much sense to most people is to find a better dac and listen to it until you get used to how it sounds. Then go back to a lesser dac and you might be able to hear all the quickly changing garbage the brain usually tries to ignore and or notice all the information on a CD that is not being audibly reproduced by a lesser dac (perhaps due to dynamic noise-masking). That's pretty much all I can say about it. I won't attempt to tell you what you should buy or from whom, or if you should buy anything at all. I can't solve your problems nor answer questions in a simple way that don't have simple answers.
By the way, ESS has talked about some of the above, there are patents to help fix some of the problems, there is a chapter on 'non-idealities' in the book 'understanding delta-sigma data converters,' and so on. Information is scattered around but people always ignore it when links are given. Almost nobody bothers to seriously try to understand it all. Nothing I can do about that either.
Also, distortion phenomena in a dac can be similar to the model we use for amplifiers, but it can also be much more dynamic in modern dacs due to complexity of delta-sigma modulation (and complexity of subsystems like ASRC). Similar to distortion, noise can be correlated the audio signal (distortion correlation with signal is 100%) Noise can be produced that sounds grainy, ugly, etc. It can sound very unlike the smooth while noise associated with resistors. Some of the above can be measured after a dac is built using special test signals, but measurement equipment may need to be connected somewhere in the middle of the output stage rather than at the user audio output jacks. Other things can only be practically measured by the dac chip manufacturer at this time (such as state-variable settling after transients), although external tests for some artifacts might be developed in the future.
As a result, the only thing that will make much sense to most people is to find a better dac and listen to it until you get used to how it sounds. Then go back to a lesser dac and you might be able to hear all the quickly changing garbage the brain usually tries to ignore and or notice all the information on a CD that is not being audibly reproduced by a lesser dac (perhaps due to dynamic noise-masking). That's pretty much all I can say about it. I won't attempt to tell you what you should buy or from whom, or if you should buy anything at all. I can't solve your problems nor answer questions in a simple way that don't have simple answers.
By the way, ESS has talked about some of the above, there are patents to help fix some of the problems, there is a chapter on 'non-idealities' in the book 'understanding delta-sigma data converters,' and so on. Information is scattered around but people always ignore it when links are given. Almost nobody bothers to seriously try to understand it all. Nothing I can do about that either.
Last edited:
Ok, I'll bite. I assume you are referring to the 2nd edition of that book (by Pavan, Schreier and Temes). What specifically in that chapter relates to audible differences between Optical and Coax SPDIF or the sound quality of Oppo's DAC?
Go back and read what I said about optical and coax in oppo's dac. In that case its mostly the ASRC masking differences, not the modulator. Why would you think that every possible problem is found in one chapter of one book? Go back and read what I said about information on dac distortion and correlated-noise mechanisms being spread out in many places.
BTW, here's some of what AKM says about the new replacement for AK4499:
"In addition to adopting a completely separated digital/analog circuitry solution, the design was further refined based upon AKM's latest high-quality sound design techniques resulting in a perceived improvement of the ratio of signal to noise from all perspectives. The system reproduces all musical information, down to the most minute detail, to achieve a 'sound as if you were there' experience." https://www.akm.com/eu/en/about-us/news/2022/20220420-ak4499ex/
Any idea what they might be talking about there? Looks to me like its about less masking of low level details by signal-correlated noise. Doesn't make sense to me as being about anything other than that.
Steady state noise can be measured with an FFT (like its done on an AP), so people are fooled into thinking the noise stays that low when music is playing. They can't believe the manufacturer would not disclose if that were not the case. I keep trying to explain in the forum that manufacturers don't feel obligated to tell you everything they know. They tell you why you should buy their product instead of the competition's and they give you a little helpful information to use the product, that's all. You are more likely to find out there was a problem only after they have fixed the problem as a feature of a new product.
"In addition to adopting a completely separated digital/analog circuitry solution, the design was further refined based upon AKM's latest high-quality sound design techniques resulting in a perceived improvement of the ratio of signal to noise from all perspectives. The system reproduces all musical information, down to the most minute detail, to achieve a 'sound as if you were there' experience." https://www.akm.com/eu/en/about-us/news/2022/20220420-ak4499ex/
Any idea what they might be talking about there? Looks to me like its about less masking of low level details by signal-correlated noise. Doesn't make sense to me as being about anything other than that.
Steady state noise can be measured with an FFT (like its done on an AP), so people are fooled into thinking the noise stays that low when music is playing. They can't believe the manufacturer would not disclose if that were not the case. I keep trying to explain in the forum that manufacturers don't feel obligated to tell you everything they know. They tell you why you should buy their product instead of the competition's and they give you a little helpful information to use the product, that's all. You are more likely to find out there was a problem only after they have fixed the problem as a feature of a new product.
The reason why your "advice" is ignored is that instead of focusing on the issue you drag every possible bit of information you can find in the internet mostly falsely assuming it validates your views. Then you repeat that same stuff ad nauseam in every thread related to audibility.
bohrok, you ever listened to closely to a variety of dacs? You can't hear the dynamic noise masking of low level details? I have described what it sounds like in the dac in my laptop. It sounds like you are hearing everything there is to hear, with no loss of low level details. Is smooth noise like resistor noise that that's gone when the music stops playing.
Again, get a good dac, get used to how it sounds, then go back and listen to a cheap dac. If you can't hear the difference then maybe you are one of the people who's threshold of hearing for that is worse than average. Just because you are smart enough, and therefor likely above average in things like math, to become, say, an engineer, doesn't mean you have to be above average at everything in life.
Again, get a good dac, get used to how it sounds, then go back and listen to a cheap dac. If you can't hear the difference then maybe you are one of the people who's threshold of hearing for that is worse than average. Just because you are smart enough, and therefor likely above average in things like math, to become, say, an engineer, doesn't mean you have to be above average at everything in life.
Attach this link above, with Amirm's measurements and then there are comments from several Oppo Sonic users and seasoned technicians who give up all that talk of Mark.... and also, much more entertaining to read than what was proposed by the... ...
😴
https://www.audiosciencereview.com/...ppo-sonica-usb-dac-and-streamer-review.11845/
😴
https://www.audiosciencereview.com/...ppo-sonica-usb-dac-and-streamer-review.11845/