This DAC must sound truly awful. It doesn't have anywhere near enough electrolytic capacitors.
If the goal of sighted listening is to find preference, which is the most likely reason for listening evaluation, why would anybody else know what I prefer better than myself? Strange for you to even suggest that unless you are yourself easily led.
You prefer R2R while somebody else prefers something else. Neither is right nor wrong. Preferences are subjective.
Some good points above regarding sound preference. I would just add that sometimes sighted listening is used for discrimination purposes rather than for preference. For one example, professional mastering engineers use sighted listening for discrimination purposes all the time. However, they are experts and knowing how to do it for purposes such as "translation" in different distribution formats, while at the same time satisfying the requirements of the client.
If I would think I heard any difference at all between two reasonably good DACs and if it could not be attributed to intersample overshoots, I would like to know if there was really an audible difference or if I was just fooling myself. Hence, I would try to find a practical way to test it blindly.
Anything better than 30 euro DACs where the entire reconstruction filter is left out and all coupling capacitors are class 2 ceramic to make it as cheap as possible.
Wrong!! you need to read more carefully. I said
"The ones in this thread that "think" R2R is old technology and that's bettered by delta Sigma technology because it "measures" far better."
Cheers George
Now that does seem strange. Why would your senses wait until you are listening to an audio device to decide now would be a good time to fool you?
iiuc the op mainly claims that his dac is better "architecturally", not the objective performance or subjective sq, hence no measurements whatsoever
dac with good architecture does not necessarily mean good implementation or performance or good sound
so without any numbers to backup this product is not for me as objective performance is my top priority
so you know what you want?
dac with good architecture does not necessarily mean good implementation or performance or good sound
so without any numbers to backup this product is not for me as objective performance is my top priority
so you know what you want?
Better believe it, our Sydney Opera House is an architectural marvel, but the main Concert Hall sounds like s**t (no power from the SSO, even the power of the 10000 pipe mechanical organ gets lost ) no matter where you sit, I've tried them all. It's an acoustically dead room, with zero sound reinforcement from it. I get cleaner and far more top to bottom range and far more power from my system at home.
Even with the "newest" BS sound aids to fix it, the dozen or so huge dreaded domed sound (upper mids and high only) reflecting "rose petals", they say if you can see a part of the orchestra in each of the "rose petals" it lifts the level of those parts of the orchestra, but does nothing for the low and mid bass they just disappear up into the pointed loft. https://media.365project.org/1/9325188_foquvwx023_l.jpg
Cheers George
Last edited:
How sad... and such a strange seating. I ended up in the choir seats behind the orchestra once - it was awful. I now only buy 10th-ish row center i front of the stage.
//
//
Yeah the old Sydney Town Hall built in 1869 concert room just up the road sounds sooo much better.
https://www.andrerieu.com/site/assets/files/3696/sydney_concert_-_4.jpg
And the organ they say is also good but not as big? (I haven't hear it yet)
https://upload.wikimedia.org/wikipedia/commons/4/46/Sydney_Town_Hall_Grand_Organ.jpg
Cheers George
Last edited:
OK but why the money value has to be involved always?
This is no-mcu/cpld DAC.
And this example is what i was pointing at.
(Not listened/build but i beliece that sounds verey, very good.)
Some people PM me to ask if they can build a MarcelvdG RTZ dac if they have a $200-$300 budget. Then I have to explain to each person individually what the options are. Then to put things into perspective, I would like to point out there are costly serious commercial dacs that are made by people generally regarded to be non-purveyors of snake oil products. It is such commercial dacs we would like to compete with. For some people, they will have to decide if they want to save up for something they would like to have, or if they want to build incrementally, or if they would rather just go for a Topping or SMSL. There are also some other popular dac board projects in the forum to consider. In these considerations cost is often an issue.
Actually, there is a USB board, and there is an FPGA board for PCM2DSD. Those things can make a fair bit of electrical noise and of course some jitter. So, we just have to deal with those things somehow. If we wanted to consolidate everything onto one board we probably could and still get good results. Cestrian has already consolidated the clock board, the PCM2DSD, and the I2SoverUSB board into one compact composite module. We could also probably put all the power supplies on the same board too. We might need to use 6 or 8-layers, and put the noisy stuff on the bottom side of the board, or otherwise as far away as we can get it from the sensitive stuff. We might also need to put a shielding can over some of the circuitry. But I think it could probably be done pretty well if mass production was being considered. OTOH, a less rigorously designed 2 or 4-layer dac board with MCU/CPLD might have some SQ problems.
Actually, there is a USB board, and there is an FPGA board for PCM2DSD. Those things can make a fair bit of electrical noise and of course some jitter. So, we just have to deal with those things somehow. If we wanted to consolidate everything onto one board we probably could and still get good results. Cestrian has already consolidated the clock board, the PCM2DSD, and the I2SoverUSB board into one compact composite module. We could also probably put all the power supplies on the same board too. We might need to use 6 or 8-layers, and put the noisy stuff on the bottom side of the board, or otherwise as far away as we can get it from the sensitive stuff. We might also need to put a shielding can over some of the circuitry. But I think it could probably be done pretty well if mass production was being considered. OTOH, a less rigorously designed 2 or 4-layer dac board with MCU/CPLD might have some SQ problems.
Last edited:
Yes but these things are from pure Digital domain. They are necessary and far away from point of the conversion in another module of DA...
But the DAC is from DA segment. And it is not the same if we have cpld/fpga just 3 points before the conversion happening. I gave my notion based on listening tests and measurements confirmed higher amout of HF noise in the audio BW.
.
You can make with glue logic I2S to DSD format too. Avoiding programmable, high density, high energy, small die units, which are actuay very good - for some other purposes...
.
Another thing, most of R2R IC dacs has thermometer code first 3 to 4 MSBits, as a segment before R2R ladder network for other LSBs. THis is 7 and 15 bitts to have non-R2R conversion type - but DSD type with just one R value
Not very practical in this case. There is maybe 32-bit or 56-bit multiple stages upsampling with particular interpolation filter coefficients, then dithered modulation from PCM to DSD256.
So we have to learn how to deal with it. Just because one person tries it once on a 4-layer board and it turns out to be noisy is not proof of impossibility. Even in the worst case a motherboard and one or more daughter boards could be used to make it work. Cestrian already showed his reclocker motherboard which also holds the USB and FPGA boards works great without audible problems. Of course, it was designed to make the system very likely to would work as intended. Also it took more than one iteration to reach the goal.
Last edited: