watchnerd
Grand Contributor
I am holding on to my 20+ year old PCM63 DAC! I'll try to measure it one day, once I dig it out from storage
Original packaging, I hope!
-
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!
Delta-sigma vs “Multibit”: what’s the big deal?
- Thread starter THW
- Start date
Original packaging, I hope!
Esotechnik
Member
Discrete R2R DACs have termo-EMF performance leak.
Can someone test DS ADC as in figure 5.39?
https://www.analog.com/media/en/tra...ndbooks/Data-Conversion-Handbook/Chapter5.pdf
Delta-sigma converters spoil the statistical characteristics of a signal.
Attachments
I have no idea what "termo-EMF performance leak" means.
The histogram test is a standard data converter test and delta-sigma designs do just fine on it. What "statistical characteristics" are spoiled?
I tend to use ENOB as a measure of performance, which in turn is based upon SINAD. If you want to read about testing, you can look at references like IEEE Standards 1057 (Transient recorders), 1241 (ADCs), and 1658 (DACs).
The histogram test is a standard data converter test and delta-sigma designs do just fine on it. What "statistical characteristics" are spoiled?
I tend to use ENOB as a measure of performance, which in turn is based upon SINAD. If you want to read about testing, you can look at references like IEEE Standards 1057 (Transient recorders), 1241 (ADCs), and 1658 (DACs).
termo-EMF? I'm familiar with the temperature effect on a ladder, but what does EMF have to do with it? Well-designed discrete R2R DACs have built-in compensation for temperature and aging effects. Mine does.
Esotechnik
Member
Gaussian (normal) distribution codes for noise. Linear distribution for triangle.
What is tones 190 and 210 Hz?
vs. 20-year PCM1704:
The ones at -142.5 dB? In response to -100 dB output and below? Seriously? The spectral distribution of quantization (and other) noise is quite different for delta-sigma designs. You could just as easily say that, since the noise floor of the DX7s is much lower than the ML 360 between 192 and 196 Hz, that the DX7s is a much better design.
I still think we are being played here...
I still think we are being played here...
Agreed.
What are you talking about, I can hear a needle fall just fine while standing next to an F22 during take off.
Actually the Topping has better linearity. The peaks snapped to every 5dB grid closer than the Mark Levinson.
Only went to -135, but still...
https://www.audiosciencereview.com/...ts/single-tone-linearity_-umc404hd-png.13993/
https://www.audiosciencereview.com/...ts/single-tone-linearity_-umc404hd-png.13993/
Esotechnik
Member
At medium and high frequencies, the results should be worse.
If you set a goal to use the SAR ADC to identify all the D-S DACs errors (artefacts), then it is possible, but this website will cease to bring advertising to mainstream producers.
If you drive the "objectionable" R2R DAC into clipping, it will become worse than others according to the measurement results.
According to unverified information, scientific (non-audio) true 20-bit ADCs (1997yr) and 24 bit DACs are made on heat-conducting ceramic boards and placed in a thermostat.
https://www.baspi.ch/low-noise-high-resolution-dac
Ceramic also: https://news.thomasnet.com/fullstor...rk-is-designed-for-a-d-d-a-conversions-478945
Last edited:
I am not sure the relevance of this; the performance of virtually all data converters gets worse as frequency goes up, though the rise over the audio band is usually small.
Why would I want to compare a D-S to an SAR DAC? What are you going to compare the SAR to and how do you de-embed its own artifacts? Why not make measurements using any of the known standards and compare to ideal performance?
This website does not have mainstream manufacturing advertisements.
I have built ADCs and DACs on a lot of different substrates. Ceramic is a poor heat conductor, BTW; alumina might be better, and diamond substrates do real well in that department (seriously, using synthetic diamond material). But I still do not see the relation to this discussion.
I see no point in continuing this. Seems like a disinformation campaign, or just somebody getting his kicks from wasting other folks' time.
I love horses , but hobbyhorse's not so much.
@Esotechnik nobodies buying what your selling . Time to dismount.
If you have other areas of interest in audio I'd invite you to seek the appropriate threads out and leave this subject for the time being.
Cheers
@Esotechnik nobodies buying what your selling . Time to dismount.
If you have other areas of interest in audio I'd invite you to seek the appropriate threads out and leave this subject for the time being.
Cheers
solderdude
Grand Contributor
For frequencies < fs/2 the NOS filterless DAC is perfectly accurate (Ok, there is some attenuation towards fs/2 caused by the sample and hold).
All inaccuracies are in the folded spectra >fs/2. If the rest of the system does not make any new signals out of these in the <fs/2 range, and the ear perfectly ignores all of these folded spectra, a NOS dac could sound ok
Incorrect. It does NOT adhere to the sampling theorem at all. It needs a reconstruction filter for that to happen, soft/firm-ware and/or hardware.
Audible frequencies above 5kHz are amplitude modulated which is what causes the 'roll-off' its not similar to an RC roll-off.
A filterless NOS DAC can sound good with 96 and 192kHz files.
Just look at the actual waveforms not just at a sweeped frequency response with 44.1kHz files.
The ear ignores the ultrasonics just fine.
Some tweeters and amplifiers do not.
There is no "deal" apart from hype and snake oil. The Holo Audio May proved that multibit or R-2R DACs can indeed match or beat chip-based implementations... At the immense cost of $4000 USD.
How is this thread 16 pages long and growing anyway?
How is this thread 16 pages long and growing anyway?
Your (and my) comments don't make it any shorter...
Amplitude modulation?
Ok, I dont think so, I will do a simulation to see if you are right, but to my knowledge a 10Khz tone out of a NOS DAC will have a constant 10KHz component attenuated by -0.7 dB and a lot harmonics above Fs/2.
Do you have a reference to that amplitude modulation?
solderdude
Grand Contributor
Attachments
For sure that doesn't look nice... But... its all because of the >fs/2 harmonics...
Set the measuring bw to fs/2 and it will look much better ..
Similar threads
