Some DACs have higher output on DSD, notably some AKM DACs. And higher is often perceived as better.
-
Welcome to ASR. 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!
Modern Multi-Bit DAC vs Delta Sigma, specifically AKM's newest flagship, but also others
- Thread starter mike7877
- Start date
Let’s not confuse pre ringing and pre echo. pre ringing is the mathematical consequence of a phase linear sharply band limiting filter. And yes, that only affects signals with energy around the band edge frequency.
Pre echo is the mathematical consequence of periodic pass band ripple, ie a pre or post echo causes picket fence ripples in the frequency domain. These periodic ripples is the result of using Parks MCClellan algorithm for equi ripppe filter design.
A seemingly benign 0.1 dB pass band ripple translates into a -40dB pre and post echoes separated by the length of the FIR filter. the echos can be seen more easily if you filter and downsample the impulse response.
These echos affect all signals irrespective of frequency content. This is not MQA mumbo jumbo.
I think it mostly concerns the question about the ringing. correct me if I am wrong.
Audiophile Myth #260: The Detestable Digital Filter Ringing and Real Music...
A blog for audiophiles about more objective topics. Measurements of audio gear. Reasonable, realistic, no snakeoil assessment of sound, and equipment.archimago.blogspot.com
true. Sony did very hard attempts to do the full chain keeping it in DSD. however, they ended up having to remodulate for each step.
One example where it made sense is the digitisation of master tapes for transport without the tape to the cutting machinery. However, the audience felt betrayed (not mentioning any names or brands here)
Last edited:
To the contrary, there are many: all SA-CDs mastered and authored with the very first Sony/Sonic Solution DSD stereo editing system or with multichannel-capable digital audio workstation (DAW) implementing Sony Oxford D-MAP processors, such a the Sony Sonoma, latter passed on Super Audio Center, Genex Mix+, Sonic Solution DSD.1 and DSD.x, and Sadie Series 5, have been passed through digital signal processing without change of the DSD sample rate, for reason thoroughly explained by the Sony Oxford's development team in a paper presented at the AES in 2001.
According to an insider who has worked for a joint Philips/Sony SA-CD production support office at the beginning of the format to keep track of all SA-CD titles amongst other tasks, the majority of SA-CDs put on the market by mid-2014 (about
Last edited:
I don't think that MoFi SACD's were mastered with any intermediate PCM processing.
The MoFi SACD for Van Halen II was critized by some for not fixing some tape issues, maybe so so done tape splices from the era, or tapes being in less than optimal condition (It sounds terrífic to me to me, and I know all Van Halen very well).
Could this have been on the DSD domain without converting to PCM and then back to DSD. I don't know.
But one things is sure, MoFi on Analogue Productions when they máster an analogue tape sourced album for SACD, they keep all the process on the DSD domain, straight Analogue to DSD.
The MoFi SACD for Van Halen II was critized by some for not fixing some tape issues, maybe so so done tape splices from the era, or tapes being in less than optimal condition (It sounds terrífic to me to me, and I know all Van Halen very well).
Could this have been on the DSD domain without converting to PCM and then back to DSD. I don't know.
But one things is sure, MoFi on Analogue Productions when they máster an analogue tape sourced album for SACD, they keep all the process on the DSD domain, straight Analogue to DSD.
glad to hear it was used widely by Sony. I had many AES convention chats with Peter Eastty who headed up the Oxford team. They had read my PhD Thesis and other papers on Sigma Delta.
Last edited:
Pre-echo is typically an artifact found in lossy compression codecs, not digital filters. The fact that you say you see this in the impulse response already shows that this is just ringing, exactly what I talked about. And no halfway serious DAC has only .1 dB passband ripple. Even a lower-end AK4490 has .005 dB of ripple, and .01 or better is fairly typical, even for budget codecs like AD1938.
Got any reference for this?
KSTR
Major Contributor
Lars is correct, please believe him.
You can try the reverse thing as well: Make an impulse response with just one centered single-sample dirac pulse, then add a small similar pulse, say at -60dB, left and right of the main pulse at a specific distance. Now look at the frequency response (centered and non-windowed FFT) and guess what you will see? The mentioned linear-phase passband ripple.
I would agree, though, that with most of todays DAC filters the passband ripple is a) small enough in magnitude and b) irregular enough in pattern so that the pre- and post-echos are unlikely an issue. Filter designers have learned a thing or two over the decades.
All of this I would still consider as ringing, or just ripple, as much more knowledgeable people call it:
www.audiosciencereview.com
This is NOT called pre-echo, although many audiophile forum posts try to peddle this to make it sound much more scary than it really is.
Minimum Phase vs Linear Phase
This choice shows up in various places: 1. Many DAC's, including several reviewed here. 2. Most implementations of SOX 3. Rephase (default is minimum phase.) My understanding is linear phase filters have pre-ringing. Minimum filters put all the ringing after the pulse. When lossy audio...
www.audiosciencereview.com
This is NOT called pre-echo, although many audiophile forum posts try to peddle this to make it sound much more scary than it really is.
Last edited:
again, please, let’s not confuse pre ringing and pre echo. As i said earlier : pre ring is a consequence of phase linearity and pre echo is sa consequence of periodic pass band ripple.
The pre echo affects all frequencies obviously. That is the mathematical consequence.
The pre echo affects all frequencies obviously. That is the mathematical consequence.
It actually makes a lot of sense, because you can apply volume control. First replace 0 and 1 by +v and -v where v is between 0 and 63. Then run the low pass filter that is applies anyway in an AKM chip. You get 64 levels of volume (including silence). This is just the simplest thing, in practice it is more sophisticated.
“By” = “my”?
yes my
I need to get your PhD Thesis then, I am quite curious.
It's not audible on a properly conducted DBT volume matched test because of nyquist-shannon sampling theorem and proper anti-imaging low pass filter.
With very long and steep filters though especially using external software like HQPlayer you can actually notice delay before playback as a consequence of extremely long pre-ringing in the time domain.
which test do you refer to? reference?
I don’t see how the Nyquist sampling theorem has anything to do with audibility of a pre echo?
Similar threads
