The signal is reconstructed to look the same, but it doesn't sound the same. This is the moment in time when you should stop using the calculators and looking at the oscilloscopes, and start listening to various DAC types. The irreversible harm is done in the recording studio when the analog stream is converted to digital. MSB promotes their super low phase noise oscillators to be used in the recording studios during A to D conversion, as being the most important place in the whole digital sound reproduction chain.
I know it will be hard to let go of your beliefs.... have you listened to a NOS DAC with a simplistic approach to digital sound reproduction? USB audio in, I2S straight to a DAC chip with no digital filtering of any kind? The resistor I/V straight to an amplifier? No Kalis or other resamplers of any kind? The lowest phase noise oscillators you can get your hands on?
Thanks Alex for your constructive reply. £125.00 is the cheapest I found it on ebay
If they are still alive I doubt they're going to contribute here, now it's descended into a NOS Bashing thread !! That wasn't my intention and as one poster said, I wish I hadn't mentioned it now !
So Mods, feel free to close this thread.
P.
I guess you're already familiar with this website but just in case you've not seen it before:
Philips Compact disc history and technology (Philips, 1984)
Some great info on there I'm sure you'll enjoy.
Best regards, Mike
You see ghosts. Imagination makes people sick. Today, electrical engineering can process and measure about 1000 times better than we can hear. What are you looking for? What you think you hear is imagination. Digital reproduction has been audibly improved by no penny since 1979. There is nothing to improve, because the few improvements are not audible.
I have - though not from USB, rather a QA550 wav file player. Sounds better with an anti-imaging filter after the DAC, no question.
If the "irreversible harm" is done at the ADC stage, how and why should NOS DAC somehow negate that harm ? And let's not forget the degradation coming from microphones... Once music is turned into an electrical signal, if it looks the same, it is the same.
edit: and yes, I've heard quite a few NOS DAC, they were all the rage a few years back in the French DIY scene. They make a truly minor difference in practice that some might prefer.
Last edited:
If it sounds better to you, then fine. I do not want to start a discussion about what "better" really means to you, or me... You found the type of DAC that suits you and that's what matters.
Inappropriate posts have been removed. Kindly keep the discussion civil or the thread will be closed.
Not just to me. There are others on DIYA who'll be receiving their NOS DACs with filtering over the next couple of weeks or so. They'll be giving listening feedback on DIYA at some point.
PDF preview
Seems to be a good text, but note that it's 9 yrs old now.
A free 30-pg preview is here:
http://download.e-bookshelf.de/download/0000/0051/11/L-G-0000005111-0002334262.pdf
Seems to be a good text, but note that it's 9 yrs old now.
A free 30-pg preview is here:
http://download.e-bookshelf.de/download/0000/0051/11/L-G-0000005111-0002334262.pdf
Last edited:
A weak point of most oversampling DACs is that they have no headroom at all for intersample overshoots. Due to the practice of normalizing the largest samples to 0 dBFS during mastering, most of the modern digital recordings have intersample overshoots and drive most of the oversampling DACs into clipping.
Intersample Overs in CD Recordings - Benchmark Media Systems, Inc.
Non-oversampling DACs with or without reconstruction filter can normally handle intersample overshoots with no problems, as can oversampling DACs that are designed to play music rather than to have as high a SINAD as possible at 0 dBFS.
Intersample Overs in CD Recordings - Benchmark Media Systems, Inc.
Non-oversampling DACs with or without reconstruction filter can normally handle intersample overshoots with no problems, as can oversampling DACs that are designed to play music rather than to have as high a SINAD as possible at 0 dBFS.
Upsampling...
Unclear where Benchmark is going with that hypothesis?? This is the first I've heard of this being an issue. Were the Philips engineers and scientists aware of overshoot in the original development of digital audio?
About Benchmark ... all their A/D and D/A products employ oversampling (TTBOMK).
If the oversampling is reasonably high, methinks the "end product" [the converted signal] averages out fairly well.
That said, the overshoot issue -- if it is an issue -- may be a reason many commercial "audiophile"- or professional-grade DACs and ADCs upsample before oversampling.
As Benchmark notes on that same linked page:
Unclear where Benchmark is going with that hypothesis?? This is the first I've heard of this being an issue. Were the Philips engineers and scientists aware of overshoot in the original development of digital audio?
About Benchmark ... all their A/D and D/A products employ oversampling (TTBOMK).
If the oversampling is reasonably high, methinks the "end product" [the converted signal] averages out fairly well.
That said, the overshoot issue -- if it is an issue -- may be a reason many commercial "audiophile"- or professional-grade DACs and ADCs upsample before oversampling.
As Benchmark notes on that same linked page:
Last edited:
Hello,
Almost funny how some call others opinions based on experience and testing "beliefs" but completely refuse to do any kind of actual blind or scientific testing to back up their claims.
Seems they are so afraid they won't be able to hear what they claim that they deny actual testing has gotten us to the planets and down to the quantum level as well. Testing just couldn't apply to anything that they figure it wouldn't support.
Of course when you KNOW what it is you like, anything different is wrong. Regardless if it is more accurate to the original. Like I said, a religion, only belief is needed.
Almost funny how some call others opinions based on experience and testing "beliefs" but completely refuse to do any kind of actual blind or scientific testing to back up their claims.
Seems they are so afraid they won't be able to hear what they claim that they deny actual testing has gotten us to the planets and down to the quantum level as well. Testing just couldn't apply to anything that they figure it wouldn't support.
Of course when you KNOW what it is you like, anything different is wrong. Regardless if it is more accurate to the original. Like I said, a religion, only belief is needed.
Back then the recordings were not peak sample normalized yet. They were recorded at a level that left plenty of headroom for occasional peaks, so the issue didn't exist at all.
I personally knew Rudy van de Plassche, who designed the first Philips DAC's and literally wrote the book (actually several of them) he would laugh at the concept of NOS DAC's.
On the other point no one has demonstrated to me that a recording worth listening to has enough inter-sample overs to create a "sound" rather than a few very short clips per minute if even that. I studied this for ADSL once you are at 10dB of crest factor they are very infrequent.
Last edited:
"HIGH-HEADROOM INTERPOLATION"???
And that's what Benchmark claims in that page you linked...
Frankly, all that sounds a bit unsubstantiated to me.
But that said ...
I own a portable DAP made by Colorfly, that uses the CS4398 DAC chip. This DAC has an internal digital volume control, which the Colorfly DAP uses for both its Line Out level and Headphone Out. I only use LO.
Not sure this is related to the overshoot issue, but I have noted that this DAP sounds its best at about 2/3 (66,67%). Theoretically, "Max Level" (100%) should shut off the volume control.-- and, and hence, optimize SQ. However, the subjective SQ suffers at 100% (0 dB).
It may have NOTHING to do with that "overshoot" thing, though! The volume pot on my outboard headphone amp (or external preamp) may simply operate in a more linear region (turned up higher).
And that's what Benchmark claims in that page you linked...
Frankly, all that sounds a bit unsubstantiated to me.
But that said ...
I own a portable DAP made by Colorfly, that uses the CS4398 DAC chip. This DAC has an internal digital volume control, which the Colorfly DAP uses for both its Line Out level and Headphone Out. I only use LO.
Not sure this is related to the overshoot issue, but I have noted that this DAP sounds its best at about 2/3 (66,67%). Theoretically, "Max Level" (100%) should shut off the volume control.-- and, and hence, optimize SQ. However, the subjective SQ suffers at 100% (0 dB).
It may have NOTHING to do with that "overshoot" thing, though! The volume pot on my outboard headphone amp (or external preamp) may simply operate in a more linear region (turned up higher).
Last edited:
That 3.01dB number is an extreme example, the spectral nature of real music (falling in energy at ~1/f) would make this very rare. The data on some recent CD remasters literally fills the screen rail to rail, who would care to listen to these on a $2000 DAC?
- Status
- This old topic is closed. If you want to reopen this topic, contact a moderator using the "Report Post" button.