[gilency]

Mmmh....Aha!
I know nothing about music theory and never played the piano or composed my own music either.
I certainly enjoy my resistors more than the music. Isn't that what's all about? 

 

[00940]

Quote:

.

 
I really wonder what you mean to say ? The OS/NOS debate isn't about science, it's just about making technology operate properly inside its given theoretical framework. As it is now, there is no measurable fact that would question this theoretical framework. Purely subjective impressions are not very useful facts in physics.
 
And anyway, the only field of science NOS fans can hope to push forward at this point are psycho-acoustics and belief-oriented economics

 

[gilency]

Actually it is. NOS-DACS are old technology.
I am glad you liked it and that works for you

 

[Baxide]

Quote:

Purely subjective impressions are not very useful facts in physics.

But it does not stop us from deciding which dish tastes better than another, even if both are made with the same ingredients and cooked for the same amount of time.

 

[00940]

@Baxide: mmm culinary chemistry... nothing better.

  I'd argue that the analogy is a bit weak though. There is never quite "the same ingredients" in cooking. Also, we are less likely to argue that what tastes "better" is so because of a technical reasons. We will usually more readily accept our cultural and personnal bias and how such decisions are quite arbitrary.
 
@gilency: if you're answering to me, it's a bit cryptic...

  NOS/filterless DAC are indeed the poor implementation of an old technology, technically flawed. I like the results of oversampling and filtering, that works for me

 

[Baxide]

Quote:

@Baxide: mmm culinary chemistry... nothing better.

  I'd argue that the analogy is a bit weak though. There is never quite "the same ingredients" in cooking. Also, we are less likely to argue that what tastes "better" is so because of a technical reasons. We will usually more readily accept our cultural and personnal bias and how such decisions are quite arbitrary.

Ears, eyes,and tongue contribute to our sensory perception. Whilst you are correct about the "same ingredients" you should then also give credit to the use of the same bit rate, sample rate, and volume level during the A to D process.
In the case of NOS, the sample rate is lower, which means there are more "quieter" gaps in between the actual music signal being played. That means there is less information to process aurally. It's like listening to a conversation by one person in a quiet environment, and listening to the same person in a busy bar. The extra background information can be distracting.

 

[00940]

Quote:

In the case of NOS, the sample rate is lower, which means there are more "quieter" gaps in between the actual music signal being played. That means there is less information to process aurally. It's like listening to a conversation by one person in a quiet environment, and listening to the same person in a busy bar. The extra background information can be distracting.

 
Sorry but that doesn't make sense. If anything, you have more unwanted information at the output of NOS DACs because of the difficulty in filtering. Dan Lawry explained all that at length at the beginning of the thread.

 

[Baxide]

Quote:

 
Sorry but that doesn't make sense. If anything, you have more unwanted information at the output of NOS DACs because of the difficulty in filtering. Dan Lawry explained all that at length at the beginning of the thread.

There is less information in a 16Bit/44.1Kz non-oversampled file. You only have to look at the file size to detect that. My 24Bit/192KHz WAV test file of Hotel California is about 400MB. My 16Bit file is about 35MB. If you consider that both of them would have been derived from the original analogue recording done in the early 70's, it's clear to see which file would have more irrelevant information.
It is not difficult to filter a NOS signal. Its difficult to do it effectively with a minimalist filter circuit.

 

[00940]

Uh, no.
 
1/ If the sample rate is sufficient (above 50k), increasing it through oversampling will not provide more information, relevant or irrelevant. The same physical sine wave will hit your ears, if adequate filtering is provided.*
2/ About no current NOS DACs filter the signal correctly (I posted about that early in the thread, which means years ago). Zanden was an exception. And 5th or 6th order filters are quite a pain/expensive to setup.
 
 
* The only problem with high frequency rates is with analog to digital conversion; too high a frequency sample rates could capture high frequency noises, noises that could modulate the audible signal in the reproduction chain. That's why we need to limit the bandwidth of what is captured. But it doesn't really concern OS DAC, rather what's ahead of the ADC side.

 

[Baxide]

You have just made the case against DACs with 384KHz capability.
 

 

[00940]

No, against source material with 384khz sampling rate.
 
You need 384khz DACs to oversample 8X the 48khz source material. Oversampling (or upsampling or whatever way you want to call it) doesn't create problem at that stage, it doesn't add any information.
 
Lavry has a a good paper on why high sample rate materials are not useful.

 

[Duncan]

Quote:

I thought the reason early CDs sounded so bad was due to the recording processes back then. The existing technology for recording was more customized for vinyl and cassette tapes which are both analog. On top of that the average engineer(or whoever) was more skilled and experienced with analog media than they were with the new digital format. To me that equals that the data on the CD itself was distorted -- not that the CD itself or the player could not faithfully reproduce what was on the disc.

So at the time, because of existing tech, vinyl probably did sound better than CD. But mostly because you had a mature technology vs a baby one.

Thats why I am not so sure about the analogy between NOS dacs and tubes/vinyl. With tubes we are talking about something very simple -- they are either buffering or amplifying - it colors or in some cases softens the sound to some degree but nothing is lost -- tubes are about taste. Vinyl is analog and CDs are digital -- nothing is missing from a vinyl reproduction - CDs are the approximation. No whether you like the vinyl sound is something else.

The reason things sounded bad to begin with was the use of 14 bit DACs (TDA1540 was the first DAC commercially used, and it was flawed, which is why Sony had to step in and help Philips to make a true 16 bit DAC)...
 
Having just been on a mini adventure with an NOS DAC, then to an oversampling one, to an upsampling one (there are differences between over and up!), I have to say that I personally preferred the sound of the oversampling DAC - sounded warmer, lusher...
 
Arguably that could be more to do with implementation, rather than technology involved, however - can that not be said for any of the arguments - for, and against? 

 

[405line]

Maybe the cooking analogy works if you had three meals identically prepared and then to taste the results from gas, electric and microwaved cooked.
For me NOS sounds like a "digital turntable" 3D, layered, natural, not bright or dull, that depends on the recording and importantly for me, not at all localized around the plane of the speakers.
 
"Since the sampling frequency of CD is 44.1kHz, each delay time for the 1 x sampling is 22.ms per tap. To achieve 8 x sampling, SM5842 repeats 2 x sampling three times, and each step incorporates the taps of, 169 degrees for 2 x, 29 degrees for 4 x, and 17 degrees for 8 x. The accumulated delay of each step becomes, 1.92ms, 0.16ms, and 0.05ms: total of 2.13ms. Our auditory sense does the frequency analysis at every 2ms interval, and 2.13ms of delay can be caught by our ear. P.S not sure if it is still done this way as this is from a 1997 paper.
 
"If the speed of sound is 346m/s, the total length of the row of speakers becomes 737mm. ( In the diagram, the distance between each speaker is presented by the total delay divided by the total number of taps" 
 
"The difference between the non-oversampling DAC and the conventional DAC with the digital filter lies whether you attach importance on the accuracy in the time domain or in the frequency domain. In other words, whether you choose the musical performance or the quality of a sound"
 
To me fits perfectly with MY empirical evidence of a sort of 2D quality (annoying if you've listened to records before) and a slightly edgy sound from OS type dacs.

 

 

[gilency]

Quote:

Quote:

Originally Posted by gevorg /img/forum/go_quote.gif Dan, thank you for your detailed explanations and bringing it down to layman's words. I have no reason to doubt you and the whole pro-audio industry that by science the OS way is more accurate/superior. I just don't understand why so many people (here on head-fi, other hifi forums, and even reviewers) find the so called NOS dacs to be more "musical", "vinyl like" and "less fatiguing". Its hard to call it placebo since there are a lot of different people who think this way of NOS dacs. Then can it be just euphony? Is it possible for people to just prefer the less than perfect sound of NOS? If that's the case, then I wouldn't mind to sacrifice a bit on the highs to get an overall sound that is more "musical", "vinyl like" and "less fatiguing", since most of the music occurs at well under 20kHz.

You are in fact asking me to get off the technical and talk about non technical stuff. I rarely do it, but I will make an exception.

Most often, when folks want to believe something, and they have nothing else to hang on, they try to claim that many others or that the majority share their view. Upon careful examination, most often such claims do not hold.

There are also a lot more people that do NOT like or use NOS DACs. In fact, I am not aware of any serious professional studio that uses NOS. I am a "major force" in the pro audio industry (music production), and I was not even aware of the fact that NOS is still around until I came to this forum! To my knowledge, this NOS is rather "localized small cult" in audio.

There are no IC makers that support NOS for audio for many years. Why do you think it is so? Can you answer my question?

I have no desire to get into argument about subjective information. I saw a whole nation, the USA get off track and back up the wrong guy and the wrong war for 8 years. I saw big marketing and advertising money used to wrongly stir a whole audio industry to go for 192KHz sample rate. Why be surprised at a small group that believes something is better when it is not? You can always find a few people that will like something, and others that will dislike the same thing. Folks talk each other into all sorts of things.

I do not argue about tastes, one can prefer the sound of finger nails on a black board, or the sound of digital clipping, and as long as it is a personal taste, there is no point to argue it. I am not surprised to see some small groups here and there that hang on to various false notions.

I prefer to talk about facts, thus stick to the objective. I talk about technical concepts. You may have noticed that I do not use posting to promote my gear, or to out down other people or companies. I stay technical, because it enables me to be objective and correct.

For me, good audio is about the ability to bring a performance into your listening environment with as little distortions and unwanted alterations. I pointed out that NOS is inferior in that regards, it ruins the response, it outputs unwanted high frequency energy, it alters the phase... A real good experienced ear, and/or a "guy with a scope" will realize what I say is correct. What else do you need to know? The acoustic performance when played by NOS is altered significantly. It lacks much of the highs, it pounds unwanted high frequency energy into whatever it drives, it has very high amount of phase non linearity… It can not sound right, if you want to hear music as it was in the original performance. When you listen to a real clarinet or violin, do you wish it did not put out the highs? Do you wish the phase to be re arranged before it gets to your ears?

And those old DA's IC's used by the NOS are from the days where distortions and noise were very high. There was no real 16 bit in 1990! It is not only about the loss of the highs

Of course I stand by my professional knowhow, which is based on technical reality and "mother physics", and it is no coincidence that the gear I design has so much presence in the so many of most respected recording and mastering facilities in the world. But when I some hear people come up with all sorts of ridicules observations, I do not want to offend them. Personally, I often wonder how many of the comments I am amazed by are by people that do not calibrate their ears with enough exposure to real quality acoustic performances. But at the end of the day, folks are free to like what they want.

I love music (I am a musician) and I love electronics design (I am a designer) and I am very dedicated to better audio, and my contributions are for those that share my passion for better audio.

Regards
Dan Lavry
Lavry Engineering

I like Dan Lavry's comments earlier in this thread.

 

[00940]

Quote:

Maybe the cooking analogy works if you had three meals identically prepared and then to taste the results from gas, electric and microwaved cooked.
For me NOS sounds like a "digital turntable" 3D, layered, natural, not bright or dull, that depends on the recording and importantly for me, not at all localized around the plane of the speakers.
 
"Since the sampling frequency of CD is 44.1kHz, each delay time for the 1 x sampling is 22.ms per tap. To achieve 8 x sampling, SM5842 repeats 2 x sampling three times, and each step incorporates the taps of, 169 degrees for 2 x, 29 degrees for 4 x, and 17 degrees for 8 x. The accumulated delay of each step becomes, 1.92ms, 0.16ms, and 0.05ms: total of 2.13ms. Our auditory sense does the frequency analysis at every 2ms interval, and 2.13ms of delay can be caught by our ear. P.S not sure if it is still done this way as this is from a 1997 paper.
 
"If the speed of sound is 346m/s, the total length of the row of speakers becomes 737mm. ( In the diagram, the distance between each speaker is presented by the total delay divided by the total number of taps" 
 
"The difference between the non-oversampling DAC and the conventional DAC with the digital filter lies whether you attach importance on the accuracy in the time domain or in the frequency domain. In other words, whether you choose the musical performance or the quality of a sound"
 
To me fits perfectly with MY empirical evidence of a sort of 2D quality (annoying if you've listened to records before) and a slightly edgy sound from OS type dacs.

 

 
Problem being that your experience cannot fit with that explanation because this explanation is completly off. This was explained on p7 of this thread.
 
To quote Dan Lavry's reaction on this: The only problem that is a complete and total nonsense! It is NOT TECHNICAL, out of mid air bunch of made out nonsense. This is beyond words!