[arnyk]

   
The purpose of filtering in A/D and D/A conversion isn't at all intuitive.... so perhaps a proper explanation would be in order.
 

The proper explanation is that recording at sample rates up to 4 times the standard CD sample rate is a simple matter of lining up the readily available 24/192 recording gear (I've had the highly regarded LynxTWO gear for over a decade), and make the recording. Been there done that many time. If the 16/44 has any inherent sound quality loss, then downsampling the 24/192 or 24/96 recording to 16/44 should be instantly discernible. If you eliminate the usual audiophile crutches of sighted evaluations, there is nothing that is discerned. In this day and age 24/192 recordings are a simple matter of a little google searching and downloading. Downsampling can easily be done with readily available high quality freeware such as Sox. The proper listening test can be coordinated and run for the interested individual with readily availble high quality freeware such as Foobar2000. Why waste time with pedantry when it is so simple to obtain the evidence of your own ears?

 

[RRod]

 

Spoiler: Warning: Spoiler!

The purpose of filtering in A/D and D/A conversion isn't at all intuitive.... so perhaps a proper explanation would be in order.
 
I'm sure you're familiar with the basic fact that a digital audio file can ONLY be used to store information up to just under the Nyquist frequency - which is 1/2 of the sample rate. For a 44.1 kHz file, the Nyquist frequency is 22.5 kHz, which is why a CD can't contain any information above that frequency (and using 20 kHz as the cutoff frequency does give a tiny safety margin there).
 
However, in reality, any source is going to contain some information above 22 kHz; which will include high order harmonics produced by some instruments like cymbals, as well as actual noise from equipment like preamps and other electronics, hiss from analog master tapes, and even noise present in the room where the recording was made. The problem is that this extra "unrecordable" content doesn't simply disappear when you feed the content into your ADC. In fact, the opposite is true; if there is any content present above the maximum frequency that can be encoded "properly", it is converted into a very audible and unpleasant distortion during the encoding process. The actual process involved, and the exact results, are somewhat complicated, but the net result is that a significant portion of it is "folded back around the Nyquist frequency" into the audio INSIDE THE AUDIO RANGE OF THE ENCODED AUDIO.
 
Let's take our CD as an example: the sample rate is 44 kHz, and the Nyquist frequency is 22 kHz. Now let's assume that, mixed in with my "audible source material", there is a 28 kHz tone (it could simply be some ultrasonic harmonic of some instrument, or just some 26 kHz hiss in the microphone preamp). If I were to feed that source into an ADC which lacked the proper filtering, that 28 kHz ultrasonic would be "folded down around the Nyquist frequency". This means that most of the energy contained in that 28 kHz tone, which starts out being 6 kHz ABOVE the Nyquist frequency (22+6 kHz), and which cannot be encoded into our output file, will be converted by the encoding process into an equivalent amount of energy 6 kHz BELOW the Nyquist frequency (22-6 kHz). So the energy from our ultrasonic 28 kHz noise source, which was totally inaudible, will appear in our encoded file as noise at 16 kHz. This process will occur with all signal energy which is present in the source that is above the Nyquist frequency - and will essentially end up as noise/distortion in the final file. The actual process is somewhat "messier" than my simplified explanation, so that 28 kHz tone will actually cause noise spikes at other points inside the audible range. And, of course, I used a single tone as an example, while the reality is somewhat more complicated. (Think of it sort of like a distorted and inverted reflection in a camera lens interfering with the desired image.) The purpose of the filter in the ADC is to remove any content outside the frequency range which the encoder can handle properly so this doesn't happen (which means anything above the Nyquist frequency).
 
Unfortunately, most real-world recordings contain all sorts of energy above 20 kHz, ranging from harmonics of actual instruments, to room noise and such, and even to analog distortion products. ANY of this which the filter fails to remove will end up as distortion and noise inside the audio band in the resulting encoded file, which is why it is critical that the filter remove it - or at least reduce it to a very low level.
 
(Oversampling avoids this issue by shifting the Nyquist frequency upwards, thus making it easier to design a filter that is flat to all audible frequencies, yet still has sufficient attenuation above the Nyquist frequency. In order to make a "clean" conversion at 44 kHz, to the level required for a CD, you would need a filter that is flat to 20 kHz, yet down about 80 dB at 24 kHz, which is very difficult to achieve in practice. To get a similarly noise and distortion free output at a 96k sample rate, the filter would have to be flat to 20 kHz, and down 80 dB at 50 kHz, which is a far gentler slope, and so much easier to design and build.)  

 

 
I'm well aware of both aliasing and imaging and how things like higher sampling rates and upsampling can help make the job of avoiding them easier for the analog components (by shifting some of the duty to the digital realm). That doesn't change the question: how bad exactly were the pro-level ADCs used for making earlier PCM recordings for CD release. If you were saying that ±1dB up above 18kHz was the problem, then I'm saying that for many of us such issues are already negligible in effect due to the limits of our hearing.

 

[arnyk]

   
He won't accept ABX as the means of testing it.

 
 
Probably an example of cherry-picking testing methodologies in order to obtain the desired results.
 
The problem is not the rejection of ABX, because ABX is just one of many good listening test techniques.  Some are more ideal for some tests than others, but they all can be used to obtain valid results.
 
The probable desire is to continue with some variation on the now-totally-discredited sighted listening evaluation technique or some variation on it. Of course, that can't be accepablable because of the extreme propensity for both false positives and negatives.

 

[arnyk]

   
I'm well aware of both aliasing and imaging and how things like higher sampling rates and upsampling can help make the job of avoiding them easier for the analog components (by shifting some of the duty to the digital realm). That doesn't change the question: how bad exactly were the pro-level ADCs used for making earlier PCM recordings for CD release. If you were saying that ±1dB up above 18kHz was the problem, then I'm saying that for many of us such issues are already negligible in effect due to the limits of our hearing.

 
While frequency response limitations can profoundly affect listening tests, they aren't the core of the problem. The core of the problem is masking,  a profound effect that affects hearing sample rate related effects even among listeners with ideal or near-ideal hearing.

 

[Gr8Desire]

 
I can do that all the time, very easily.  Just close my eyes and hit this:
 

 
That's even tougher than a regular ABX because the ponoplayer's signal chain plays all those formats about as good as it can. Some people complain that the Revealer feature is counterproductive because the PP plays MP3's so well. 
 
Fast switching between songs is always a problem. I prefer to let it roll and listen for overall room sound and decays.

 
Fascinating feature!  

Downsampling 192/24 to 96/24 will create a near perfect result (care to verify?).  But downsampling to a non-integral clock rate of 44.1 and for MP3 (also 44.1kHz) with induce significant dithering (introduction of random sample values to make up for missing time-specific data) during the conversion process.  

Pono is PURPOSELY choosing inappropriate sample rates to make themselves look good. They should be using 48/16 for their so called 'low-res' playback.  That would produce a GREAT result. Not what Pono wants.  

Made my day, this one.

 

 

[RRod]

   
While frequency response limitations can profoundly affect listening tests, they aren't the core of the problem. The core of the problem is masking,  a profound effect that affects hearing sample rate related effects even among listeners with ideal or near-ideal hearing.

 
True, but for masking to even matter you have to be able to hear the particular frequency to begin with. What I'm trying to get at with my questions to Keith are exactly what kind of frequency response issues did early ADCs have (perhaps you have some data on this?).

 

[jcx]

its possible some "worse" pieces of the chain could have distortion that makes otherwise ultrasonic content audible as difference IMD products
 
this was a demonstrated error of some ultrasonic hearing listening tests - both tweeters and some amps were found to be at fault - "audiophile" reputation doesn't guarantee great linearity, particularly of "no feedback" amps driven to high levels
 
so its entirely possible for someone who hasn't verified low distortion with a mic and headphone coupler to really hear a difference even in a properly controlled, blind test - because there could be conventional audio frequency sound differences when the amp or headphone has high levels of ultrasonic signal to distort

 

[Ruben123]

Is there audible dither when comparing 44 vs. 48khz? I know of problems when the sound card can't handle one or the other though

 

[arnyk]

Is there audible dither when comparing 44 vs. 48khz? I know of problems when the sound card can't handle one or the other though

 
 
Sample rate conversion with no other changes does not require  changing the dither. The dither gets resampled along with everything else.  Decreasing the number of bits per sample point does require changing the dithering ..

 

[Ruben123]

Sample rate conversion with no other changes does not require  changing the dither. The dither gets resampled along with everything else.  Decreasing the number of bits per sample point does require changing the dithering ..

OK thanks !

 

[arnyk]

   
True, but for masking to even matter you have to be able to hear the particular frequency to begin with.

 
 
Not a  problem because it is generally the low frequency content that masks the high frequency content.
 
 
Originally Posted by RRod /img/forum/go_quote.gif
 
 

What I'm trying to get at with my questions to Keith are exactly what kind of frequency response issues did early ADCs have (perhaps you have some data on this?).

 
I've previously posted such data as I was able to acquire myself here.  
 
Digital Audio started out with short data words (8-9 bits) and low sample rates (8-12 Khz) if memory serves, but the purpose was telephone service.
 
Prior to the introduction of the CD  data words were generally 12 bits or more, and sample rates were 32 Khz or higher with some of the better examples up around 50 KHz.
 
The first two CD players on the market (Sony CDP 101 and Philips CD 100)  had pretty close to 16 bit resolution (measured) and reasoanbly flat frequency response.
 
Here's the worse of the two - the analog filtered CDP 101:
 

 
About a half dB down at 20 KHz.
 
 
The Philips CD100 had a digital filter and far better response.

 

[RRod]

   
 
Not a  problem because it is generally the low frequency content that masks the high frequency content.
 
 
Originally Posted by RRod /img/forum/go_quote.gif
 
 
 
I've previously posted such data as I was able to acquire myself here.  
 
Digital Audio started out with short data words (8-9 bits) and low sample rates (8-12 Khz) if memory serves, but the purpose was telephone service.
 
Prior to the introduction of the CD  data words were generally 12 bits or more, and sample rates were 32 Khz or higher with some of the better examples up around 50 KHz.
 
The first two CD players on the market (Sony CDP 101 and Philips CD 100)  had pretty close to 16 bit resolution (measured) and reasoanbly flat frequency response.
 
Here's the worse of the two - the analog filtered CDP 101:
 
 
About a half dB down at 20 KHz.
 
 
The Philips CD100 had a digital filter and far better response.

 
Interesting, thanks!

 

[KeithEmo]

   
True, but for masking to even matter you have to be able to hear the particular frequency to begin with. What I'm trying to get at with my questions to Keith are exactly what kind of frequency response issues did early ADCs have (perhaps you have some data on this?).

 
Unfortunately I'm repeating this from memory - and it's mostly based on engineering design articles (as quoted in articles with titles like "here are the problems with old style ADCs and here's how to avoid them"). Since I don't produce recordings, and so have no control over what equipment was used to convert a particular CD, I haven't bothered to keep track of individual examples. Basically, the situation where the gain of a filter rises and falls several times over a relatively narrow range of frequencies near the cutoff frequency is known as "in band ripple", and "engineering best practices" suggest that it is something to be minimized. I can tell you that this type of ripple is easily audible at lower frequencies, where it actually produces an audible "warble" in instruments or voices that move up and down across the frequency range over which it occurs, but I haven't heard of any specific testing to determine the limits of audibility.
 
Since modern oversampling technology has rendered most of this subject moot anyway, it doesn't receive much discussion lately. I did have an opportunity recently to compare specifications on several "professional sample rate converters" sold during the early days of CD production. (I was wondering if any "old pro units" available on the used market would be suitable for use with a DAC.) One unit I recall was very proud of the fact that their unit could reduce jitter "to as low as a mere 2 nanoseconds". 2 nanoseconds is 2000 picoseconds, which would be considered barely passable for a piece of low end consumer gear today. I suspect that many of the really early ADC units probably didn't even include specs for things like jitter, because their importance simply wasn't recognized at the time, which would make comparisons difficult unless you actually were to secure samples of early units and measure them. In all fairness, I should also mention that some particular early models of ADCs had a reputation for sounding very good, and some, like one model from Pacific Microsonics, continue to be used today.
 
(Many of the articles published on this subject predate the popular use of the Internet, so they're difficult to search for and locate.)

 

[dprimary]

 
It could be but it's not because I STATED exactly what I was talking about.
 
Network bandwidth and storage is EVERYTHING in digital audio.  It's the main count of everything, assuming you are using the same encoding and file containers, and in this case I clearly said STEREO PCM at what resolution and if lossy or not.  I was very clear and guess what - out came the personal attacks again.
 
I didn't say 8-bit mono 655k sample rate. No one cares about that for music production.
I didn't say 10 tracks of 14bit/36k sample rate.  No one care about that for music production.
Stay on topic.  Commercial music production, stereo, PCM. Thats all I'm talking about.

 
Originally Posted by FFBookman 
 
 
Until then I'll just remind you all that 320k PCM < 1400k PCM < 3000k PCM < 5600k PCM.  That's available bandwidth at the various resolutions.
 
 
Where do you see anything about bandwidth or resolution? Bit rate is not bandwidth. I know the sample rate and bit depth I'm perfectly capable of calculating storage and network bandwidth requirements. 
 
I have needed to use both sample rates and bandwidth in my examples while the 8 bit one was for a biologist. Needing 10 tracks at 14 bit at 36k is not that unusual. In fact the early digital music recordings were 14 bit and roughly 36 k sample rates.  
 
Here is little history
 
http://www.aes.org/aeshc/pdf/fine_dawn-of-digital.pdf
 

  KeithEmo is right, I'm right, the rest of you are choosing to degrade your audio because you don't think you can hear the degradation.  That's your choice but you should degrade your own files, not continue to claim that I'm imagining things, and not add to the confusion about what people can and should hear.
 
 
I have to ask, is there anything else your body can clearly and repeatedly do that you don't have math or testing to confirm?
 
You "16/44 is the highest" people are the most ridiculous lot of brainiacs I've come across. Continue to deny the sensations of the physical world, it could take another few hundred years to devise the math behind what we all do all the time.  I've done many things in 1 hour this morning that you have no mathematical model for.
 

 
The topic is not about lossy or lossless compression. It is about is anything beyond 16bit 48k even needed for a release format. Bit rates are just a distraction, it is 16/48 we know the bit rate.
 
Confirm it! all we need is one person in billions on the planet to be able to constantly pick out 16/44.1 from a 16/88 or 24/44 recording in a blind repeatable single variable test. Yet Arthur never shows up to pull the sword from the stone.
 
I have enough reason to believe people might be able to detect above 20k that I'm open to the possibilities. Till I can repeat the test.
 
Anyone jumping up and down claiming it is night and day my BS meter is so pegged the needle broke. Yes I can reliably  pick out a well encoded 320kbs AAC from a well recorded raw PCM file but it not trivial. Can I pick between and PCM and AAC of every pop/rock song, the answer is no,  please refer back to the "well recorded" requirement. If it is trivial maybe the encoder is garbage.
 
Don't worry mathematicians are generally about 200 years ahead of the rest of us. The math is figured out centuries before we can possible build it. 

 

[icebear]

...  
Anyone jumping up and down claiming it is night and day my BS meter is so pegged the needle broke.

ROFL

made my day!