Hi Ken,Hi, Gerrit,
I've seen two definitions of subjective transparency among audiophiles. One definition is: The revealing of more information. Which, is what, I believe, you are addressing. The other definition is: Not altering the information. Yes, shades of the power amplifier distortion debate. I'm not a pure objectivist, so bear with me for a moment.
Sorry for the belated reply to your post of #885.
Audiophile definitions of transparency as "revealing of more information” appear in comparative contrast to "not revealing of more information” . Whereupon the other audiophile definition “not altering the information” is in seeming contrast to “altering of information”. The difficulty is that the descriptors in quotes reflect the range over which judgments are made. Hence transparency isn't being defined, rather is a measurement range descriptor. Alternatively, given that sonic presentations are variant from being "revealing or not revealing of information” or “altering or not altering information” it does not spring to mind that these are definitions of transparency, or alternatively definitions of obscurity. Ultimately the conciseness of a word being defined seems must be interchangeable with its definition. In other words, it doesn't seem that the word being defined can be guessed at from its definition.
It would be nice when some more listen to 88.2/24 and compare it to the original 44.1/16.
So far only DDAC did the NOS Dac test.
Statistics based on one experience is a bit meager
Hans
Hi Hans,
I would like to listen to them though I am not sure where are the files are located
It would be nice when some more listen to 88.2/24 and compare it to the original 44.1/16.
So far only DDAC did the NOS Dac test.
Statistics based on one experience is a bit meager
Hans
I am a bit confused over the goal of this exercise.
First test was if upsampling+downsampling with long FIR lengths were audible.
This test now tests only the upsampling with the same long FIR filters?
Ok, so this only works if I would play the 44.1 as NOS I guess?
(which I dont have)
It would be nice when some more listen to 88.2/24 and compare it to the original 44.1/16.
So far only DDAC did the NOS Dac test.
Statistics based on one experience is a bit meager
Hans
I believe that Frank (Lampie519) also will conduct the experiment, if he can figure a way to play the PGGB 88.2 files from his PC, over to his SPDIF input DAC.
I've also invited Frank to share his subjective comparison of the PGGB 88.2 upsampled files, to their 44.1 source files upsampled to 88.2 by his reference Chord M-Scale device.
This is an age old audiophile debate. Whether or not, altering the signal in some certain way enables more musically significant information to be revealed. I think that an argument could be made either way, depending on how holistic a system view one takes. If the room and the ear-brain system are excluded, I think it correct to view any physical change to the signal as degrading it's transparency. On the other hand, once the signal leaves the loudspeaker, it interacts with the room and, most importantly, with the human ear-brain psychoacoustic system. Which has masking effects, and non-linear distortion that could interact with signal alterations in a way which conceivably makes the information (the music) more intelligible or emotionally communicative. Possibly, improving the subjective transparency. I do believe that one day, however, the two will be completely merged, once there is a full and complete understanding of the signal's interaction with all psychoacoustic factors.
The kind of transparency which we were testing for in that first digital filter experiment was differential subjective signal transparency. This experiment was to explore the hypothesis that typical OS interpolation-filters produce audible artifacts that create sound which is subjectively different from NOS. So, Hans' interesting suggestion was to resample a 44.1 song track up, then back down to 44.1 again, and then listen for whether it still sounds like the original 44.1 song track when both are then played on a NOS DAC. Should we find some resampling software product which statistically sounded the same as the original track, it would indicate that artifacts from inadequate OS interpolation-filter design/implementation were, indeed, responsible. Which, the 44.1 PGGB resampling experiment indicates, although, admittedly, with too low of a sample size to place statistical confidence in.
Here, is the link to the Dropbox containing the 88.2KHz files upsampled by the PGGB resampling software product. The objective here is to subjectively compare the 88.2 files to their original 44.1KHz source files, and listen for whether they sound the same as, or worse than, or better than their source files.
The ultimate goal of this thread is to determine a method/formula for digital playback that's subjectively superior to both NOS, and to typical OS. Essentially, combining the best of both.
Dropbox - NOS1_88.2 - Simplify your life
I suggest, trying the experiment with your OS DAC anyhow. Compare the 88.2 files with their matching 44.1 source files, and tell us what you hear, if anything.
Thanks for the files and your previous post Ken. Unfortunately my DAC only supports 44.1 and 48KHz hence I need to find some DAC that can take the files. Overall it is not my intention to undermine the objective outlined, rather to support it by challenging the use of the word "transparency", that in my view seems limiting the success of that objective.
Any two recordings can be compared on the basis of relative transparency or obscurity. In other words one file can be found more obscure than the other, or one more transparent than the other. My concern is that by limiting the descriptors of comparison to transparency alone, as seems universally the case in the thread, this can suggest or imply that transparency is reflective of reality as thereupon also being suggested reflecting neutrality.
To one of your points brings about a question: Prior to being imposed on human hearing should un-amplified naturally occurring acoustic waves in free space be considered transparent or obscure? My contention is that such waves should be considered neutral, being a reference that upon subsequent manipulation by human hearing, or other mechanisms of variance, can become more or less obscure or transparent, both potentially divergent from neutrality.
For the sake of argument, consider a file that is neutral being imposed upon a spectrum analyzer. The degree to which the spectrum is detailed is dependent upon the bandwidth being swept over the frequency of interest. This is to conclude that the increase in visible harmonics is equivalent to an increase in detail and thereupon an increase in apparent transparency. This is to suggest that the degree to which harmonics are transparent isn't necessarily related to the nature of the file. In a similar way, the degree that individual sounds in a field of sounds is variantly transparent is also a function of the direction that the horns (our ears) are variantly pointed, thereupon being more restrictively focused.
Transparency and obscurity relate to the manner that sounds in an acoustic field are either already manipulated or as further manipulated by the hearing mechanism, and that in variant combinations contribute to variant psycho-acoustic experiences. Hence there are numerous comparisons that can be made upon the presentation of two recorded files. One test is a comparison of the two files A and B, the others in the way file A and B are compared to ones memory and belief of a perceived neutral reality, or other files that are considered more or less reflective of a perceived reality. As an example, some files are perceived overly transparent, to the extent that obscurity, being naturally manifest as background noise or absorbency (as occurs in an anechoic environment) seemingly disappears altogether in becoming of descriptor overly analytic.
That's not the objective of our differential transparency test. We are not testing for which of two files is the more transparent. We are only testing for whether the two files sound the same. Which is determined statistically. This is a valid objective when one of the two choices is directly derived from the other, via some process under test. Should the files sound the same, then the process which created the derived file is subjectively transparent. Should they sound different, then that same process is not subjectively transparent. This relieves the listener from attempting to determine which is the 'better' sounding file, only whether they sound the same, or not.
I have no issue with those priorities. I only point out that certain alterations to the physical signal may produce more emotionally communication, primarily do to human psychoacoustics. I feel that both of these are valid, depending on main the objective of the listener. In listening to music, do they most desire an emotional reward, or an intellectual reward?
I'll only say that I'm not interested in such a divergent, and well worn topic as a subjective versus objective debate in this thread.
The word transparency has a multiplicity of words synonymous to it. This promotes its meaning being taken out of context. In other words, one could equally state Should the files sound the same, then the process which created the derived file is subjectively the same. This brings about the question: "What is the objective intent or advantage of using the word "transparency" in lieu of "the same"?
Completely agree. Positive human experiences (seemingly the fundamental cause to produce or reproduce music) often comes down to a head vs. heart engagement that is often subject to variant moods.
Neither am I, rather to support greater clarity of objective advantage in order to advance the cause.
I suggest, trying the experiment with your OS DAC anyhow. Compare the 88.2 files with their matching 44.1 source files, and tell us what you hear, if anything.
PM to Hans send
Hans asked me to post directly. Not expecting influence I guess... It was a few posts ago.
To recall, I could not really hear any significant difference between source and the up-sampled 88.2 versions (up-sampled with the 4-letter word software I keep forgetting )
Only after I up-sampled with Roon I could hear a difference with the original 44.1 source (I liked the the original better here)
so feel free to post your observations directly here?
To recall, I could not really hear any significant difference between source and the up-sampled 88.2 versions (up-sampled with the 4-letter word software I keep forgetting
) Only after I up-sampled with Roon I could hear a difference with the original 44.1 source (I liked the the original better here)
so feel free to post your observations directly here?
I tested the 4 pairs again with ABX in foobar.
I could only hear difference between :
File A: 13 Bach_ Pastorale In C Minor.wav
File B: 15 Bach_ Pastorale In C Minor.wav
----------
Total: 4/16
p-value: 0.9894 (98.94%)
It seems I prefered A 13.
All the others resulted in in non significant diffrences.
I could only hear difference between :
File A: 13 Bach_ Pastorale In C Minor.wav
File B: 15 Bach_ Pastorale In C Minor.wav
----------
Total: 4/16
p-value: 0.9894 (98.94%)
It seems I prefered A 13.
All the others resulted in in non significant diffrences.
I believe, that I've sufficiently explained my reasoning for that. However, it's fine with me if you prefer to utilize different words.
As a general statement, we realize that we're not conducting some AES approved listening experiment. The very nature of our being a group of internationally dispersed hobbyists precludes conducting a properly scientifically-controlled investigation. Even so, we hope to discover a method/formula that enables subjectively superior sounding digital playback, and to have good fun along the way while we investigate that question.