I am on the front lines in the battle for maintaining dynamic range in music both in FM broadcasts and web streams. I have had fierce knock-down, drag-out arguments with other radio engineers on the subject, you want to see impassioned discussions! Geesh! It is analogous to M.A.D of the cold war...commercial stations are so convinced if they aren't the loudest the listener will tune right by.
If you are willing to be patient with college DJs and the ever-changing mix of genres you could listen to the WXYC stream which has zero multi-band or overall compression, just long time-constant AGC to compensate for lazy-a$$ DJs not watching board levels: WXYC Live Stream
If you are really interested I can give you the URL for the 320 kb/s confidence stream via PM...
Cheers,
Howie
Hi Howie,
Just for my benefit as I confuse easily. Given that most people listen at -20dBFS =75dBa (headbangers at THX 85dBa) were you listening to something at effectively 25dB or cranking 50dB of gain in to hear all the low level grot?
Thanks, Howie. Sadly work's blocking it. I'll give it a try at home. Keep up the good work out there and hopefully more stations come around to the idea that louder isn't better.
The Denon disc has the same orchestral piece played at various levels, from peaks at "0" digital to -70 below this. This lowest level confines the bulk of the music to the bottom bit or two + dither. Of course that meant cranking the gain up 50dB to really hear it well. It is pretty good at showing differences between different dither methodologies as well as lowest bit linearity. The best DACs sound pretty smooth and unchanged as the level drops except for the noise floor (dither) begins to come up.
There is an interesting effect with certain dithers which are weighted heavily in amplitude in the 15-20 KHz region: as the music is stepped down in level, it becomes more "airy" either due to the perception of the dither or some aural IM between the music and dither.
I have used this test series to diagnose DACs with missing or non-linear properties, and when we were working with Apogee to build a DAC for our digital transfers (a while ago...) they used a test series very much like it to audition different dither spectra.
And yes, especially in the early 2000s many DACs had a lot of grot. LOL
Cheers,
Howie
I believe you are making a false distinction - auditory perception is "what our brains process and feed our consciousness" there is no such thing as "raw sensory perception" - it's a psycho-physical system using the physiology of the ear & the processing of the nerve impulses by various parts of the brain.
In general, I agree but often the only "art" considered is one of electronics theory & seldom, if ever is "one educated in the art" to mean the workings of auditory perception.
Often, this is the great disconnect found on this forum.
I believe that it's possible for many people to report the same observed perception without there being a sufficient theory or explanation (or existing measurement) in electrical theory to explain what underlies the perception. I suggest that in most of these cases people are looking for answers in the wrong place - they should be looking for possible explanations in the field of auditory perception first & working back for how this may be attempted to be measured in the soundfield. I'm not saying that this is an easy task but too often this approach is ignored.
Last edited:
I gave an example last year of a strand in auditory perception research that interested me - one analysis technique that seems to be used in auditory processing is using summary statistics of the temporal structure of the sound stream - "time-averaged sound statistics"
A research paper was published earlier this year which makes this even more relevant to audio playback - I've maintained this for a while - that background sound texture (ambiance) which is the sound from which the foreground sound (music) emerges is analyzed by the auditory system by using statistical averaging operating over multiple seconds
Adaptive and Selective Time Averaging of Auditory Scenes
So what? I've long held the belief that a subtle change in the background ambiance from which sound emerges is not consciously noticed as a change in this aspect itself but rather it effects how we perceive the foreground sound (music). That this background is being processed/analysed as a summary statistic ranging over many seconds opens up a number of considerations.
Well, maybe devising some measuring techniques which could examine the background 'texture' sound (excluding the foreground sound) using some temporal statistical averaging might be worthwhile?
A research paper was published earlier this year which makes this even more relevant to audio playback - I've maintained this for a while - that background sound texture (ambiance) which is the sound from which the foreground sound (music) emerges is analyzed by the auditory system by using statistical averaging operating over multiple seconds
Adaptive and Selective Time Averaging of Auditory Scenes
So what? I've long held the belief that a subtle change in the background ambiance from which sound emerges is not consciously noticed as a change in this aspect itself but rather it effects how we perceive the foreground sound (music). That this background is being processed/analysed as a summary statistic ranging over many seconds opens up a number of considerations.
Well, maybe devising some measuring techniques which could examine the background 'texture' sound (excluding the foreground sound) using some temporal statistical averaging might be worthwhile?
Last edited:
We are in agreement, if you read what I write, that is indeed what I said. My point being one can never access the "raw" event, our perception is always that which is processed through the auditory cortex.
Cheers,
Howie
Price range has nothing to do with it, I would not have much use for it since I've returned to mostly LP listening.
I have no contacts here with folks that listen to digital mainly, just thought it would be nice to compare some of these boxes. Maybe I'm tainted by years of listening to big valve/LP setups. They all sounded different tuned to individual tastes.
Please clarify one thing, does the Denon disk have the different dithers applied at the A/D end? My early SONY disk had these tracks but they were truncated undithered 16bit so game over at the start. Same with the files used for standard perceptual effects that Jakob linked, I posted pictures here the flute nicely suddenly goes to 0.
IME some of the discussion of these issues is around basically flawed experiments. BTW this is yet another hidden Windows feature, some SW takes 16bit truncated data and adds 24bit dither. You are then fooled by seeing 24bit LSB's in the data.
Last edited:
Hey Scott,
Good question and since the disc was made in 1984 I can't find any information on it regarding the type or presence of dither. BTW I was wrong stating the lowest level of the orchestral recording was -70 dB re normal level on that disc, it is -60 dB, sorry. Here is a link to the track listing:
No Artist - Denon Audio Technical CD (CD, Album) at Discogs
Either way, in my home use of that disc I heard a difference between the OPPO analog and Vega DAC with those tracks so I was not hearing a non-dithered last bit dropout which would have been present in both.
Early ADCs like the Sony PCM1600 allowed the dither to be turned off and the difference was easily noticeable when running low-level tests. Being a novice in digital audio in the early 1980s when I first started experimenting with it, I was initially against the idea of adding noise...until I heard the difference and saw the difference in low-level waveforms. I'll be honest, when running dither spectra tests with Apogee I only really heard a difference with low-level passages in orchestral music like that "airy" enhancement I mentioned before with certain dither spectra.
Cheers!
Howie
Good question and since the disc was made in 1984 I can't find any information on it regarding the type or presence of dither. BTW I was wrong stating the lowest level of the orchestral recording was -70 dB re normal level on that disc, it is -60 dB, sorry. Here is a link to the track listing:
No Artist - Denon Audio Technical CD (CD, Album) at Discogs
Either way, in my home use of that disc I heard a difference between the OPPO analog and Vega DAC with those tracks so I was not hearing a non-dithered last bit dropout which would have been present in both.
Early ADCs like the Sony PCM1600 allowed the dither to be turned off and the difference was easily noticeable when running low-level tests. Being a novice in digital audio in the early 1980s when I first started experimenting with it, I was initially against the idea of adding noise...until I heard the difference and saw the difference in low-level waveforms. I'll be honest, when running dither spectra tests with Apogee I only really heard a difference with low-level passages in orchestral music like that "airy" enhancement I mentioned before with certain dither spectra.
Cheers!
Howie
I have that disk, the use of PIL as "pop" music was funny. BTW that cut is contaminated with TV horz sweep. Pretty sure it was not dithered either.
To clarify my statement: I only heard a difference between the different dither spectra being tested in low-level orchestral passages, not between dither and no dither. I would always use a fixed dither signal in an AD process, having experimented with dynamic dither I object to the noise modulation.
Cheers,
Howie
In a related issue, my copy of "Metal Box" was stolen in college...cool concept and quite a breakthrough for them at the time, one of my housemates would scream "TURN IT OFF!!!" when Lydon sung "Chant."
Good times...
Howie
Thats exactly as I hear high THD %-age of distortion in analog amps ---Airy.
THx-RNMarsh
PIL had 2 singles in the UK top 10 so that does class them as 'popular' by one metric
.
PIL had 2 singles in the UK top 10 so that does class them as 'popular' by one metric
I wonder what Dick and John think of PIL.
Definitely non-mag here, eye doctor seemed to know exactly what he was talking about. Am surprised that so many are using CDs actually, but then I only listen to favorites, less variety on a regular basis.
12s from a kit/diy sd player into a decent dac is better to my ears than a usb input or CD player. Much better to be able and manipulate the power side than be stuck with what companies like oppo provide imo, for audio anyways.
Then you have to find SD cards that work for audio files, new horizons...
- Status
- Not open for further replies.