The Black Hole......

...Here is an interesting comparison..... I first heard major differences between my own analog Master tapes and CD.
YouTube
Do they all sound the same to you... i am using the earphones for listening thru my computer DAC and hear very different sounds from each format. As I do differences between CD and 24/192+...THx-RNMarsh

Hi Richard,

While I support remastering from the original hires source and the concept of higher than CD res for many albums, the video you linked to is invalid, the presenter has not verified the same master was used to make all the media he was comparing.

As such, the comparison is apples, oranges, cherries, grapes, kiwifruit, pawpaw, and pomegranates. His ignorance of the difference between masters is typical of people making invalid media comparisons.

Having done the same myself with a variety of source masters in a studio environment that has access to a full CD/DVD replication plant, I can tell you properly executed digital storage of 16/44.1 only compromises some highly dynamic orchestral music and perhaps (though it is in doubt) some music with very high >15KHz content. This has been my opinion, although it has been shared by many other respected engineers like Doug Sax, Jack Renner and Tom Jung.

Of course DAs and ADs with good filters like the Apogees are necessary to keep from mangling the high end, and proper level setting is critical.

It is for these reasons I chose 24/96 myself when making masters in the studio and for archiving, not 16/44.1. In retrospect this gives the option of a hires release without remastering, although at the time it was merely to preserve dynamics.

Cheers!
Howie
 
Ahh ok got it. I guessed it was something along these lines but I was not aware of the detail. This in effect then band limits the signal in the case of a non repeating waveform, but in repeating waveforms, you can actually see ‘into’ much wider bandwidths- one example would be fast rise-time signals.

(There are still functions on my scope I haven’t got around to trying)

Yes. For the 500 MHz case, it can display a full waveform accurately, but requires the input remain exactly unchanging for 20 cycles or 40 nSec.

If the signal changes during that 20 cycles, the display will not be an accurate portrayal of the signal.

Jn
 
Seriously. The M2s are safely in indoors storage. Now using Sound Lab electrostats (645 frame size). Very painful even though I only had to pay for new plates. They kill the M2s, but they are huge things in the living room.
SOUND LAB
SOUND LAB

Those remind me of the best imaging speakers I have ever heard: the Beveridge 2SWs. Acoustic memory is of course highly suspect, but I remember a pair of these in a showroom which projected an amazingly holographic image which remained stable, even walking through and around in it. They have been and are still way out of my pay grade...

Cheers!
Howie
 
What Howie no bananas? Oh wait there are a lot of those folks already here!

CDs will be considered perfect once major production stops and they become a specialty item.

The last radio station I built, had a problem getting cymbals through their air chain. Problem got resolved when they fired their staff "Engineer." Never clear to me what he screwed up, his replacement, I think just reset the Studio to Transmitter chain back to their normal settings.
 
This has been my opinion, although it has been shared by many other respected engineers like Doug Sax, Jack Renner and Tom Jung.

Of course DAs and ADs with good filters like the Apogees are necessary to keep from mangling the high end, and proper level setting is critical.

Howie,
Although not at the level of the guys in your list, I agree completely.

A problem some/many people have reproducing HF from well made CDs to their satisfaction is with their DACs. Of course, some people are satisfied with lesser DACs than you, and some want DACs that are even better.

I suspect that is part of what is bothering Richard about CDs.
 
Last edited:
Those remind me of the best imaging speakers I have ever heard: the Beveridge 2SWs. Acoustic memory is of course highly suspect, but I remember a pair of these in a showroom which projected an amazingly holographic image which remained stable, even walking through and around in it. They have been and are still way out of my pay grade...

Cheers!
Howie

The Beveridge were unique insofar as they used acoustical lenses for horizontal dispersion and while I haven't hear the Beveridges a similar approach was this kind of acoustical lens on a huge oversized Audiostatic element and that was an experience similar to what you've desribed.

Of course good electrostats are IME always great at imaging, though.
Not to mention the other good things, like resolution, so-called microdynamics and so on......
 
All that's needed is fast enough sampling for the required BW, whatever that may be'
At least 40-50KHz is needed. But "for the required BW"..... there is NO required BW as there is in a communications systems (channels, groups, super groups etc). And, we dont have to fit data onto a rotating mechanical disk any more (LP or CD). We can be wide open if we wish. 20KHz was a minimum which has audible consequences with too low sampling rate, sharp cut-off filter system.

The 40-50kHz, albeit debatable is at least a realistic suggestion.
As for the rest, whatever the means chosen for distributing a digital signal, this will constitute a communication system.
There has to be an imposed BW limit in every communication system. Nothing comes free in technical and economics realm. (formal training, work experience?)

CDs will be considered perfect once major production stops and they become a specialty item.

Where have we seen that happening before? 😀

The last radio station I built, had a problem getting cymbals through their air chain. Problem got resolved when they fired their staff "Engineer." Never clear to me what he screwed up, his replacement, I think just reset the Studio to Transmitter chain back to their normal settings.

OK we got it. Cymbals got through the air despite the 15kHz freq and % modulation limit.🙂
47 CFR § 73.317 47 CFR SS 73.317 - FM transmission system requirements. | CFR | US Law | LII / Legal Information Institute
Carson's bandwidth rule Carson bandwidth rule - Wikipedia

George
 
Seriously. The M2s are safely in indoors storage. Now using Sound Lab electrostats (645 frame size). Very painful even though I only had to pay for new plates. They kill the M2s, but they are huge things in the living room.
SOUND LAB
SOUND LAB

OK I will buy the Sound Labs next. Really, they are similar to the new QUADs in clarity etc. But, larger. The reason I still have the QUADs. Very accurate. However, I appreciate the near field experience with minimal room interaction as much. We will see what happens when I get the right cross-over with it and my new Demair mono blocks. hahha F--ing expensive at $32000 USD a pair (MSRP). Need 4.

You may like the QUADs also as they dont play loud and as you know I like to 'feel' the bass when it plays. 🙂

Happy for you. Finally got some better speakers that will further your quest for perfection.


-Richard
 
Last edited:
Hi Richard,

While I support remastering from the original hires source and the concept of higher than CD res for many albums, the video you linked to is invalid, the presenter has not verified the same master was used to make all the media he was comparing.

Cheers!
Howie

Hi Howie,

I think you missed a subtle comment at the end. Can you hear a difference?

I am not sure from the comments I get on this forum here that some would hear no difference between any of them.

If they heard differences, then they have a chance to judge the sound of CD vs 24/96 and above.


Thats all I meant by that question.


Myself, i use my own analog masters (30 ips 1/2 track) and make my own red book CD in 16/44 (could do 24/96 too) and compare. using Master Link. Old but reliable work horse. Alesis Masterlink ML-9600 I only bought it just to do the experiment and to know for myself. Didnt want to invest in any learning curve for software i wouldnt other wise use. Just push the button and get my answer.


THx-RNMarsh
 
Last edited:
You may like the QUADs also as they dont play loud and as you know I like to 'feel' the bass when it plays. 🙂

Sound Labs have very impressive bass you can feel (an understatement) and they can get plenty loud with powerful amps, no worries. It has been said that new or reconditioned ones need breaking in to open them up a bit: Turn up pink noise loud when you go out of the house. Takes a couple of weeks of stretching out the more flexing parts of mylar for them to fully open up. That's the short version, Jam can give details.
 
Last edited:
From the list that Braca suggested, I've taken a part of a 96/24 track.

There are 3 versions.
1) the original 96/24
2) the original BW filtered at 20Khz but still as 96/24, and
3) version 2 but now limited to 16 bits also in 96/24

However, to make things a bit more puzzling, I have added non correlating content above 20Khz to version 2 after BW filtering.
And also did I add non correlating content between 16 and 24 bits to version 3 after having reduced it to 16 bits.
The added non correlating frequency content has roughly the same power spectrum as the original HF spectrum in version 1, so nothing can go wrong with your tweeters.

To my feeling having non correlating content instead of zero content in those frequency and bit bands, should make the difference with version 1) even bigger.

The 3 files are in random order in the dropbox file below as A, B and C.
Let's see what you can make of it.

Hans

Dropbox - DiyAudio - Simplify your life
 
Looks quite similar to the measured (?) performance graph in the AD713 datasheet.
So doubling the capacitors would result in a cut off frequency below 20 kHz (~17 kHz) which might be too low if used as an anti-aliasing or anti-imaging filter in a 44.1 kHz system.

-3dB @ 16kHz, -17dB @ 20kHz, -24.4dB @ 22.05kHz but as Scott suggested the sexy spec is the benign group delay.
Who knows, even RNM has given up to the idea that the freq cut off isn’t everything.

George
 

Attachments

  • Scott's Antiimaging filter double capacitance.png
    Scott's Antiimaging filter double capacitance.png
    60.9 KB · Views: 193
Sound Labs have very impressive bass you can feel (an understatement) and they can get plenty loud with powerful amps, no worries. It has been said that new or reconditioned ones need breaking in to open them up a bit: Turn up pink noise loud when you go out of the house. Takes a couple of weeks of stretching out the more flexing parts of mylar for them to fully open up. That's the short version, Jam can give details.

I understand that break in process for the membain. next is to measure them for frq response flatness where you listen. Seems like a lot of room interaction making them hard to be without room coloration.


THx-Richard
 
-3dB @ 16kHz, -17dB @ 20kHz, -24.4dB @ 22.05kHz but as Scott suggested the sexy spec is the benign group delay.
Who knows, even RNM has given up to the idea that the freq cut off isn’t everything.

George

??? Its getting more samples per cycle that make a big difference. [Zillion plus 2]

I agree, you want a flat group delay.


THx-RNMarsh
 
Last edited:
CDs, Hires

All,
After re-reading what I sent about CDs I think it is important to give historical context for this issue, as well as clarify the point i was trying to make (and missed).

Studios in the early 1980s were mastering to analog tape for the most part which had excellent HF extension, but merely adequate noise floor. There are labels which had superbly aligned equipment for sure including those for which JC modded decks, but many had poorly aligned mastering decks, and many had poor master copying stations...many which were used to send out masters for CD replication.

People form their opinions of media when initially exposed to it, and many on this list are um, mature enough to have first encountered CDs in the 1980s. At that point in time, the highest fidelity consumer medium was LPs which when properly set up had huge dynamic range, great high frequency response and a high noise floor. CDs killed the noise floor, but it was years before many of the bit linearity, jitter and filter details as many here are experienced with were refined and more importantly the proper mastering techniques for them...perhaps into the 1990s even.

Studio equipment generally leads consumer equipment in terms of advancement, and in this same period studios progressed quite a bit. Sure in the late 1970s and early 1980s there was the Soundstream and Denon 900 (which I hated), but the Sony 3324 and 3348 DASH machines became popular 16/48 machines and made some great albums prized by audiophiles. Of course once the various tracks were summed a 16-bit multitrack recording is arguably greater than 16 bit dynamic range. Many were retrofitted with Apogee filters, and in the late 1980s with 20-bit converters, and later in the mid-1990s with 24-bit converters...and then cheap HDDs and DAWs killed them.

These days just about all recordings are mastered in 24-bit initially, with a down-sampled 16-bit master made to send to replication.

In order to reduce the RHF (Retirement Home Factor) with this post and make it relevant to the discussion, my point is as the source masters improved, limitations in the consumer delivery format become the limiting factor. Twas ever thus, and now CDs are a limitation for uncompressed 24 bit master sources.

Sorry if I mangled the concept I am trying to convey...
Howie
 
Howie - "properly executed digital storage of 16/44.1 only compromises some" ... "Of course DAs and ADs with good filters like the Apogees are necessary to keep from mangling"

Having experienced a "mangling" effect imparted by particular hardware, I have to believe the rest of the electronics has more to do with good sound than merely the sampling particulars. It seems we've assumed that part's all transparent in the discussion - while anything different in the electronic makeup of that which surrounds the (AD / DA) converters is probably as significant as different source masters.
 
...make my own red book CD in 16/44 (could do 24/96 too) and compare. using Master Link.

My opinion only:
Given the CD recorder's age IMHO it is doubtful its data converters are up to the finest performance levels. I would want to put it up against those used by guys like Bob Ludwig and compare the sound.

More opinion:
What makes me think the above is my work with DACs (and ADCs). The best I have gotten out of a CD with AK4499 is very good indeed. Doesn't comport with my prior impression of sample rates such as with the very good-measuring Lynx II data converters. Its ADC/DAC loop clearly sounded better at higher sample rates. However, the more professional level 'Crane Song HEDD 192' ADC sounds about the same at all sample rates, more in keeping with what theory says should be possible.