|
Audio Asylum Thread Printer Get a view of an entire thread on one page |
For Sale Ads |
216.129.31.125
Of course, I'm assuming the best practices for the ADC and DAC processes. Anyone agree?This is important for the practical application of frequency and time/phase correction using digital domain devices, e.g. TacT, DEQX, Behringer DCX2496 and/or DEQ2496.
Bill Bailey
___________________________________________
See my stereo config ... Good enough is good enough
Follow Ups:
1. Take a live feed, convert it to digital, and back to analog, and listen to it. Does it sound as good as the original feed? Clearly not, even with the best equipment.2. Take your record player and phono amp output, do A/D and D/A, and listen. Does it sound as good? No. You recognize the digital nasties after a while.
3. Take your CD player, play a CD (which of course has already been encoded into digital) do some additional D/D and D/A processing. Now that IS likely to sound not significantly worse than playing the CD straight out of the analog outs of the CD player, but with added jitter, RFI, etc. as TK pointed out. I think this is the question you originally asked.
If you want to do digital EQ, best to start with an already digitized source. Digitizing the signal from your phono preamp might be interesting in terms of manipulating frequency equalization, but you will lose most of what is good about analog.
Fremer's experiments with SOTA analog discs, TTs and preamps notwithstanding....he was essentially remastering and improving the quality of poor CD masters.
My question was referring to the No. 1 scenario. Of course we aren't talking about live "live", we're talking about direct, analog feed of live. Your answer, "No". even with "best equipment" we have today. OK.But as for No. 2, I've heard people who disagree but perhaps they were less critical that you.
And as for No. 3, aren't we really talking about upsampling? That seems to be a whole other debate that drags in issues like synchronus versus asynchronus upsampling. There are cetainly people who say they can here a difference there.
Bill Bailey
___________________________________________
See my stereo config ... Good enough is good enough
Upsampling, sure, but also digital EQ, as you mentioned, which would certainly be noticeable and sometimes desirable.I have to agree with T. Krieger, there is something sonically wrong with asynchronous upsampling.
Good 24/96 native PCM is pretty damn good, but it still doesn't have the same ease and "grounded" feel as good analog reproduction. Maybe a 32-bit, 384kHz native PCM recording would have it...but would take up a hell of a lot of data space.
1/ asynchronous upsampling is obviously
not the best practice.
2/ One can well hear AD/DA stages
nt.
Accuphase??
Bill Bailey
___________________________________________
See my stereo config ... Good enough is good enough
I am afraid that cd is too flawed for even an Accuphase to make perfect. However, you want to hear the most realistic digital playback available? This may surprise you...It comes from noneother than their original seperate's, the DP80 & DC81, first made in 1986!
I have respect for Accuphase' reputation. If I could ever come up with the price of the Accuphase models, I would certainly check them out. And I really like the looks of Accuphase: classic, no nonsense elegance.
Bill Bailey
___________________________________________
See my stereo config ... always look for cost effective improvements
As top contender's, I also recommend the DP-70v (which I own), and the original DP-75.
It would be transparent if jitter, RFI, and data corruption didn't exist.....
And isn't inconsistent with my hypothesis.
Bill Bailey
___________________________________________
See my stereo config ... Good enough is good enough
"And isn't inconsistent with my hypothesis."Hmmm.
In theory it's possible for one monkey to write the complete works of Shakespeare if given one typewriter and one month, never mind giving one thousand monkeys one thousand typewriters and an infinitesimal amount of time.
In practice you'd probably get something resembling the script for Southpark, perhaps something a little more coherent.
:0)
Not one professional I know who actually works with digital audio on a day to day basis would ever say that, or agree with it.Even more amazing, not one of the serious pro companies that make ADC's or DAC's say's that their products are sonically "perfect". They know better than to make the claim to all those pro's I mentioned above.
There is a penalty to go into the digital domain from the analog, and back again, and it is not necessarily trivial.
Now, suppose we postulated both the ADC and the DAC were literally SOTA, the best there is right now. Yes, you would be hard pressed to hear them in a signal chain using pre-recorded analog music as the source, but with a live signal, they are still detectable.
As for the typical mid-fi (and I include all those devices you cite in that category, because none of them is even remotely capable of true SOTA performance), forgetaboutit.
BTW, "good enough" don't cut it. :-)
What about staying in the digital domain (until the very end)? Digital out from the CDP to digital in on a DEQ, digital out to a dac--the eq'ing between the cdp and the dac, aside from whatever intended effects the eqing may have--does this necessarily degrade the signal? Thanks. (Got your quick and dirty bass traps in my corners, so this is not a troll but a genuine question).
An ADC and a DAC have very specific and defined jobs, the measurements, as lame and inadequate as they still are, are dialed in to test them for those attributes they are believed to require to perform their jobs well, etc.A digital EQ on the other hand, is deliberately changing the bits, and I can assure you, it is NOT just the EQ of the amplitude that is going on. Unfortunately, the measurements we are currently limited to DO NOT truly test the digital EQ's properly for all that they are doing to the signals.
Various algorithms must be used to create these digital filters, and while most of these algorithms are designed to EMULATE an analog circuit and it's amplitude response, they ALL introduce other problems during the signal manipulations to achieve the intended changes in the amplitude response.
All of a sudden, all kinds of subtle and bizarre things rear their ugly heads to sully the signal: digital multipication, addition, subtraction, dithering, truncation, buffering, and so on, not to even enter into the extremely difficult realm of which algorithm to use.
Many digital EQ's use a form of the digital bi-quad algorithm, which is KNOWN to be suboptimal, but it is what the "free" software that the DSP or the digital filter chip they used came with, and 9 times out of ten, they use the default programming, the default software and algorithm, etc., you end up with the cheapest digital filter solution on the market. Guess what, it don't sound so hot.
Some companies, such as Apogee, Wadia, Arcam, etc. create their own digital filter algorithms, as do some of the digital EQ folks, such as Lake, and some of the others. Are these the ultimate? Who knows, but at least they tried to go one better than the default freebie software the hardware came with for bare minimal functionality. Someone actually listened to those algorithms to see if they sounded better than the cheapie defaults, and in most cases, they do sound much better. Are they perfect? Probably not yet.
Does YOUR particular digital EQ have custom algorithms that were carefully researched and listened to and tweaked by ear, or are they the "out-of-the-box" algorithms that came with the chip-set?
Who knows?
For instance, Behringer tends to "borrow" other companies algorithms, but they don't always pick the 'good ones', since they are limited to who else is using that particular chipset and general hardware. Most of what I have heard from the Behringer stuff is pretty much mid-fi quality, not that good.In theory, routing the signal chain as you desribe CAN provide the potential for a much better situation, but ONLY if the digital EQ is really up to par, which right now, means a near SOTA unit with totaly custom and tweaked algorithms, and these aren't readily available for less than about $4-5k or so.
Thanks Jon for taking the time to give me such a complete answer. Always looking for a free or cheap ride, I have a DEQ2496. Can't afford a Rives or whatever. I see all over AA posts re digital eq having no distortion penalty, and I even think AHC in TAS said something similar recently (although I don't have the mag at hand and if Im wrong I admit it up front), so I was interested as to what you had to say. I have seen no arguments on the side of no digital distortion added, and your arguments coming down on the negative side of that seem pretty compelling. Thanks again! Regards, Tom.
digital out from the CDP into your processor
is the best that you can do.
Now there might be jitter issues,
pbs with the asynchronous upsampling etc...
If you have a vacation this year Jon, I invite you to hear a display of digital that me and a few others believe to be SOTA, but, unconventional compared to todays commercial offerings.Lets just way, it takes some rather large file sizes.
Well, OK, if you're saying TacT is mid-fi then I have no rejoinder. I'll take your word for it that it doesn't meet the highest professional standards.For home application, though, TacT is about where it's at for home application.
Bill Bailey
___________________________________________
See my stereo config ... Good enough is good enough
I know nothing of the long (winded?) theoretical endless binary number crunching conversation and puritan positioning that this thread is likely to evoke based on past similar threads (wow do I have audiophile issues?) - but I do listen to my turntable nightly as output through the Behringer DEQ2496.I don't actually use the DEQ for A/D - rather, I use my Alesis Masterlink (a system judged so transparent by Stereophile that the reviewer noted one could easily archive the sound of various tonearms and cartriges to CD-R for replay in future reviews for reference). I stream the 24/96 output of the Alesis to my Behringer DEQ2496 for playback. Changing cartridges and/or phono stages is much the same with the digital EQ in stream as without. I can A/B to bypass the DEQ and no major difference I'll every cry over, except for a ridiculous amount of bass booming all over my room (which admittedly makes exacting comparisons very difficult).
More to the point, vinyl still sounds like vinyl in all it's glory and despite my very humble turntable system, it sounds simply fantastic and sprays music all over the room, completely detached from the speaker positions - wide and deep. So let's just say it preserves enough phase to phase me. Furthermore, whatever minor losses there are in the 24/96 A/D process are stomped into insignificance by the gain in having a system EQ'd to my room - my system is otherwise unlistenable.
For digital sources I apply digital EQ at the source in computer based playback - but for the phono that's not possible - god bless my DEQ2496. I think I'll buy another one and have sendler do me a direct out mod - keep meaning to do that, never quite get around to it. I also have a SRC2496 but I'm just too damn lazy to try it, things sound just fine as is.
To me the question isn't so much whether it's 100% measurably transparent as whether utilizing such devices bring an overall improvement to your system or not - it sure does mine - I was never so happy in my music loving life as when I sold all my tube traps, acoustic corner pillows, etc - what a joy. Without the DEQ I'd have no turntable. My advice would be for less than the price of most audiophile interconnects, experiment with the DEQ2496 and decide where to go from there. $300 doesn't even count in audiophile prices.
I listen to a MMF-5 with Monolithic Sound Phone Pre /w upgrade power supply through my Panny XR55 biamping my Silverline Sonatina speakers. I still feel vinyl sounds like vinyl or 'analog' compared to my Squeezebox (getting closer but not close) or my SACD thru my philips 963sa.
Probably playback thru a CD player has more jitter or some damaging jitter characteristics than what the analog signal goes thru in a digital equalizer.
Not that, in practice, AD-DA is actually totally in audible, but that the net result can yield a net benefit. Also, that vinyl still sounds like vinyl, which tends to put the lie to the notion that analog technology is inherently superior and that digital cannot sound good by comparison.
Bill Bailey
___________________________________________
See my stereo config ... Good enough is good enough
. . . when good enough is good enough.
Nothing wrong with that I guess (???)
Bill Bailey
___________________________________________
See my stereo config ... always look for cost effective improvements
I actually understand the intent of your original post. Converting into and out of digital is good enough for your needs, and so you deem it transparent. My response to that is, if you can ignore the conversion process then you can draw a straight line between the two points of A/D and D/A. If you can't hear the difference or the difference is so trivial then go with it. I would say that is more of a practical view than a purist one.As far as answering the question "is it transparent?", then my answer is a definite no. It doesn't mean that it is bad, it just means that it is not transparent.
That's my final answer Regis.
It's been years since I followed the psychoacoustic literature in JASA, but I think your supposition is correct about insensititvity to a correctly performed conversion process.And it serves our auditoy process well to be insensitive to many distortions, so we can generalize over disparate conditions. Think of our auditory system from a teleological perspective. Except for binaural phase for locating a danger, for example, phase perception is simply not important to our survival, and a few centuries is insufficient to evolve sensitivity to such. OTOH, sensitivity to acoustic level and frequency content is obviously critical to survival -- who can we eat or who is aiming to eat us, where are they, how close?
(But does adquately handle level and frequency?)
Bill Bailey
___________________________________________
See my stereo config ... Good enough is good enough
No, I was not suggesting anthing about conversion between analog and digital domains. I was suggesting that perception of monaural phase is not important for human survival, whereas level and frequency are. Binaural phase to some extent but differences in binaural level are useful for indentifing the direction of potential danger.A classic demonstration is to set the level of a signal that is inphase in both ears so it is just masked by noise. Then flip the phase of the signal in one ear; the signal becomes very audible, and needs to be attentuated as much as 15 dB to again be masked. The magnitude of the effect depends on the frequency, being greatest at lower frequencies. This effect, called the Masking Level Difference (MLD) or sometimes the Binaural Msking level Difference (BMDL), is an intereting psychoacoustic phenomenon that almost always amazes the observer.
This post is made possible by the generous support of people like you and our sponsors: