Hi-Res - Does it matter? Blind Test by Mark Waldrep

[gmgraves]

12 minutes ago, Miska said:

 

Like Sennheiser MKH-8020 for example that goes to 60 kHz? Or the DPA microphones like 4006 and especially 4007 that go to 40+ kHz? Or Sanken CO-100K that goes to 100 kHz, used for example by Five Four?

 

What I've looked, for example in recordings made by 2L, content goes up to about 60 kHz. So you need at least 120 kHz sampling rate for those.

 

Those are the ones, alright! I was unaware of the Sennheiser, but the DPA and the Sanken mics I was aware of!

2L is a Swedish audiophile label, and I have a couple of their Blu-Ray releases. If you want to capture the 60 KHz on them, you do definitely need at least a 120 KHz sampling rate (likely either 176.4 or 192 KHz), but I seriously question why. BTW, I doubt seriously (in fact , I know) that you won’t find the mikes you just mentioned being used on a regular basis at MCA, Warner, EMI, TelDec, or DGG!

[gmgraves]

12 minutes ago, Miska said:

 

I think I have only three DACs that don't support DSD. All the rest do support DSD, even up to DSD1024...

 

When I buy content I prefer to buy it in the original recording format, not some conversion.

 

I very rarely send LPCM to a DAC though.

 

 My DAC in my main system is a Yiggdrasil, it doesn’t do DSD. Like I said, while my    Oppo UDP-205 with it’s high-end ESS SabreDACs, does do DSD (well, it does SACD, anyway. I’ve never tried it for DSD files), I don’t have the audio output of the Oppo connected to my amplifier system. Right now, I am using an AudioQuest Cobalt in my office system, AFAICS, it does not support DSD, but the Chord Quetest that I have on loan does, and I have used it to listen to my DSD master files. Without that, I have to rely on my Kong MR-2000s or MR-1 to play them back!

[gmgraves]

17 minutes ago, Miska said:

 

I know both especially the DPA but also the Sennheisers being used for lot of classical recordings by different labels. Especially if you look at 5.1 channel Decca-trees or similar suspended from concert hall ceilings. Like here in Helsinki. DPA used to have 130V phantom versions of 4006 etc, and especially those are used in high quality recordings. Many still have the original versions under B&K brand, before DPA split.

 

When I'm looking at for example HiFi-News reviews of recent hires recordings, many of the 96k tracks, or even most, have content reaching the 48k Nyquist.

 

If someone is making hires releases, it only makes sense to invest into hires capable microphones as well.

 

P.S. 2L is Norwegian, not Swedish...

 

Forgive me for the error, you are right 2L is most assuredly Norwegian! I agree that it does make sense that modern Hi-res recordings use microphones with response at least to 30 KHz. But most hi-res being done by and for the majors simply use the normal microphones that they have on hand. And many of the hi-res releases are of older material, either originally analog, originally 16/44 digital, or early attempts at 24/96 or DSD. 

[gmgraves]

1 minute ago, Miska said:

 

OK, the Chord converts DSD to PCM first and then back to SDM. But the Korg does have actually DSD capable DAC too. And a real 1-bit DSD ADC (any PCM recording with it is just DSD-to-PCM conversion).

 

Correct. The two Korgs still do the best job of rendering the DSD that was captured on them in the first place!

[gmgraves]

1 minute ago, sandyk said:

 

Barry Diament for example uses microphones that are only 1dB  down at 40kHz for his 24/192 recordings.

Ok, I suspect our friend Mario Martinez also uses the latest wide-band microphones. When Telarc was still recording, they decided to release everything after around 2000 in DSD, but Bob Woods, their engineer used the same B&K omnis that he always used. While these were “calibration mics”, and had very flat response, I recall that they didn’t go very far above 20K.

[gmgraves]

5 minutes ago, Miska said:

 

You can still get the same stuff under DPA brand, like 4006 and 4007...

That’s right. I had forgotten that B&K’s product line was taken over by DPA. The DPA 2006 is very close to the B&K used by Bob Woods at Telarc. I do not know whether or not DPA has changed it since they took over, but the 2006 has a frequency response of 50 - 20,000 Hz +/- 2dB (and, ostensibly, has usable response to 20 Hz, but DPA doesn’t say how far down the mike is at 20 Hz) and does have response to around 30 KHz, but is more than 10 dB down at that frequency, so I suspect that it is suitable for Hi-def recording. Now, whether or not the B&K equivalent had similar specs, I have no way of knowing.

[gmgraves]

2 minutes ago, Speedskater said:

Robert Woods was the Recording Producer. Jeck Renner (RIP) was the Recording Engineer. Jack passed away in July 2019 at the age of 84.

Thanks. Yes, that’s true. But When I met Bob Woods at a SF AES convention, he said that he decided what equipment, including mikes that they used. Renner was a fan of C.R. Fine of Mercury Living Presence fame, and used his spaced omni recording scheme (with which I do not agree, BTW). Telarc freshened it with better microphones, that’s all (Fine used an omni Telefunken from the early 1950’s. It was only “somewhat” omnidirectional; which is the only reason Fine’s stereo recordings worked as well as they did. Telarc’s B&K mikes were true omnis, so their stereo was not as good as Mercury’s [IMHO, of course]).

 

Nice discussion, BTW. A lot of knowledgable people involved. 

[gmgraves]

1 hour ago, Jud said:

 

Have also heard good things about these guys: http://recordinghacks.com/microphones/Earthworks#:~:targetText=Microphones

Looks good. I am unaccustomed to seeing electret mikes with decent bass response. These, however seem to have it in spades (as well as extended high-frequency response and 140dB+ of headroom. Impressive. As a recording mike, though, it’s limited to spaced array “stereo” which is unfortunate. 

[gmgraves]

10 hours ago, sandyk said:

 

Pleasing colouration perhaps ?

Agreed. I once had a B&O (Bang & Olufsen) stereo ribbon mike. Two figure-of-eight mike units were stacked, one atop the other and the top one could turn, right to left with regard to the bottom one. The mike was gorgeous satin chrome and it came in a beautiful, padded rosewood case with gold lettering. Aside from the fact that at the time, I couldn’t find a mike preamp that was quiet enough to give a decent S/N, and that the top end only went to about 13,500 Hz, recordings made with it sounded marvelous in spite of the noise! I finally sold it due to my inability to find a suitable mike amp (one of life’s little regrets). The mike was designed to be used with a B&O stereo tape deck which contained a pair of proprietary transformers to boost the minuscule ribbon output to a level where it could be electronically amplified without the amp needing 70 to 80 dB of gain! The transformers were not available separately! Today, of course, there are cheap op amps that could do the job easily (Like the TI LME47910, for instance with it’s 2 nV/root-Hertz of self noise). But in the 1980s when I owned the B&O stereo mike, a transformer was de riguer.

[gmgraves]

26 minutes ago, sandyk said:

 Hi George

 I presume that you meant the LME49710 ?

The HA (metal can) version is even better, and I use a couple of them in my DIY DAC

 

Kind Regards

Alex

Yes, I did mean the LME49710. Typos happen. And I agree about the HA variant. It seems to be even quieter than the DIP version (!!??). I suspect that the metal moves heat away from the chip better than does the encapsulating plastic of the DIP.

[gmgraves]

43 minutes ago, esldude said:

Ever think that maybe ribbons with a slower transient response are providing a beneficial limiting upon the rate of change in a digitally sampled system?

With the lightness of the ribbon, I would say that its transient response would be faster than say, a condenser mike. Especially older ones with etched metal diaphragms. I’m sure that the limited high-frequency response of most ribbons  have some affect on transient response, but low inertia due to low mass, would allow it to start and stop very quickly.

[gmgraves]

5 hours ago, Rt66indierock said:

 

Going back to my original vaporware post you would have to say jazz, classical reggae, new age, world and children's aren't doing too well. If you aren't up there with stage and screen at just under 3% of the US Market in 2018 can you say things are OK?

I really don’t understand your use of the term “vaporware”. In the computer world, vaporware is a product that developers/manufacturers keep promising to bring to market, but never do. If I’m not wrong, you are using the term to mean a product that doesn’t perform to it’s makers’ promises. Is that right?

[gmgraves]

8 minutes ago, Rt66indierock said:

 

Nor does a few thousand tracks make a product.

No, but it’s not like it DOESN’T Exist, though. I have Tidal, and I look for the “M” after new titles when they are released every Thursday. What bothers me about MQA, is that while Tidal’s software tells you that you are playing an MQA title, what I don’t see any indication of what is the actual bit depth and sampling rate of the selection to which you are listening. Same is true with the AudioQuest DragonFly Cobalt the dragonfly logo glows purple when a MQA file is encountered, but again, is it 16 or 24 bit? 48 KHz? 88.2 KHz,? 96 KHz? There is nothing to tell you to what you are actually listening!

[gmgraves]

3 hours ago, esldude said:

No the condenser will have the better transient response.  Ribbons while light have to move more.

That seems counterintuitive to me. Microphone ribbons of the type I’m thinking of  weigh less than the same size feather! In fact they are so massless that if you blow into one incorrectly you’ll tear the ribbon! Obviously, I’m not talking about an RCA 44BX or a 77DX. You’re probably right about those.

[gmgraves]

12 hours ago, Ralf11 said:

whether condenser or ribbon, shouldn't the mass of the transducer approximate the mass of the ear drum??

I don’t see what one thing has to do with the other. The ear is in an “interpretive” mechanism, meaning that the brain interprets the sounds that fall on the ear, making what one hears a combination of physical characteristics and what the brain makes of those physical characteristics. A microphone, OTOH, is merely a mechanism with a measured set of physical characteristics bestowed on the device by engineers who will do whatever they need do in terms of materials, techniques, and operating principles to achieve their desired result. For instance, condenser mikes with the best bass have big diaphragms, often over an inch in diameter (the stereo mike pictured next to my posts, for instance, has a pair of diaphragms that are 35mm in diameter.). Certainly they are definitely not ear-drum sized!

[gmgraves]

9 hours ago, esldude said:

So how can ribbons have better transient response if they are so truncated in frequency response?

 

OTOH, not all are. Like this Samar MF65 ribbon.

Up to about 10 KHz this mike does very well (obviously omni or figure-of-eight. No cardioid could ever do a flat bass response like that). Even that slight peak at 10 KHz is insignificant, but that -10dB negative peak at around 16 KHz is difficult to account for as is the rising response above that. At US$2000 each, this Chech-built beauty certainly ain’t chicken feed!

[gmgraves]

50 minutes ago, esldude said:

It is only a -5 db dip there.  So not great, but plenty of mikes do worse. 

Yeah. -5dB. Not so bad, but the second peak above 16 KHz is still puzzling to me.

[gmgraves]

5 hours ago, John Dyson said:

And my heart *breaks* about Linux taking over...   Oh well, deep in my past, and now I use it daily, and wouldn't nowadays know how to boot FreeBSD (even though I wrote much of the original kernel).   I wouldn't mention anything except I *REALLY* did develop a lot of OS code, even FreeBSD with AT&T Bell Labs official permission, while working at AT&T Bell Labs as my full time day job...  In fact, Bell Labs had offered to support my development as a job, but I opted for a Bay Area tech boom job instead...   (For people who know the history, it might sound strange because of the lawsuits between AT&T and the BSD community -- but I was in a very special position, and had a hell of a lot of fun until I quit.)

 

If conditions are correct for its use, Linux (or any Unix structured OS) is the most amazing platform for raw/hard-core/no-nonsense productivity.

 

*Was gonna digress with a long history and diatribe which is of ZERO consequence in this discussion, with mildly negative personal sentiments about Linux, except that I also mildly endorse it as a very good OS*    Consider the 'starred' comment as a kindness that I didn't waste any more bandwidth on a boring and off-topic history lesson :-).

 

John

 

I feel that way about Mac OS X. It’s UNIX with the slickest of GUIs on top. And the Command Line is just a click away if you want or need it!

[gmgraves]

1 hour ago, John Dyson said:

They DID originally start with the FreeBSD userland for supporting the effort, I don't know how much of FreeBSD userland is left.  Their kernel (AFAIR) is their own, but has a POSIX (Unix-like) API.  (I'd suppose that they started with the FreeBSD userland because one of their MacOS people was on the FreeBSD core team, Jordan Hubbard.   I was an early member of -core also, which was a council of about 8-12 or so lead developers for decision making.)

john

[gmgraves]

1 hour ago, John Dyson said:

They DID originally start with the FreeBSD userland for supporting the effort, I don't know how much of FreeBSD userland is left.  Their kernel (AFAIR) is their own, but has a POSIX (Unix-like) API.  (I'd suppose that they started with the FreeBSD userland because one of their MacOS people was on the FreeBSD core team, Jordan Hubbard.   I was an early member of -core also, which was a council of about 8-12 or so lead developers for decision making.)

 

John

 

It’s been a long time, but I seem to recall that the. Kernel of OS X was the one used in the NeXT  computer. NeXT was BSD UNIX. One thing that the UNIX core of OS X does is make the system bulletproof. While the Mac interface is logical, fun to use, and robust, what makes it really powerful is the easy access to the CLI.

[gmgraves]

5 hours ago, Jud said:

 

Since "improvement" is to some extent personal, that's always difficult to assess. At minimum I would have to understand what the difference was that listeners were being trained to hear, and see whether it corresponded to greater or less distortion (or possibly neither). If hi res corresponded to less distortion, in that event I'd have to decide how much the slightly greater distortion bothered me, keeping in mind there are psychology experiments that have shown we can sail along without conscious realization of a situation that subconsciously is causing us enough anxiety to literally break out in a cold sweat.  (See https://en.wikipedia.org/wiki/Iowa_gambling_task .)

Ain’t that the truth! People buy expensive interconnect cables and misinterpret some frequency response anomaly wrought by the ‘cable’ as an “improvement” (remember, a cable is passive, it cannot add anything to a signal, it can only subtract from the signal passing through it. If it effects a change in sound, it’s because it has altered the frequency response by attenuating some portion of the pass band).

[gmgraves]

7 hours ago, sandyk said:

 

 OR, it is not adequately screened against RF/EMI , or is located alongside sources of RF/EMI instead of passing across them at 90 degrees.

I’m simply not convinced that’s a real concern. Maybe it is in some rare instances, but I have a 100 MHz ‘scope and I’ve looked at audio signals on hundreds of interconnects, ranging from the cheapest of throw-always to the most expensive of boutique cables. And I’ve never seen any RF interference, of any important amplitude, on any of them. Any extraneous signal that I’ve seen on any of them has been down in the microvolt range, and in some cases, the nanovolt range. On a cable ostensibly carrying an audio signal with a nominal amplitude of one to two volts, I simply do not see that signals that small could have any affect on the final sound of a component. And mind you, finding any extraneous signal at all is very rare in my experience.

Maybe it’s different in OZ, but I can’t for the life of me imagine why it would or should be.

[gmgraves]

22 hours ago, John Dyson said:

The noise might be rare -- maybe not.  Just looking at the DolbyA cat22 circuti with protective rolloff and filtering -- they don't do it for fun.  In fact, if doing a real professional microphone amplifier -- I'd certainly provide some EMI filtering on it, even being balanced.

 

Truly, I don't know if people can hear the noise -- but there is some anecdotal evidence there there is noise coming from somewhere.

 

There is too much consideration for interference in quality circuitry for the issue to be non-existent,  Also, analog circuitry often becomes disappointingly nonlinear as frequencies increase - with the audio quickly becoming susceptable to noise sources well outside of the expected spectrum resulting from out of band signals and IMD in the circuitry.

 

My tinnitus filled ears cannot detect the amazingly low noise and distortion that they used to -- but it would be a REALLY interesting exercise to do a well controlled experiment on the noise leakage, analog cabling differences, etc.   If there are applicable studies, doing tests in a contemporary computer environment doing audio processing and presentation, it would really be interesting to see the results.

 

There are apparently some steatlhy issues with signal integrity, and might be help for someone to do a study some day.  My own bias suggests that significant quality issues from audio signal cabling differences don't happen very often -- but certainly can happen, and given the integrity of people that I know who might make note of the issue -- very likely does to some degree.  Leakage of noise coming from poor PCB layout/shielding or even gross impedance mismatches/reflections can certainly happen also.

 

Skepticism is a good thing, and I simply cannot be a denier.  I was certainly around in the TIM days -- amazing the very competent people who might be naysayers.

 

John

 

I said that the noise was, in my experience, rare. I didn’t say that it was non existent.

And I‘d like to make a comment about Dolby A. I had two channels of Dolby A when I was using a pair of Otari MX5050s (half-track, 15ips) and doing analog recording. I’ve read where a lot of people have trouble getting Dolby A to “Track” perfectly. I always put calibration tones at the head and the tail of every reel, and I never had that problem. I also might add that I always meticulously aligned and set-up my tape decks before every recording session, and before actually making dubs of the tapes. 

[gmgraves]

26 minutes ago, sandyk said:

Hi George

 It could be an interesting exercise, given that your tapes have calibration  tones, to see how much difference John's highly accurate S/W decoder could make to the sound of your recordings , assuming that they were recorded at 16/44.1

 

Kind Regards

Alex

Well, the tapes have been transferred to 16/44 a long time ago. Most of the symphony tapes to which I am referring were recorded on Ampex Grand Master brand tape. Unfortunately, this tape had a “friction” coating applied to the back-side of the tape base, and this coating has long since turned to goo sticking the layers of tape on the reel together making them unplayable. Luckily, most of the tapes were transferred to DAT (and since to CD).

[gmgraves]

23 hours ago, John Dyson said:

You are 100% right about aligning the decks.  Any error is bad, and if you spend time making sure that your decks are REALLY flat to at least 15k, and everything is within about 0.1dB or better, then there is ZERO real problem.  There is still the fog, but the DolbyA 'fog' isn't really all that bad sounding -- it is just foggy :-).  I don't know how bad head bumps really are -- but even keeping the LF band well aligned with respect to the MF band helps to mitigate the generated LF distortion.   In the middle signal levels, there is a big bump about 100Hz in true DolbyA (and DHNRDS) -- that bump comes from a relatively high Q filter (about 1.070) at about 74Hz.  When the gain of the LF band and the MF bands aren't mutually correct, that 100Hz bump wobbles a lot, and creates an effect that sounds something like conventional distortion.

 

My experiments show that the absolute best sound is obtained when there is less than 0.05dB error in calibration.  Tilt in frequency response on the high end can also make sibilance a bit strange -- you can hear it in some of the Karen Carpenter recordings.   Balancing the response between 6kHz range and 9kHz range is especially important.  Alex and others with good hearing beat me up all of the time (less so lately -- learned my lesson) about sibilance.

 

the LF bands are low enough frequency that can signficantly modulate the gain -- even in the MF band, so a good deck with relatively good phase behavior and good LF behavior is helpful.  (The attack/release is fairly fast 4-60msec variable attack, 60-120 variable release time for the 0-80 and 80-3000 bands.)

 

John

Well, while I appreciated the increase in S/N afforded by Dolby A, I noticed later when transferring the tapes to DAT that those masters captured WITHOUT it (and in spite of he increased tape noise) SOUNDED better.