Tape is in good shape, just some riffling on the edges that any decent machine will take in stride.
Wilder knows the tape and the catalog and he's really good. Enough said.
Matching the sound that people are used to only makes sense if the sound they are used to is correct. Otherwise they should have remastered it 1/4 tone off-pitch. Accurate playback never undermines the validity of an archive or a remaster, it's the remastering choices that can screw it up. Which in this case clearly didn't happen.
I guarantee Mark would never just "antique" a tape playback for marketing purposes. He's too good. Running something through tubes is not why Mark may (or may not) have used a Pultec, it's simply because that's what was used on the original, and Laico knew what he was doing, and to get it to sound right you gotta use the same filter curves. If those curves were readily and exactly available in top-notch DSP (they're not) then that would have been fine. That's not "thinking too much as a scientist or engineer" that's thinking as somebody who has never yet seen a case where a faulty or diminished playback yields more aesthetically valid results compared to a more accurate rendition of the recording's original intent. And Wilder didn't play it on an Ampex 300. QED. He knows that the ATR is more neutral, and more of the music will come through.
We tried to get in with Battery on this one and missed, not sure why - and pitched hard for it, on the premise that we might be able to pull slightly more from the tape with our electronics than the 1975/Spitz ATR, as we have done elsewhere. And that the tape speed issues would automatically correct themselves, that part is a no-brainer. Mark expressed the reasonable disinclination to play the tape again... the producer OTOH actually states that often he favors the lower-fi vibe of his recollection than going for precision audio. I think he suspects that if it's too pristine it will be less emotional. I don't see how that's the case, seems to me that whatever takes the intervening noise and distortion out of the way of the performance is to the good. Above my pay grade. He gets good results, and is well-regarded and on this point we simply disagree a bit.
But there's no way a slightly clearer tape reproduction would have thrown anybody out of the record. And again, 90% is the mastering skill - obviously having the best raw source is important. And a top-notch playback with speed correction is not going to circumvent the art and tech burned into the tape, it will simply convey it more exactly. And then Mark does his thing, and the result is beauty.
Another thing I think is commonly romanticized about the older equipment and which is misguided is the notion that their guts are mystical, and that you gotta have these because they sound "better" than anything contemporary, which is nuts. The reason the vintage gear is popular (maybe too popular) is because it's the stuff that has survived because it was good. API 550s would not have worked on this record because the Q is sharper and the frequency points are different, not because of the transistors. It's a compatibility issue, in a sense. You simply could twist the API's knobs until doomsday and not match the Pultecs. Neither is better than the other. They're different, not so much because of how they achieve gain, or the circuit topology - but because the knobs sound different when you twist them. And unfortunately most of the recent digital stuff doesn't really operate like the old stuff - because the old stuff was designed with a slide rule based on what sounded good to the engineer, not the other way around. A lot of the effects devices of that era were designed by ear, and the curves were purpose-built for the studios and styles of the day - they worked. They call it "work" for a reason. These were working guys, listening and twisting knobs until it "worked". The Rupert Neve 80xx equalizers were broad as hell, and had really useful carefully chosen frequency points (by real mixers) and that's why they sounded good and have survived as classics, not because they were using slow 3055's driving a transformer.
In the case of this recording, where Teo and Laico made a great effort to be as realistic as possible, and where some correction had to have done on the original LPs - because nothing matched the control room without more "work" done on it - you gotta assume that they were the best judges at how to get the control room sound onto the vinyl. And using a Pultec or LA-2A or whatever doesn't do that because its tubes, or its vintage, it does that because that's the tools they used, those are the filter curves they were hearing. Recently I've been trying to figure out how to un-hype the elevated top of a Bones Howe recording, wherein the tones are perfectly spot-on. Notwithstanding the fact that I don't have the exact mastering gear he and the mastering guy did, it's possible to mock up a system that is similar to what they might have had in their homes when listening to the check-lacquers that this EQ master is a derivation of. And listening on a pair of AR 4xs lent great insight into what they thought sounded right, because on those speakers it sounded like it should - "by the tones" on an Plangent/ATR - only if the tweeter level was backed off a quarter turn. Which they probably did to make sure they didn't release something too dull for AM radio. So there you go, Reverse engineering that transfer-function with a DSP equalizer came very close to their original intent on modern equipment, and when it got to be exactly what it took to un-ring the bell of the AR speaker (considered as a "filter") suddenly the tape sounded right, all on modern gear. A lot of cross-EQ but it wasn't done by going back to ancient equipment So there's where the use of vintage equipment can lend insight into original intent without being strapped with the obvious downside of lower-fidelity gear. . And probably if I could mime the exact equalizers in the mastering chain i'd do even better. And probably the easiest way to do that would be with a Pultec likely to have been in the mastering chain on the day it was cut - but not because it's tubes, but because that's probably what was in the room in 1965 as Bones and the mastering guy corrected it for release.
On Nov 20, 2013, at 9:57 AM, Tom Fine wrote:
> Hi John:
>
> The rationale was three-fold:
>
> 1. play the tape as little as possible because it's not in good shape
>
> 2. try to match the sound to what people are used to. This album has a huge "sonic memory" out there with the buying public and critics. If the reissue team didn't reference the original LPs for the mix and the general sonic ambience, they'd get slammed by critics and careful-listening fans. Those who don't "get" or accept this rationale are thinking too much as a scientist or engineer and not enough as a music business person. One has to understand the market, plain and simple. Deviating too far from what I call a "sonic memory" is toxic to sales. I can cite numerous examples.
>
> 3. given that Mark Wilder has a long track record of excellent remasters that sell well, I tend to trust his judgement to mix outside the box. In my own experiments with using, for instance Izotope's "mastering EQ" plug-in in Sony Soundforge vs. going back out to analog and using my Great River mastering equalizer, I always prefer the Great River. I do not believe that DSP has gotten as good-sounding as the best analog gear for "sweetening" something to an individual's taste. I'd trust the Great River or my trusty Pultec equalizers anyday over any DSP I've heard. I would also trust Mark's analog mixer over, for instance, the ubiquitous Protools mixing interface. Again, this is not worth arguing if someone is a hardcore believer in doing these things in the box. Tomato, tomahto. Wilder and Sony have a proven track record of sales to back up their work methods.
>
> I also think they cut new LPs out of the mixdown process. Like the LP niche-renaissance or not, cutting an LP that will retail for $30 with a $10 profit for the issuer is a good business move for a popular title like this.
>
> Regarding why the mono -- purely a marketing choice. There are many fans out there, myself included, who prefer the mono version of this album and are thrilled to have it in a high-resolution new version (I'm probably not alone having worn out my original mono Columbia LP). Like the mono Beatles albums, at least up to Sgt. Pepper -- and we can debate whether this is true all the way through their last mono offering, the White Album -- with the Miles Davis small-group albums of the late 50s and early 60s, the music hangs together better in the mono mixes. I immediately notice that I hear complete ensemble songs instead of compositions of well-played sounds coming from three different directions. I feel the same way about the Blue Note small-group records up to the mid-60s. Small-group jazz does lend itself to close-in mic'ing in order to get all the details of playing. But, then taking those close-mic'd signals and building them into a non-distracting stereo image took some learning and practice. It still wasn't right in the 70s, when you'd have close-in wide-spread stereo mic'ing of the piano (inside the lid, invariably) and drums, but have single-mic'd horns then spread across the soundstage (think of Pablo small-group records made at A&R or Group IV studios). It was unrealistic, like one's head was simultaneously inside the piano, drum set and yet out in the room to hear the horns spread across a plain. One can get used to it, but an excellent mono mix keeps just the music and the ensemble front and center.
>
> Given how terrible the original CD reissues of Sony's jazz library sounded (including "Kind of Blue"), I am thrilled that lessons were learned and budgets are being used to get us modern remasters that are faithful to the "sonic memory" but also crisp and clear and dynamic as is expected in a modern setting. To my ears, the "Kind of Blue" remasters keep the balance and vitality of the original issues but remove several layers of fuzz and gauze, so you get the same instrument tonality and mix approved by the original team, but it's now like you're hearing it out of the original signal chain instead of behind a layer of cutting-master tape smear and LP fuzz. Oh, and this stereo version is speed-corrected, unlike the original LP.
>
> -- Tom Fine
>
> ----- Original Message ----- From: "John Haley" <[log in to unmask]>
> To: <[log in to unmask]>
> Sent: Wednesday, November 20, 2013 9:30 AM
> Subject: Re: [ARSCLIST] The new "Kind of Blue" remasters explained
>
>
>> Of course they should have used the session tapes, not later mixdowns.
>> That's a given. The part I don't get here is doing DA and AD conversion
>> just to use the analog mixer, if I understood that right. The resulting
>> 192/24 signal has thus been unnecessarily converted twice already and
>> subjected to a bunch of old analog electronics. Also, I don't get the need
>> for a mono version derived from the same tapes, and personally, instead of
>> that I would much rather have had a three-track SACD version, which they
>> have precluded, but that's me.
>>
>> Best,
>> John Haley
>>
>>
>>
>> On Wed, Nov 20, 2013 at 6:44 AM, Tom Fine <[log in to unmask]>wrote:
>>
>>> https://www.youtube.com/watch?v=z_p7Qbb_LAo
>>>
>>> Miles Davis "Kind of Blue" archival transfer made by Mark Wilder on an
>>> ATR-100.
>>>
>>> This text (below) copied from the Kind of Blue description at HDTracks.com
>>> (HIGHLY recommend the new 192/24 downloads of BOTH stereo and mono, they
>>> sound fantastic): The new mono mix is also in the new Miles Davis Mono CD
>>> box set. As I understand the description below and in other interviews with
>>> Wilder and Berkowitz, the 192/24 transfer from the 3-track was a straight,
>>> high-quality NAB playback. Then all remixing and remastering was done by
>>> bringing the 3-track high-resolution digital back out to analog, mixing and
>>> processing using analog equipment, and then back to a 192/24 stereo (and
>>> mono) master.
>>>
>>> -----------------------------------------------
>>> Kind of Blue Becomes Digital, by Engineer Mark Wilder
>>>
>>> "Since the Kind of Blue mixed masters are multiple generations from the
>>> original (due to excessive play/wear), we decided to go directly to the
>>> original session reels. Not only does this put us at the original session
>>> as a starting point, but it also allows us to deal with the pitch issue as
>>> well.
>>>
>>> The three, 3-track half-inch tapes are in good condition, but age has
>>> force them to "scallop" a little, meaning that the edges curl away from the
>>> tape head. This changed the initial focus from mixing from the originals to
>>> archiving them before mixing and working from the archive files. This
>>> allowed us to gently guide the tape against the playback head to get
>>> optimal contact and fidelity.
>>>
>>> The archiving was done at 192kHz/24 bits, played from a modified Ampex ATR
>>> 104, and hard-wired to HDCD Model 2's directly patched to a Lynx 2 sound
>>> card.
>>>
>>> An upside to working from the archive files was the ability to chase the
>>> original fader moves done during the mix in 1959. We constantly compared to
>>> an early pressing - mono and stereo - and worked bar by bar to duplicate
>>> the level moves on the three tracks to match as well as possible.
>>>
>>> Each channel was converted to analog and passed through a GML mixer,
>>> bussed to stereo or mono - depending on the release format - and converted
>>> once again to 192Kc/24 bits. At the GML, we inserted processing where
>>> needed."
>>>
>>> - Mark Wilder, Battery Studios
>>>
>>> ----------------------------------------------------------
>>>
>>> -- Tom Fine
>>>
>>
|