Audio Overkill? AP Article Lands
In early April, I received a phone call from an AP reporter named Ryan Nakashima. He told me that he had been assigned the task of writing a story on the merits of “so-called” high-resolution audio and wondered if I could make myself available for an interview. I was glad to spend an afternoon with him in the studio explaining high-resolution audio in general and demonstrate the misrepresentation of most tracks on sites that offer high-resolution music.
The article came out this morning and it’s worth a read. You can check it out by clicking here. It’s well researched. Ryan took a great deal of time to investigate various claims and counter claims regarding high-resolution audio. He talked to a PonoMusic customer that had downloaded Bob Dylan’s “Shadows In The Night” and also owned the CD version. To his amazement, they sounded identical and he pressed the Pono site for a refund.
You can guess what satisfaction he got from Pono and their Vice President of Content Acquisition Bruce Botnick who is also a very highly regarded audio engineer (he worked extensively with the Doors). According to the article, which quotes Bruce, purchasing a “high-resolution audio” download from the PonoMusic site is “a case of “buyer beware”. Until the provenance of each and every track can be researched and verified, customers may want to stay on the sidelines when it comes to older analog recordings remastered to large bit buckets. Especially when it comes to PonoMusic because virtually all of their catalog comes from ripped CDs…not too hard to figure out what the provenance of those files are.
The major labels and download sites could easily put an end to the many articles that express doubts about their newly issued “hi-res” files. All they have to do is properly identify the production path for each album as it travels from the tape vault to a 192 kHz/24-bit PCM WAV or FLAC file. It may be challenging but it’s not impossible. For example, anything recorded prior to digital audio was likely done on analog tape…just say so.
I’m going to write about my experience at the 25th Anniversary celebration of Sony Music Entertainment’s Legacy Recordings last Wednesday evening in New York City. It was a wonderful experience and deserving of multiple posts thanks to the very engaging presentations by their engineers and John Jackson of Legacy Recordings. But I thought I would share a brief comment here because so much of the material that is being issued as high-resolution music comes through the detailed processes that engineers Mark Wilder, Vic Anesini, and Matt Cavaluzzo do at Battery Studios. These guys and the facility have the expertise and equipment to produce the very finest versions of classic albums. We got the chance to hear a lot of tracks as they make their way through the process. And they sounded amazing! I was especially impressed with the original 3-track of “Take 5” by Dave Brubeck and “The Sounds of Silence” by Simon and Garfunkel.
Figure 1 – A visit to Battery Studios listening to Roger Waters “Amused To Death” with Matt Cavaluzzo at the console.
And Legacy Recordings’ John Jackson described the end result as “high-resolution transfers”. I love it! This just might be the right way to describe all of the reissued standard-resolution tracks that dominate the music retailers. They sell “high-res transfers”, which are distinctly different from “high-resolution sources” or “high-resolution music” releases such as the new James Taylor project or my own stuff, which originate as high-resolution projects. I’ll talk a lot more about this idea as I get time. I spent some time on the plane last evening crafting a logo that might work to identify the work that is being done in the mastering rooms at all three major labels. The work is meticulous and the output “hi-res transfers”.
The AP article is really good. When highly regarded engineer/producers like Giles Martin say, “You can’t upscale audio. There’s a compromise in having huge high-res files that don’t sound any different than other ones,” it puts some credibility back in the business of better quality music.
I’m encouraged by some of the new developments. Transparency, information, education, demos, and guidelines for engineers and producers might be able to get the high-res initiative back on track. I said might.
I have accepted that you will keep hammering Pono.
Do be inclusive to mention that while two million is a big number, somewhere between 10-15,000 of the tracks are hi-res masters at 24/96 and above. The dumping on analog tape has two sides since I just read a big article concerning the high degree of care being afforded to the transfer of precious, older tapes of great music.
It is also reasonable to ask that you do make clear to folks that purchase of a Pono player in no way obligates using the Pono Music service. Your material, HD Trax and all the other sites which have hi-res downloads available can be used to load the player if someone is averse to their service. Level playing field department…please, no stacking the deck.
The numbers that the three big labels contributed to their panel in NYC this week totaled about 5000. The 10,000 to 15,000 includes all of the smaller labels that HDTracks supply. They are actually not Hi-Res but transfers of older content. Analog tape is wonderful but when you see what the engineers are doing at places like Battery Studios, you realize just how fortunate we are to have any tapes…many are worn out and some won’t play at all.
You think I’m harsh on Pono…here’s a section take from the Bob Lefsetz newsletter. This one is entitled “Failures”…
“PONO
An unmitigated disaster.
First there’s the threshold issue. Do we need better sound? No, let me put it more succinctly, can we hear better sound?
Try this test, it’s fascinating:
“How Well Can You Hear Audio Quality?”: http://n.pr/1ALMuFQ
We had to listen to the ravings of Mr. Young for months as he trumpeted a product that almost no one wanted, other than the early adopters who purchased it on Kickstarter, and there weren’t many. If, like with Garth Brooks, the business had been a success, we would have heard about it, these stars love to get their flacks to crow about their wins, and the press eats it up, but we haven’t heard a peep since the launch. And we can criticize the dearth of available material, i.e. how much can be gotten in hi-res, but come on…
You’ve got a misshapen box playing overpriced files in a world where everything goes through your smartphone and access is key, ownership is fading away. Furthermore, people expect it all for one low price a month. Introducing Pono and its store is like selling premium gasoline to Tesla drivers, assuming everybody has a Tesla, and anybody with cash has a smartphone.”
Bob’s newsletter reaches over 40,000 people…
I’m just reporting the facts. People deserve to know.
Way over-heated analysis. John Atkinson’s Pono overview and player review in Stereophile are the only articles I’ve read that did not obviously have a poorly concealed agenda and bias behind them. Why don’t you quote some of the words he wrote too?
And, just for fun, his follow-up article re: accountable journalism v. adaptive journalism should be used to understand that 85% of the attacks etc., are from folks who wrote the travel article, the cooking article, the cigar article, etc., or even some audio journalists who thrive on being sh-t disturbers to keep their by-lines active.
Insight from only one vantage point does not exactly demonstrate wide-angle awareness, just the opposite in fact. I have no trouble understanding where you are coming from, and it would be really nice if just once in a while you would acknowledge that there in fact two sides to every coin, although I’m not holding my breath either.
Find me a player that costs 399.00 and sounds even close. I have the 300.00 Sony here, it gets badly beaten, and the pricey A&K’s have very little if any sonic advantage. I personally do not find the shape problematic either.
And if Pono were to go away, are you thinking your sales will explode or something?
Mark you mentioned “…I was especially impressed with the original 3-track of “Take 5″ by Dave Brubeck and “The Sounds of Silence” by Simon and Garfunkel”.
Wasn’t this originally recorded on analog tape? And yet after being impressed, you feel that the performance of it cannot rival or out do “high resolution digital”?
I’m confused! 😉
The whole evening involved playing back high-resolution (96 kHz/24-bit PCM) files on digital audio workstations of original master tapes or copies of analog tapes. I’ve never said that analog tape can’t sound terrific. It can’t produce the same fidelity as high-resolution PCM but maybe it doesn’t have to. The recordings I heard sounded wonderful.
Oh wow, I’m going to fall over, you actually included analog tape in the same sentence as the word ‘wonderful’.
Even on a tech basis, the Ampex ATR-100 running 1/2″ tape at 30ips which we used exclusively over 20 years of live to two track recording was flat to 50khz even well past 0db, signal to noise at that speed is vanishingly low and we never recorded over a used piece of tape. The mics were some of the most prized then and today.
Judgement by specs has been proven to be among the reddest of audio herrings; we’ve all heard units that had impressive specs but failed to engage. And no, I’m not saying that something can measure poorly but still sound great either. Good specs are an indicator that something might be expected to sound good, but not a guarantee that it will.
No question, ‘wonderful’ would be the best adjective to describe the recordings we made on that machine, so it’s nice to see even a begrudging acknowledgement. Find me a CD or LP that is an audible clone of the master tape; such does not exist. But as you say, transferring a new or older master tape,of analog or 24/96 digital origin, does give us a virtual clone of the master, or vanishingly close. This is the beauty of hi-res; provenance is not what we hear when playback commences, but instead the product of transfer and mastering to a medium that does not have the limited signal capacity of CD or LP. That is what counts to the ears, no more and no less.
Ampex AtR 100s are very fine machines…and once thought to be the ultimate recorder/reproducers. But you might be surprised to read about these machines at ATAE. Fred Thal, an acknowledge expert in the world of analog tape machines. His assessment of the ATRs? Read the whole article by clicking here.
“In the USA countless retired recording engineers, fiercely loyal to the Ampex brand, assert that the ATR was the finest tape machine ever built. Respectfully, we must disagree. The ATR makes a poor candidate for building a state-of-the-art reproducer. Here are a few of the reasons why. The ATR has fixed-pin lifters which are damaging to tape. Its transport is a force-guided design, incorporating tape-damaging edge guidance at its tension sensors and in its headblock. To achieve dynamic azimuth stability, the ATR’s headblock forces the tape to travel a reverse-camber path with three ceramic edge guides that create high scrape flutter. This scrape flutter is not fully attenuated and it can be audible (see footnote). Additionally, the ATR’s un-clamped capstan servo tape drive (which eliminates a pinch roller) requires undesirably high tape tensions for proper operation.”
Good machines but not the best.
You and plenty of others like to bypass specifications when they prove troubling or run counter to your own beliefs. I’ve worked with analog tape and love the catalog of great recordings that have been done using it. It’s not difficult for me to shout out that fact…no hesitation from this writer. But it’s not capable of meeting my needs for high-resolution recording. The engineers at Battery Studios aren’t transferring back to analog tape…they’re using a 20 year old ADC to 96 kHz/24-bit PCM. There must be a reason for their choice.
CDs that capture all of the fidelity of an original master do exist. There are many CDs from MoFi that accomplish that feat. Moving on.
The HRT idea is even better than the idea presented yesterday. As the XFR logo suggests: let listeners be excited by the “quality” of the original recording; which doesn’t conflict with advocating JAS specifications for today and the future. The JAS and HRT “specs” are compatible with a slight upgrade in the “Made for iTunes” spec.
BTW The other commenters may want to remember that a well-engineered analog or DSD recording (eg Blue Coast) is far better than a 96/24 PCM done with a cheap USB microphone on a laptop. What PCM allows is the full signal quality to be always available.
I think the HRT or high-resolution transfer might be acceptable to the major labels. We’ll see.
Brubeck … Desmond…..lead me to anything 3 track and beyond ! Somebody please do Brubeck. His later recordings actually included one DSD I believe ? Paul always sounded good standing an arms length from the left side stereo mic low bandwidth and all ! Guess we can attribute that to musician’s dynamics !
THe “Take 5” recording sounded terrific. Would it have sounded better if I have engineered it with my technique and equipment…yes, it would. The piano especially is a little dull and distant…but it was a great listen.
Too bad Mobile Fidelity trademarked ‘Original Master Recording’, that would have been perfect and wouldn’t have included any mention of High Res anything.
Mark, Do you know anything about the provenance of James Taylor’s new Before This World release. Can’t find any info on how it was recorded.
I enjoyed the music but don’t see myself listening to it once a month or more as I have his classic 70s stuff for the last 40 years.
I’ll be writing about the James Taylor “Before The World” in depth soon. I sat down and spoke with the producer and engineer while I was in New York.
Your daily blogs have inspired me to reconnect my old VHS player to my new 4K TV. Why? Because the content on some of my VHS tapes now exists nowhere else.
As we speak, I’m comfortably seated, waiting for the player to warm up and display that 230i content on my “big-container” display.
Should definitely be better, considering the technology this TV has.
Great idea…in fact, you’re doing the very best playback of your videos. None of this transfers or remastering to UHD…stick with the tried and true VHS deck to give you the full analog experience we all loved in the 80s. I don’t think I have a functioning VHS deck.
Just wondering if you saw the comments to the AP article from HDtracks about their quality control? Are their competitors checking quality this closely? It appears only frequency content is being checked.
Before I purchase a track from any store, I check the dynamic range database first. The Simon & Garfunkel albums from HDtracks sound much better than the CDs remastered in 2000. The Aerosmith tracks from HDtracks sound quite compressed compared to the CDs remastered in 1993. I think all the high-resolution music stores need to insist that the record companies give them files that have dynamic range and good frequency content. A direct tape transfer with no mastering would be perfect.
I did. And I spent way too much time correcting the comments by Bob Rapoport…his explanations were completely wrong. The issue of QC at these places is challenging…it takes a lot of time and experience to be able to sort out the facts. Even the DR database is flawed. They scan the entire file including the fade in and out thus skewing the results to increase the actual dynamic range.
Mark, you’ve made this comment about fade ins and fade outs before, so I looked up the documentation for the software. It appears the algorithm is comparing the peak with an RMS value calculated from only the top 20% of loudness measurements – representing the average of the loudest passages in the track. It is, by no means, a perfect system, and doesn’t measure what you are looking for to determine whether a recording qualifies as high-resolution, but fade ins and fade outs shouldn’t affect the results.
Thanks Andrea…but I have to wonder why I get different values when I avoid the fades? It doesn’t change things by much but it does change. I prefer to use the analysis tools included in Adobe Audition.
Including the fades will change the distribution slightly, but DR scores are supposed to be rounded to the nearest whole number. It would be a very odd track where the fades could cause a change of that magnitude.
These are not big changes but they do happen.