Discussion in 'Music Corner' started by Steve Hoffman, Nov 30, 2007.
I apologize. Dear Host; how are we to interpret your bumps?
That's a rather personal question! Someone wanted it bumped, and I'm home because of our 6:00 PM curfew so there ya go.
Thanks Steve. This is all quite interesting and while I don't necessarily understand it all, it does illuminate some things that I think about from time to time. I'd like to add a wrinkle here from my own experience. I've done a modest amount of studio recording on guitar, electric bass and primarily double bass. Some of this was a home studio situation with pro tools, some other was a more modest professional situation and a few hours were at Sound Techniques in Boston back in 1996 or 1997. My sense is that no recording sounds exactly like what the musicians hear while they are making the takes (I realize that I'm speaking of the studio sound vs. the analog master sound here). Some get very close, so this begs the question: what does the best sounding recording even mean?
Is it the one that sounds most close to what things actually sounded like when the take was made or is it one that brings out something more? Obviously, the answer to this question is quite subjective, but to illustrate my point, I'm going to refer to a few John Coltrane releases that were recorded late in his life, but I have on "20-Bit Super" mapping for CD. As a double bassist, I restrict my comments to that reproduction. The releases are: Living Space, Sun Ship and Stellar Regions.
The first thing that strikes me is that Garrison's bass is massively round and deep. And to be honest, it sounds great to me. But there's no way that any of the musicians or engineers ever heard the double bass like this. Jazz double bassists were still using gut strings for the most part then (if my memory serves, steels first came out in the mid-'60s but weren't mainstream for another few years). And even if Garrison had been using steels, they would not have sounded like they do on these 20-bit recordings when laying down the tracks in the open air.
So, my question is (if I have a question at all): What is going on here? Is 20-Bit Super Mapping bringing out something that the master tape captured but we weren't hearing when actually laid down, or is it some form of enhancement that creates depth in the bass part that wasn't really there to begin with? I tend to be a "I want to hear it like it sounded at the time" type, so I'm curious.
Boy, what if I were to ask a truly personal question! Thanks.
And thanks for all the great music. I still have a number of your masterings (regular labels, DCC, etc.) that are just terrific.
Super-bit mapping means nothing. The engineer either recorded it like that or the mastering engineer fiddled with it, or both.
CDs will always be 16-bit resolution.
In the first half of the 90’s, 20-bit A/D’s (analoge-to-digital) converters became the new standard in the professional record industry. Everything else equal, a 20-bit A/D converter will convert analog master tapes more accurate with more subtle musical details into digital, compared to a 16-bit A/D.
Instead of just remove the 4 least sigificant bits of the 20-bit digital studio master to match the standard 16-bit CD format, they came up with a techique that used noise shaping to fold down more of the musical information from a 20-bit digital master to a 16-bit CD. Now I have explained Super Bit Mapping to you. Bit Mapping = Noise Shaping.
Would you conclude now that 24 bit digital brings you as close to the master as a record?
On my home system, SACD has the most analog feel of any digital format I’ve ever owned and been able to play regularly. I’m not even sure whether my SACD player does pure DSD, or converts the signal to PCM, but yesterday I was listening to the Mobile Fidelity SACD of Carly Simon’s No Secrets, and it has all the goodness of analog, with stunningly beautiful and pure reproduction of Carly’s voice, but none of the hassles of vinyl, i.e. no popping, clicking, warping, record cleaning, stylus cleaning, etc. to deal with. I’m a huge believer in DSD and SACD.
I thought the question percolating here is comparing DSD (as sampled for SACD) to analog. I was under the impression that our host does not find the 24bit PCM gets close enough to analog. I hope I am not putting words in his mouth.
SACD sounds less incisive to me. I listen mostly to classical. Where SACD does clearly excel is on the violins and violas. But it seems bit rounded off on transients, like cymbals. Could be that my CD player, Marantz 8004 isn’t really That great on SACD. It is exceptionally good on red book.
I don't know that model. But some SACD players actually just convert SACD to PCM. In which case the value of SACD is rendered pretty much moot.
I don’t think the Marantz converts to PCM. The CD layer, which is normally a DSD master downsampled to PCM, definitely sounds different from the SACD layer. On classical, I normally do prefer the SACD layer, but on most rock/jazz/contemporary, I prefer the crispness of the red book layer. Again, that’s on my CD player. It might well differ on other players.
When CD players was reviewed in the old days, HiFi magazines from time to time described the very marked or pronounced starting transients from CDs, that sounded unnatural. This was often described as «digital sound».
Todays players are much better and this problem are nearly gone. But I know some audiophiles always has listened to music with this slight transient edge to it, and when it’s gone, like in pure DSD, they think something is missing.
I love DSD because of very natural sounding transients, and the speed of transients, the latter is very important, Imo.
I looked up the 8004 and based on its audiophile pedigree, it does look like it directly converts DSD. By the way, the redbook layer is going to sound different even if the unit was lower end and only converted DSD to PCM. Redbook 16 - 44 often does not sound as good as higher res; we can argue it it is due to mastering or the conversion technology but in many cases you can and should hear the difference. Of course there are times when you don't which is a cause for consternation. There was a famous case of a Norah Jones SACD where the 2.0 track on the SACD layer was just the redbook source. In that case if I recall the 5.1 track was hi res, just not the 2.0 track.
I’ve tried to convince myself of what you say and to “retune”my ears to get the “truer“ sound of DSD, but I can’t get there. It just sounds dull to me in too many cases. There’s also evidence from the market that most people seem to be unimpressed with SACD. Despite improvements in the mastering DSD over the years, SACD still remains very much a niche product, largely limited to classical music, where I agree it does sound very nice. Also, if SACD really did sound so much like LPs in the old days, I don’t think we would have seen such a huge movement among serious audiophiles back to analog sources, in particular LPs. I think the resurgence of LPs does show that many people agree with the view that redbook cd doesn’t sound “quite right,” But it also shows that they don’t think that SACD is the answer. Somewhere in this thread, our host does say that the improvements of modern DSD mastering equipment is allowing for extremely good results, so I am definitely going to keep my mind open to SACD and DSD. But as Steve said, all of the formats can yield excellent results when used properly, so I’m not so sure how important SACD is going forward, at least for me.
I stuck with vinyl through the dark years. I have a huge vinyl collection that I’ll never get rid of. I do think SACD is the answer. I pretty much only buy SACDs now. That’s in part because I have nice playing copies of most of the LPs I care about and want to own, but in part because to me, SACD does sound right, just like vinyl does it at its best, but SACD sounds right without all of the hassles and problems of vinyl: no warps, no skips, no scratches, no record cleaning, no stylus cleaning, no tracking force gauges, no cartridge alignment protractors, etc.
If Sony had been able to deliver SACD quality at the launch of CD in 1982, I would have adopted that format and never looked back. It’s a real shame that CDs were brought to market with 1970s digital technology that was pretty good, but not as good as it could have been.
What sounds like the master tape all depends on what is meant... If you are talking about the 2 track master used to produce the consumer version, then the most accurate way to present that tape is the best.
However, more than likely, people are interested in the sound of the mixdown produced when the artist and recording engineer get together and produce the final result. In that case, it is all dependent on what happens between that mix and producing that final master for production. More often than not, there is a step that is sometimes called 'mastering'. The mastering step is not usually transparent, and there is often processing done in that step that changes the sound.
So, the answer to the question: What sounds most like the original artist's work product? Answer: Depends on the faithfulness of the mastering and subsequent processing. More often than not, nowadays, esp on the older material that I know about, there is usually a step of processing based on compression in between the original work product and the consumers' copy. MFSL has been known to try to be more faithful, but the regular labels very often have that 'extra touch' placed on the recordings. It IS intentional, and if they really wanted, you COULD have a near perfect copy of the work product, if the distributors want to allow you to have it.
While I follow your point, you are adding a second, and I think different variable -- intentional changes by the engineer. Sure, that has to have some effect, but the question here is, all else being equal, which process sounds the most like the master tape.
The question is not what sounds most like vinyl, but most like the master tape. It may be that truly high quality vinyl is the best -- a previous post by SH seemed to suggest that. Yest, speaking as a non-expert, I find it hard to understand how a medium with 40db channel separation, and considerable out of phase signal/cross-talk even at its best, would sound most like a master tape).
Okay, assuming you mean the two track master (or multi-channel) at the end of the entire process, then it would be essentially anything with properly implemented digital >=16bits and >=48k per channel.
My own desires and interest are closer to artists intent (I don't count mastering as artistic, because of so much historical damage there.) I prefer mastering be mostly 'transparent', mostly ONLY adapting to the resulting medium -- not so much adding dynamic range compression when the consumer medium has 90+dB dynamic range. Too often nowadays, unneeded DRC is added to the material before the final 2 track master (or multi-track for multi-channel.) That extra compression didn't happen when the music artist and their recording engineer produced the original result.
If DRC is needed for the final listening environment, then add it in the car or whatever playback device. Good DRC for dynamic range control isn't difficult to do nowadays -- general purpose damage doesn't need to be done during mastering. Why mess up a recording for nice, quiet home listening so it would also be applicable for listening in the car? Not a good engineering decision...
Of course, you must always listen to what sounds best and most true to you. Whatever floats your boat...
DSD and SACD can be a bit «tricky» in sort of way like there is no free lunch. You have to have things in synergy to really hear the benefits from it. That is my experience.
It’s not my goal that SACDs should sound like Lp’s. I’m happy they don’t. SACD sounds something in between like digital and analog in my ears. If you asked me 10 years ago I would say I preferred the sound of a good analog studio tape to digital. But today, I think the best digital recordings (not that many) surpasses the best analogue recordings in sound quality. Analogue recordings will always be limited by noise. And actually, after listening to high-rez digital over time on SACD, returning to analog recordings, they sound a bit «imprecise and slightly grainy».
So things are not static, they changes over time.
One reason why I have mired some of my answer in a bit of sophistry -- what is meant by master tape? Most master tapes before the middle 1990s aren't pure recordings, but are encoded by a noise reduction system. Apparently, in some cases, decoding the material has been deemed optional. Audio/recording isn't as clear-cut as people might hope that it is, and it has been disappointing for me also.
I get that both LPs and SACDS of recordings from the analog era are almost always mastered from master tapes. In case I didn’t make it clear enough, I think SACD matches, or comes very close to matching, the ability of LPs to reproduce what’s on the master tape, when everything goes right with LP playback. But with SACD, you don’t have to adjust the VTA or tweak the azimuth or own a record cleaning machine to get the best quality the format has to offer.
We can't know this of course unless we have heard both the master table and the SACD. This is why SH is in an unusual position since he has had that opportunity. Other points:
- if SACD or DSD always sounds a certain way it is a poor format, since the idea is to reproduce the original not always sound smooth, for example. It should not have a sonic signature. Most realise this but I don't get it at all when people talk about the sound of a format, it should not have one.
- we know from the often excellent SHM-SACD range that it is the choice of source tapes and care in the mastering process that counts for more than the format.
- people talk about the limitations of CD but while they exist they are tiny compared to the destruction wrought on CD sound by the loudness wars, NR, poor tapes used and so on. The best CDs sound superb.
Separate names with a comma.