Original Master Recordings. Here I always thought every CD I owned was an "Original Master Recording"! Guess not?

I just found out about these MFSL edition CD's a few days ago, and grabbed a few of them, then compared them with my originals.

I am asking if and why these MFSL CD's are better because I don't currently have the HiFi equipment to see any fine details. But I will at some point in the near future. However, it is now that I have a chance to buy a bunch of these MFSL CD's.

The only thing I have noticed is that these MFSL CD's are all even more quiet than my originals (I have only tested 3, but all 3 have the same result). And these 3 CD's are late 80's early 90's originally, thus they are already quiet compared to more modern CD's.

I have definitely owned and tested out "remastered" CD's, and I've found with those that they are generally louder.
And so I thought these MFSL CD's would be similar, but they are the opposite.

Are the MFSL CD's somehow gaining quality by being more quiet? Because in my experience, a more quiet CD usually is associated with more hiss noise from needing to turn the volume up even further than I normally would. That is barely a noticeable issue, but notable none the less.

However, I'm obviously not going to throw away MFSL CD's because of that. I don't doubt that there is still a better quality to these, I just want to hear what you guys think to find out what makes them better. I tried doing some searching the internet about these CD's and I've come up with surprisingly little info about why these are supposedly better.

If these MSFL CD's do have a better quality, then I will stick with them. But like I said, I need your comments as I don't have much more than headphones right now to work with, which explains why the only real detail I can pick out is loudness. I'm also verifying that by viewing waveforms.

If MFSL indeed have better quality, it is easy for me to apply some replay gain on the albums, so loudness really isn't too much of an issue anyhow. But if the only difference between the two types of CD's is loudness, then I see no reason to go with the MFSL CD's.

Btw, the 3 MFSL CD's I've compared with originals are Nirvana Nevermind and Guns n' Roses Use Your Illusion II and Appetite For Destruction.

CDs that are perceived as "quieter" for the same subject material usually indicates reduced dynamic range compression. For most rock albums, the master tapes have more dynamic range than the final released record. This is due to a number of reasons:

• The final record itself has physically less dynamic range (vinyl) - compression is necessary for all the detail of the recording to make it onto the record.
• Because customers expect something that is listenable at low volumes in high-noise environments (ie in a car).
• Recording labels generally wish to ensure a consistent "sound" to all of their releases. ie, the relative loudness of all of a label's rock releases in a given era is going to be around the same.
• Mastering engineers sometimes wish to impart their own unique signature to the sound.
• Noise reduction algorithms for digital remasters. These are rather notorious for introducing distortions in quiet sections of music, although modern algorithms may not suffer from this issue.

And on and on. One of MFSL's biggest selling points is that they advertise trying to avoid all of that, and present on CD exactly what was written to the master tapes. This is generally easy to show objectively, by comparing original CD releases to the MFSL releases. As a result, they're usually quieter than the original CD release. In contrast, CDs that are remastered by the original label are often simply adjusted for today's perceived tastes. As producers expect listeners to play their CDs in the car and in non-hi-fi environments, rather than the dedicated home theater rooms of yesteryear, they are generally mastered to be louder than before.

http://en.wikipedia.org/wiki/MFSL

Beyond that there's a lot of audiophile crap associated with MFSL that I don't buy into - most especially gold CDs. Moreover, that a release is from MFSL is no absolute criterion of superiority over the original CD. Their release of Frank Zappa and the Mothers of Invention's "We're Only In It For The Money" is provably from the same digital recording as the original Rykodisc CD release (which was extremely well mastered to begin with), and very little was changed in the remaster besides an absolute polarity reversal, IIRC.

Definitely not. There's no guarantee the original, zero-eth generation two track master was digitized to make a CD. (It's not a guarantee for Mofi's either, despite the advertising, but they did often use them)




Yes, because MFSL didn't tend to apply additional compression to the recording, which has been the fashion since around then to make CDs sound louder.



See above. MFSL took a more 'purist' approach than is common nowadays.



THey are generally gaining more dynamic range -- or rather, presenting the original dynamic range , rather than compressing it.



The Wikipedia entry is a good place to start, and reiterates much of what I've written here.

http://en.wikipedia.org/wiki/Mobile_Fidelity_Sound_Lab



The definition of 'better quality' is subjective. So no one can TELL you which of two CD versions is objectively better. They can only describe how the CDs differ\, and why they prefer one over another.

Replaygain will match the 'subjective' overall evels of two versions, making any comparison more fair.

care for a quick, easy answer?

they were mastered correctly - that's it. EVERY cd COULD sound like that, if only producers would "get" it.

Wouldn't that be nice?



as to your last comment:

that is partially correct - replaygain will indeed match the levels of two cd's, however, that does not take into account the fact that one disc may be maximized and compressed for loudness, while the other one isn't - better quality is not necessarily subjective in this regard. Numerous tests have been done on to prove that.

I've heard a lot of MFSL CDs. For the most part, the changes are really minor. Some aren't the best versions floating around (IMO, the 20th anniversary DSOTM is better than MFSL, for example). However, I find some really obvious. The MFSL version of Pink Floyd - The Wall is much, much better sounding than the Capitol remaster.. Also, there isn't a "blip" in the beginning of Comfortably Numb, which is nice. The Nirvana ones are nice as well.

For the most part, IMO, MFSL CD's aren't worth the extremely high prices that they are going for on eBay. But brand new, they're a good buy. I just wish that every CD coming out would be mastered in the way of MFSL...

Quick and dirty answer: superior mixing results in CD which sound great. I do not care if they get them by applying tar to the tapes. I have never had a MFSL CD which was not far better than the original.

For understanding why louder is not better:

http://www.youtube.com/watch?v=3Gmex_4hreQ

Thanks guys. All that is good info. And that youtube video is nice! However, since the CD's that I have MSFL versions of are old and not compressed to begin with (loud), I'm looking to see what else MFSL does that makes their versions superior. It actually seems that it would make more sense that MFSL put out their versions of more modern (loud/compressed) CD's rather than the old ones they are doing! So why are they not doing that? Seems MFSL's work is all from the pre-loudness era.

Simply, I will just assume MFSL releases will never be worse than the original and hopefully will always be better. That is good enough for me!

However, I have a couple circumstances here that I want to make sure will be OK.

I will be ripping them to mp3, although I may listen to the actual CD from time to time. However, will my mp3's hold the same quality that the CD does?
I ask this because of the "gold" on the disc that supposedly makes the CD sound better. Obviously, my mp3 won't have any gold on it!
So I just want to make sure that the MFSL will also create higher quality mp3's than the originals will.

Another thing is that I do travel a lot and listen to most of my music (in mp3 form) while driving. So one thing I am considering is adding some replay gain to the mp3's ripped from the MFSL CD.
Let's say the MFSL CD has an overall volume of 90 and the original CD is 93, using mp3gain.

If I decide to make the mp3's from the MFSL CD the same album volume that the original has or higher as long as I avoid clipping, then will I still maintain the same quality, just louder?
I believe the dynamic range would be the same if I avoid clipping.

So basically, if you had the choice of an MFSL CD and the original CD and they both were the same price, yet you were somebody like myself who was going to make mp3's out of the CD and listen to them while driving, then is MFSL still the better buy? And remember both CD's are the same price, and I may use replay gain on the mp3's (without clipping).

In short, are there any circumstances where the MFSL CD would NOT be the better choice for somebody (disregarding price). For example, somebody who knows they will only listen to the CD while driving?
As for myself, there will be times I will be in quiet listening to the album on a nice setup. As often as I'm able to!

I did simplify in my comment about Replaygain, it's true.

As for as 'numerous tests' that show 'better quality' is somehow objectively determined by compression level -- I'd like to see references to some. I am aware of something of the reverse -- that psychoacoustically speaking, the somewhat louder of two presentations (whether achieved though simple level increase, or via compression) tends to be reported as 'better sounding', rather than 'louder'.

My point being, evaluations of 'better' must always be subjective...there's probably someone out there who prefers the sound of music over a old Victrola, versus the latest/greatest technology. And there are people to whom a modern, clipressed remaster will sound better than a Mofi. Preference is inherently subjective.





MFSL CDs aren't remixed.




Compression has been used for decades in pop/rock music broadcasting for very much this reason: the make music sound 'better' in the noisy environment of an automobile. When the dynamic range is compressed, there is less difference between 'quiet' parts and 'loud' parts of the music, so the quiet parts don't get lost in the car noise.

Then producers/engineers began printing compression right to CDs -- even though radio broadcasts were still using compression. It's arguably more than a bit nutty, but it is standard now.

Replaygain doesn't involve compression, so in your car it will never achieve that sort of 'levelling' effect where you can hear acoustic guitar parts as easily as screaming electric guitars in the same track. So even with replaygain applied, MFSLs may not be the best remasters for car listening....unless you have a very quiet car.

That's not nutty at all - it's logical. With the advent of CD-playing head units (and boomboxes/discmen), CD's were being played in more noisy environments. Hence, a need for more dynamic range compression.

Since you're new to well-mastered CDs, you might check out a DCC CD or two if you can find them once you get your stereo equipment. IMHO, the quality of DCC CDs is on average higher than MFSL CDs.
Robert

What;s nuttty is to have all this compression hard-coded into the software, so that EVERYONE has to hear it, rather than having selectable compression as an option on the playback gear, e.g the 'night listening' modes on some AVRs.

I think the nutty thing they meant was to compress it heavily on a CD, and then have the radio do the same again to it.

And in my opinion, all the compression is very bad. In my opinion it only exists for the record companies to try and sell more records with a cheap gimmick of their records sounding a tad louder.

I havnt done any listening tests on MSFL or anything, but I have the piece of mind that it was mastered with quality in mind, not sounding a tad louder then the newest muchmusic pop music stuff.



In my opinion, compressed albums sound considerably worse then uncompressed ones (most notabley compare Pearl Jam's Ten from 1991 to their greatest hits album Rearviewmirror in 1991). I can hear that on my 100$ computer speakers. I can even hear that on my 1$ earbuds! (buying 20$ ones got to expensive when they broke every few weeks). (All in my opinion, don't hit me with whatever TOS some mods here are crazy about where you need some super professional test for every opinion on quality).

As for listening in a car, I know replay gain won't compress, but it should help a little in that I can knock the stereo volume back a few notches.
I've noticed with these MFSL CD's that there is headroom for replay gain without clipping.

Since the original versions of the CD's I've mentioned are old enough to not have compression (at least not the notable compression that is oh so obvious on more modern CD's, if these old original CD's do have compression, it is very light compression according to the waveforms), I don't think I'm losing out on much as far as how the music will sound while driving.

Both CD versions (MFSL and non) have dynamic range and little to no compression (have very similar waveforms, with the original version just being a bit taller/louder, but still similar), which is why I'm assuming I'll have successful results by using replay gain to match the volume level of the original CD.

In short, what I'm saying is that it "seems" like if I just added some RG to the MSFL CD's, they would be just like the originals as far as loudness and dynamic range goes (not talking about quality). Maybe not exactly the same, but roughly.

Now if MFSL did their version of a CD released in 2002, then it is another story entirely as far as how it would sound while driving, of course.
Before going any further with this, I really want to do some simple sound tests in my car and pay more attention to this "loudness war".

After listening to you guys, and checking out that youtube video, I think I find myself on your side of the loudness war. Bring on the dynamic range! I think that tracks without this compression honestly sound more like you are in the recording studio or at concert. What is supposed to be loud is loud, what is supposed to be quiet is quiet.

I think I would like to listen to music this way unless I'm in a really noisy environment and I couldn't hear the lows at all. Even in my car it is not that noisy to where it would block out quiet parts of music to the point where it would annoy me.

It would be nice to have like a two sided CD (compressed tracks on one side, non compressed on the other) to get the best of both worlds!

It's not bad per se, it's only bad when it's inappropriate (that constitutes the majority of cases, though). I don't want to strip you of your opinion, but it seems to me that you're missing the original point of compression.

To be fair, there's very little pop music, even from decades ago, that has much "quiet" about it.

There has been DRC for as long as there has been recorded music. Even in the acoustic recording days, the engineers would sometimes move the singer away from the recording horn when they sang a loud note! Mechanical DRC, if you like!

The kind of compression that helps with in-car listening and creating a punchy full sound has been in use for decades. It does make the recording slightly less "hi-fi" but it doesn't make it lifeless, harsh, and totally objectionable. It's squashing/clipping the peaks and pushing as close to 0dB FS that gives us today's sound, and I might be getting old, but I don't think anyone actually likes that sound.

The problem today is that very brief peaks in signal energy (typically from drum hits, the attack of note from certain instruments etc) are generally squashed across the entire mixdown, and the overall level raised. This doesn't help with car listening at all. It rarely helps with listening on a cheap stereo. All it does is make the track louder (or as loud) as the competition. The cost is that the music becomes lifeless, and sometimes doesn't sound much like music at all.

The real problem is that you don't have to be some kind of hi-fi nut to hear the problem - comparing overcompressed and OK samples back to back, most people won't comment on the sound quality - they'll just enjoy the less compressed version more, on a conscious and subconscious emotional level.

Cheers,
David.

If you make a bold statement like that I'm convinced you've never heard uncompressed music.

Compression is a vital part of recording and all music sold in shops to end-listeners will have been compressed to a certain extent. At least since probably somewhere in the late 1960's. Before that recording quality would usually be low enough not to need compression because dynamics would have been poor anyway.

If I listen to the raw 'tapes' my brother (he's a jazz musician) gets from of the studio I have a general feeling what it will sound like but it won't sound enjoyable until at least some basic mixing and compressing has been done. It is not something you'd sit back with and enjoy, it generally gets on your nerves within ten minutes and has nothing to do with musical pleasure.

So, compression is always necessary. However, the amount of compression is what it's all about. Not enough compression makes it sound crap, as will too much compression, the trick is to find the sweet spot. Where this sweet spot is depends on who you ask and generally tends to have shifted in the last decade.

To put the other side of the argument...

It's possible to make binaural or purist stereo recordings of most live events, and the "need" or otherwise for compression will be determined by the ability of the musicians and the placement of the microphones.

If you get both of these right, compression can only degrade the recording.

Multi-miked pop music, and/or the desire for a "created" rather than real sound, is different.

btw, "late 1960s" is not right - there's audible DRC on early Beatles recordings - listen to "Any time at all" on A Hard Day's Night which includes DRC pumping that stands out a mile.

Cheers,
David.

Compression has existed since classical recordings used gainriding in the 1930s.

I'd like to believe that's true (as it gives hope for a reversal of the mastering trend) but I know of know formal studies that show that. And I know plenty of people who've bought new remasters, that are far more compressed then the original CD version, who prefer the new one.

Well, that ought to be easy enough to test, right? Take a set of uncompressed samples, and compress them by an equal amount with a compressor that is representative of modern mastering practices. Loudness-match them and ask people to compare them on a subjective basis. Also record where people are listening to the tracks (car, ipod, headphones, speakers ets) and try to correlate between the two.

Alas, it wouldn't be necessarily objective, but it would answer a lot of questions.

I think what I meant was compression, as used by the record companies, not just normal compression.



Like I said, I wasn't refering to compression as a technical thing, I was refering to compression as used by the record companies in the loudness race.

I don't think ive heard totally uncompressed just recorded music, all I know is that albums that are much much louder that I listen to seem to sound bad. Even before I knew what all this stuff (loudness race and all) was, I listened to many albums including Rearviewmirror and thought "Why does this sound so bad? And why is it much louder then all my other music?".

But as Axon said in the following post, the fair test is loudness matching.

I'm as big a sucker for preferring the "louder" sound as anyone else on the planet. Take one track, make it 0.5dB louder, and very few people will say it's louder - but all kinds of serious listeners will declare it's better in any number of ways (pick any positive adjective from the audiophile dictionary!).

The only remasters which sound better when loudness matched with the originals are ones where there was some other genuine improvement (e.g. going back to the original master tapes, whereas second or third generation dubs were used originally) which brings greater sonic benefit than any extra compression brings detriment.

Of course some "remasters" are actually remixed, and that's a whole different ball game.


Can anyone suggest a good source for this test? Something that's been remastered with more compression, from the same master tapes? I think some of the stuff on Beatles One counts, but the EQ and NR is also different from previous releases, making it sound more gentle on some tracks, despite more compression!

Cheers,
David.

I've enjoyed reading this thread and can't really add anything to the ideas already expressed. Let me say, however, how refreshing it is to read a thread on here with opinions expressed without someone freaking out and screaming blue murder about TOS #8 violations. It's getting to the point where every time a good, useful thread gets going on here, someone decides to ruin it by twisting the intent of the TOS #8.

As for the topic of the thread, I remember one of the biggest arguments in favor of the CD back in the 80's was that huge 96dB dynamic range. Yet here we are with many modern recordings that are restricted to the top 5-10dB and some with only 4-5dB! Bob Dylan expressed it well in an interview this year. There's no space left in the music, it's all like a TV test pattern, just a constant square wave. MFSL and DCC discs tend to use much more of that potential 96dB dynamic range whereas some "digitally remastered" CDs from the major labels try to cram everything into the top 5-10dB. Sad.

Thanks guys for all the great comments. I'm convinced that these original master recordings (when available) are the way to go, even if I want to use replay gain on them.

I always have noticed the quieter older CD's and the louder newer CD's, but I never realized all the reasons and details behind it, like the loudness war. So thanks again for all the great info. Everything makes more sense now.

I don't like the heavy compression used in today's CD's, but loudness is a perk that people like these days as more people are listening to portable music in loud environments than ever before.
However, I think at best we should hope the music industry finds a happy medium.

Cut the compression back a little, add a little more dynamic sound range, then it is more fair to everybody's tastes.
And some of the newer CD's are so compressed that I don't think cutting back on that and adding a bit more dynamic range is honestly going to hurt the loudness all that much to where most people would even notice, except audiophiles!
So more dynamic sound range, but still try to make the music as loud as it can possibly go without clipping.

That would be ideal, for me. I hope the music industry does catch on and can hold back at least a little with the compression. Also it would of course be nice if this whole MFSL and DCC movement would continue to grow and be more popular.
If the music industry sees that these are starting to sell, then I think that is our best bet for potential change in how CD's are mastered.

Why not make your own remaster like Axon suggested? Just take a raw studio mix that hasn't been mastered, and then apply compression to one, none to the other, then wavegain them and make a comparison. I think it would be incredibly hard to compare studio masters together...well I can think of one example, the SACD Dark Side of the Moon. They were mastered off the same tapes, right? And we can all agree that the CD version is substantially more compressed.

I've done that, but the argument could always be made that my attempts at compression are bound to sound horrible, while the recording industry might like people to believe that they can do a better job.

(Which, TBH, I'm sure they can!!!)

Cheers,
David.

If you mount a pendulum on a structure, then it will look as if it stands perfectly still - but it doesn't, because of earth, moon, mars and even the milky way having a gravitational influence on it. But who cares? Good enough is good enough - making something better than necessary is inefficiency. And if all we want to show someone is a concept/principle, then it doesnt take much to be "good enough".

So far, i havent seen anything perfect in this world. And i dont need to.

- Lyx

I wasn't trying to avoid the job - I've already done it a few times (I'm looking for the tracks right now, but I think they're at home).

What I'm saying isn't "it won't be perfect" - it's that "it won't be good enough for a valid comparison".

These guys who are smashing tracks against 0dB FS are using expensive equipment. In many cases, they avoid introducing actual distortion - the sound is just loud (and flat). IME with toys like Cool Edit Pro, you can't get within 6dB of their result without introducing audible distortion.

Mind you, a bad workman always blames his tools - I make no claims to know how to do this properly - it's not my field.

I'll post what I can when I can, and encourage others to do the same. Still, I think anyone with suitable software/equipment can do better!

Cheers,
David.

Actually, many modern compressed recordings have very audible compression distortions, just not so much as a clipping or THD distortion, but audible volume 'pumping' as the compressor attacks/releases.

So far as 'expensive hardware' used for compression, I would be interested in knowing what such a piece can do, that a well written DSP algorithm (such as used in late versions of Cool Edit Pro/Adobe Audition) can not.

I have spent a few hours in Adobe Audition(the re-badged Cool Edit Pro), and initially I had problems getting results similar to the big labels. But after a few hours of tinkering, I have been able to get pretty good comparable compression(so far as lack of distortions using aggressive compression and limiting) to the 'big' labels. I do this for fun, and to help me understand exactly what is modified and what the results are, when the music is compressed to a certain degree. Also, this helps me identify problems in big label recordings when I hear them.

As far as the poster whom claims that some degree is compression is always good -- I tend to agree to a point. I do not want any compression of classical music or other such instrumental presentations that are supposed to emulate a live unamplified auditorium performance. However, with most other music, I think he is correct. I made some recordings of a friend playing guitar, and then singing later, then post-processed and mixed these. Some level of compression made the sound quality higher, in my subjective opinion. Of course, I used compression in a sparing manner, leaving on average, about 20-22 dB effective dynamic range to 0dBFs from the average RMS power level -- nothing like would be found in a major label release. My compression was more like a slight 'shift' of dynamics on a particular track of the mix, not a global compression. For example, I would take the vocals and shift some of the lower mid level volumes gently, in a non linear curve, closer to the higher amplitude, to get a more full sounding vocal, to make the vocalist sound more controlled and increase clarity. I also limited peaks of narrow time duration from the guitar track that were basically irrelevant to being very audible, as these seemed rather pointless, and did nothing but require the recording to be quieter without any substantial benefit, so far as I could tell.

-Chris

This would be interesting. But I tend to doubt you would get the quality as achieved from re-mastering early era master tapes. From what I have read on some pro sound websites, it is becoming more common for the final mixes to contain a rather substantial level of compression that was applied in the mixing stage(s). Some mastering engineers have complained that this makes their job harder, and that they want as little compression as possible on the final mixes that they have to master. Remember, the mastering engineer gets a final mix to work on, not the individual recorded tracks. Based on this tid bit of data I have read, assuming it is true, one would need to go back to the original pre-mix recorded tracks, and re-mix the music and make new master tapes.

-Chris

"...mastering systems should be neutral and transparent. The essential idea is to unveil all the detailed musical information on the original master recording without adding deterioration, coloration or other sonic artifacts." "...faithfully reproduced high fidelity recordings that would compliment the numerous advances occurring in audio delivery systems."

Links of note: http://www.vxm.com/21R.46.html and http://members.aol.com/boardwalk7/mofi/mofi.html

MSFL twisted opinion...
Quality and dynamics are the obvious MSFL differences. I cannot say about later releases as I own few. My MFSL purchases are very selective. One I enjoy immensely over the standard release is Pink Floyd's The Wall. I truly love the extended dynamic range [and yes, the quieter aspects!] of this work, especially using headphones. I use MFSL discs only in-home. I have used them with various systems from Class-A to very inexpensive and found hardware plays an important role, but environment a more curcial one. I believe the appreciation for MFSL discs, or lack thereof, is simply subjective to the listener. It’s normal to listen to an album and become accustomed to its nuances so when a different version is heard, the listener feels a loss, a lack of familiarity. I can attest to this as I only own "The Wall” in MFSL and although it contains the same music and vocals, it definitely has a different “presence” for me over other versions.

I equate MSFL discs to Linus’ blanket. Nothing is perceptibly the same to Linus without his blanket. And is there really anything wrong with owning and loving a blanket or two?

Yes, and as he also said, the test hasn't been done. I'm sure *individuals* have compared tracks this way (I have) but in no way is there data out there to make a general statement that 'most people prefer to uncompressed version'.




Well, again, 'better' here is subjective, and AFAIK there are no controlled studies to prove this to be a general trend of opinion among listeners -- though we'd all like it to be true!




I presume all released of the Yes album 'Fragile' have been remastered from original master tapes since at least the 'Gold' version mastered by Joe Gastwirt that came out in the early 90's. It was remastered again after that with MUCH more compression, for the recent Rhino version. (And re-remastered, apparently, for the DVD-A version, which is also more compressed than the Gastwirt, and lately remastered AGAIN, by Mobile Fidelity). Howveer, even in cases like this there's absolutely no guarantee that different EQ/NR moves weren't also applied.

It would be better as an experimental sample for one of us to generate a more-compressed version of an already-existing 'uncompressed' CD, mimicing modern mastering compression.

i have also heard of people who prefered bad remasters not because it sounded better/worse, but instead to AVOID level-matching - replaygain isnt mainstream yet, sadly.

The gold coating on those CDs isn't supposed to make them sound better, but to make the medium last longer.

Of cause MFSL records are mixed different. Myself has this "Sting - Nothing like the Sun" record from them.
They once claimed to offer the same quality digital recording but with better high fq control.
To be honest i always preferred my standard recording.
Thinking further reveals the original recording was already done in the digital domain and MFSL just bent the audible range a bit....