Differences between 'Re-Mastering' & 'Re-mixing'

[peace_sells]

A friend of mine (who lives in the UK) got an opportunity to speak with Steven Wilson last week. He's mainly known for being the founding member of the band Porcupine Tree (and a lot of other projects). One of the other things that he's been doing recently is creating 5.1 re-mixes of albums of classic progressive rock bands (Jethro Tull, King Crimson, Camel, etc.). One of the questions my friend asked him was : "So you've been re-mastering a lot of classic albums of late...". Wilson replied : "By the way, it's re-mixing, not re-mastering...".

This is not the first time I've heard people correct others with respect to re-masters/re-mixes. Lot of people seem to have this confusion. Even I din't know about the difference before i did some research on this after hearing my friends' experience with Wilson. So i thought of summarising whatever I learnt from various sources. Maybe it might help some of you guys.

This comparison is for re-master/re-mixes in music only.

Unless clearly specified, re-mastering by default means : digital remaster, i.e. creating digital masters from original analog masters.

Re-mastering involves : altering dynamic range, tonal balance, loudness of tracks (not to be confused with volume of track which can be controlled by the listener during playback), changing EQ (if needed).

Re-mastering doesn't change the content of music.

Re-mastering can also be done to suit the vinyl format.

When the CD/DVD era was in its infancy, the transfer into digital domain was done using inferior processing techniques. Hence, in market today, there are lot of re-masters of CDs that were issued in the 80s. These re-masters are done in order to enhance the quality of the original CD releases (however, lot of re-masters that took place during the 90s were a part of the loudness war. To know more about the loudness war : http://en.wikipedia.org/wiki/Loudness_war#Criticism).

Re-Mixing alters the original music.

This can be done for the following reasons : to make the music more suited for radio play (or DJ play), or to change stereophonic sound to surround sound (for e.g. 5.1).

This can be done by changing the pitch/speed, tempo, EQ, etc.

If I've missed out any other important information, please share.

Cheers.

 

[haisaikat]

when you say this

Re-mastering involves : altering dynamic range, tonal balance, loudness of tracks (not to be confused with volume of track which can be controlled by the listener during playback), changing EQ (if needed).

Re-mastering doesn't change the content of music.

Isnt the originality of the recording lost? Although in my personal experiance Pink Floyd - Dark of the moon I have both the old and remastered version and they sound the same (rips). Not sure my ears or the recording / rip quality

 

[peace_sells]

when you say this



Isnt the originality of the recording lost? Although in my personal experiance Pink Floyd - Dark of the moon I have both the old and remastered version and they sound the same (rips). Not sure my ears or the recording / rip quality

In a way yes. The overall vibe of the sound does change i think. But 'doesn't change the content of music' means there's no change in the song length, no additional music added or deleted. Its quality changes. Re-mixing however could mean otherwise.

 

[Thad E Ginathom]

Imagine the performance going on and the guy with that wonderful array of sliders and knobs, behind the glass panel. This is mixing. I suppose it is not so different to the mixing that is done in a theatre or auditorium. It establishes the basic sound of the music, the relative volumes of the instruments, their placing on the stereo sound stage. Effects can be used, too. Mixing takes the parts and makes them into a whole.

Unless the parts have been recorded individually, there can never be a re-mix.

Mixing is fairly easy for us novices to grasp --- partly because we have all seen that guy-and-mixing desk image.

Mastering is much harder to grasp. Not all mixes are perfect, and the mastering engineer can do some clever stuff to correct problems with individual instruments or frequencies, but, as I [only slightly] understand it, it is much more to do with the whole than the parts. Some of it will be specific to the media, but one of the things that a mastering engineer has to do is to make sure that the music is as listenable-to in your car as it is on the hifi --- or the clock radio.

The process of music production is fascinating. I love to learn about it, but when actually listening to music, it is only the end result that matters.

 

[ranjeetrain]

Isnt the originality of the recording lost? Although in my personal experiance Pink Floyd - Dark of the moon I have both the old and remastered version and they sound the same (rips). Not sure my ears or the recording / rip quality

Originality of a recording is lost every single time it is transferred from the master tapes to any other media (in analogue world), regardless however small the loss. It's inevitable due to the nature of analogue. In Digital world, originality will always remain intact, till the media gets physically damaged or has served beyond it's life span (assuming that resolution was not altered during the transfer process), unless the re-mastering engineer purposefully tweaks it.

Coming to purposeful tweaking (mostly done to enhance the listening pleasure) can actually reduce fidelity (in strictly from the purity/originality of the signal point of view). But even if this is done, and the signal is modified (regardless of the source), this is still referred to as Re-mastering (as the content itself is not modified, it's character is modified for the good (as perceived by the recording engineer)).

Remixing is absolutely different, with a non-formal but granted assumption that the piece will be modified (slightly or heavily) depending on the re-mixing artist. In short, re-mixing is guaranteed to produce a different sound from the original piece.

The easiest way to differentiate between re-mastering and re-mixing is asking the question who did it? If it was done in entirety by a recording engineer, it's remastering. If it was done by a DJ (or similar), it's remixing.

PS: A slightly deep thinking will reveal that the term re-mastering mostly implies re-producing recordings by reading from the original analogue source. If the same is done for digital recordings, it's more of re-sampling rather than re-mastering, as the capability of applying the tweaks a recording engineer would do in his studio is available on every desk (thanks to ever increasing processing capabilities of modern-day computers). An ordinary user with no knowledge of studio equipment can achieve less or more same result as a recording engineer does in his studio with expensive gear sitting right on his desk, with software as simple and free to use as WinAmp (RIP WinAmp, I will always love you). So, there really is no need of tweaking a digital master. At least I won't bother with that if I were the engineer doing a digital remastering. Just re-sample and serve.

By the way, this entire post assumes that we are talking about digital or analogue masters produced after the mandatory post-production work that needs to be done prior to making a production master.

 

[Thad E Ginathom]

A slightly deep thinking will reveal that the term re-mastering mostly implies re-producing recordings by reading from the original analogue source.

I don't think that is right. A modern production that has seen no analogue existence whatsoever is still mastered and could be remastered.

who did it? If it was done in entirety by a recording engineer, it's remastering. If it was done by a DJ (or similar), it's remixing.

Mixing is part of the original recording. It is almost part of the performance. See my notes above. Mixing may also involve the combination of separately recorded tracks into one piece of music. Recording engineers mix; mastering engineers master. DJs are nothing to do with any of this: they are doing something transient as part of the music playback in the chain. Mix may have a meaning in the language of their trade, but it is not the same as its meaning in the recording/production of music.

 

[ranjeetrain]

I don't think that is right. A modern production that has seen no analogue existence whatsoever is still mastered and could be remastered.

Could be? Sure! Even a perfectly mastered recording could be re-mastered (even if that actually worsens it). But the question is - Should that be done?

IMO, no, it shouldn't be done. Why? As I said in my previous post and quote again ---

the capability of applying the tweaks (to a digital master) a recording engineer would do in his studio is (now) available on every desk (thanks to ever increasing processing capabilities of modern-day computers). An ordinary user with no knowledge of studio equipment can achieve less or more same result as a recording engineer does in his studio with expensive gear sitting right on his desk, with software as simple and free to use as WinAmp (RIP WinAmp, I will always love you). So, there really is no need of tweaking a digital master. At least I won't bother with that if I were the engineer doing a digital remastering. Just re-sample and serve.

Mixing is part of the original recording. It is almost part of the performance. See my notes above. Mixing may also involve the combination of separately recorded tracks into one piece of music. Recording engineers mix; mastering engineers master. DJs are nothing to do with any of this: they are doing something transient as part of the music playback in the chain. Mix may have a meaning in the language of their trade, but it is not the same as its meaning in the recording/production of music.

Mixing and remixing is not same. You are mixing mixing with remixing (whoa, sounds like a good tongue-twister).

Also, for the sake of simplicity, it was assumed that (at least I assumed) there is only one (set of) person(s) - the person who is doing the whole thing. If we were to discuss this in actual terms as it happens in a professional studio then we will not only have Mastering engineer and Recording engineer, but also, mixing engineer, assistants, electricians, equipment support engineers and a whole bunch of other people. A professional studio doing remastering for the world doesn't comprise of just a recording engineer and a mastering engineer for sure.

Mixing is part of the original recording. It is almost part of the performance.

No, that's not correct. What you are talking about holds true for live performance. But here in this thread we are not talking about a live performance. The thread is a about studio remastering. In studio remastering - mixing is NOT a part of original recording, it is a part of post-production (which sometimes may happen in parallel with recording, but it mostly doesn't).

Recording engineers mix; mastering engineers master.

Recording engineers don't mix, they just record. Mixing console is controlled by different people, which in case of small studios can be same set of people, but as a task recording and mixing are entirely different (in the context of studio remastering).

DJs are nothing to do with any of this:

Re-mixing will not be possible without a DJ or an equivalent artist who provides the additional/alternative acoustics for the original piece. We can call this person (these people) whatever we like, but the role exists. Using a different term doesn't eliminate the role.

I would love to have examples where a remixed piece of music was produced without assistance an external (set of) person(s).

 

[Thad E Ginathom]

I don't know anything about what DJs do --- and, given my cultural and musical tastes, that is highly unlikely to change . But I thought we were talking about the chain between performer and listener, ie the stuff that happens to music before we buy it, whether that is on vinyl, CD/etc, download or whatever. Re-reading the original post, I still see it that way: this is all about the music production chain, not about DJs or listeners.

Sure! Even a perfectly mastered recording could be re-mastered (even if that actually worsens it). But the question is - Should that be done?

IMO, no, it shouldn't be done. Why? As I said in my previous post and quote again ... ... ... the capability of applying the tweaks (to a digital master) a recording engineer would do in his studio is (now) available on every desk (thanks to ever increasing processing capabilities of modern-day computers).

Nobody can do, at home, what a mastering engineer does, unless they have the engineer's skill, knowledge and training. Sure, the capability is there in the tools. I have a music-biased Linux system, it's all there, but could I tweak music like a professional does? Not even close. I don't even have much knowledge about what they are looking for, what they do about it, and why.

I don't understand why you should think that that the first ever master should always stand, and that there should never be any remaster. What if the guy who did it got stuff wrong? What about improving technologies? What about improving skills?

Sure, it works the other way too; what if the new master is worse, it could be, but that is no reason not to improve things.

 

[ranjeetrain]

I don't know anything about what DJs do --- and, given my cultural and musical tastes, that is highly unlikely to change . But I thought we were talking about the chain between performer and listener, ie the stuff that happens to music before we buy it, whether that is on vinyl, CD/etc, download or whatever. Re-reading the original post, I still see it that way: this is all about the music production chain, not about DJs or listeners.

DJs are very much a part of the production chain. Not every DJ plays in a pub. There are DJs that work closely with production teams. May be in your mind DJ means a hip kid spinning discs in a bar, but the term DJ goes beyond that.

Some pages that you may visit 1, 2, 3, 4, 5. A lot more if you search "DJ production". Even 10 minutes of reading will be enough to ascertain that DJ and music production (or reproduction) goes hand in hand.

I don't know why you should think DJs are not a part of production chain.

Nobody can do, at home, what a mastering engineer does, unless they have the engineer's skill, knowledge and training. Sure, the capability is there in the tools. I have a music-biased Linux system, it's all there, but could I tweak music like a professional does? Not even close. I don't even have much knowledge about what they are looking for, what they do about it, and why.

Nobody? That's a very bold statement you make. May be you mean you can't.

There is hardly anything an experienced audiophile can't do at home with a digital master. Once you have a digital master, it's the same for everyone. A digital master doesn't perform better on studio production machine's HDD than a normal user's HDD.

Any person can have access to software (some free, some need to buy) that are a part of usual re-mastering process in the digital domain. Re-sampling, Normalization, tonal-balance are the things commonly done during digital re-mastering. And all of this, yes all of it, is very much possible to do with say any gaming machine most people own or say any decent multimedia machine people own.

As for the capability of the tool vs the skills, yes, it is person to person. But any music lover who is not absolutely new, begins to understand everything fairly easily. And the software is there to assist, most of which are a lot more easier to work with than considered. It might just be harder for people who are not tech savvy compared to those who are, but everyone who wants, he can.

For those non-tech savvy ones, modern software are already doing it. Many software players do actually analyze every single digital track that is added to their collection, without the requirement of any knowledge or skill from the user. Those are the kind of things 10-15 years ago only big names sound engineers could do with mega buck equipment in studios. And here, a software does most of it, if not all, without needing any knowledge or skills from users (beauty of digital).

In the past years (bridge years) a lot of the studio stuff could be done on the desktop with tools that were once a forte of only highly skilled sound engineers. You may want to look here to see what all an ordinary user could do with a digital piece on a normal computer.

I don't understand why you should think that that the first ever master should always stand, and that there should never be any remaster. What if the guy who did it got stuff wrong? What about improving technologies? What about improving skills?

Sure, it works the other way too; what if the new master is worse, it could be, but that is no reason not to improve things.

Yes, if it does improve things. But like you agree, it need not necessarily improve things. It can go either way.

Like I said already (and at the cost of repeating myself) digital is easy to process. Every modern day software analyzes full tracks it plays and processes it before playing it back. I am sticking my neck out and claim that, gradually, the need of processing in the studio will come to an end as processing at every step will only harm the fidelity of digital masters. I would much prefer to cook my digital dish just once vs cook is many times (something like why people prefer NOS DAC).

I personally, will always take the original digital master. Once I have raw rice I can make whatever meal I want. I don't want cooked or preprocessed rice. But I agree with you, some people might not have the equipment or the skill or the will to let a software tailor their music at their end.

It's clear that there are always two camps of people. One who like to eat out, and the other who like to cook. Those who prefer eating out will always give a number of reasons why they should eat out. At the same time, those who like to cook will always have reasons to cook at home. You are in the camp of people eating out. I am in the camp who would like to cook their meal themselves.

 

[peace_sells]

Some very interesting points made by you 2 there! However, one doubt is creeping into my mind..You say :

PS: A slightly deep thinking will reveal that the term re-mastering mostly implies re-producing recordings by reading from the original analogue source. If the same is done for digital recordings, it's more of re-sampling rather than re-mastering,

This might too trivial but I'm still curious..

Now if the original digital master is re-sampled (assuming there was no analogue master for it), what is the point in re-sampling? Even if you increase the number of samples while re-mastering, won't it be kind of useless, since no data exists between 2 consecutive samples in the original digital master (I hope I've understood this correctly). I mean..there needs to be an underlying continuous time/analogue signal for re-sampling right? If the analogue master is re-sampled with a greater sampling frequency I understand, but re-sampling a digital master

 

[Thad E Ginathom]

I think resampling is the wrong word and misleading.

DJ = Disk Jockey. The "disk" (even if it is some other format these days) is what the music industry produces, and the DJ plays it --- even if DJing has become so much of a performance art that the music he plays doesn't sound anything like the media.

 

[ranjeetrain]

Now if the original digital master is re-sampled (assuming there was no analogue master for it), what is the point in re-sampling? Even if you increase the number of samples while re-mastering, won't it be kind of useless, since no data exists between 2 consecutive samples in the original digital master (I hope I've understood this correctly).

Valid point, with your assumption. But you misunderstood slightly.

What you are interpreting the term re-sampling as, is not re-sampling, but up-sampling. That is, the process of creating non-existent samples by extrapolation. Many companies (Wadia comes to mind instantly) have developed proprietary algorithms that have been very successful in what they were attempting to achieve.

HOWEVER, in my post you referred to I wasn't talking about up-sampling at all. I was talking about re-sampling. The two terms are very different. Allow me to explain.

When an engineer re-samples from a digital master, he usually down-samples (because original studio masters are higher resolution to compensate for rounding errors during post-production). To be more specific an engineer may re-sample 32/192 studio master to 44/16 (while producing a redbook CD). In this example, bit-depth is being halved and less than 1 samples is being taken out of each 4 sample present in the original digital master. This process is known as down-sampling, one type of re-sampling.

However, the reverse is true too (as you understood above), as in up-sampling is also a type of re-sampling. But up-sampling usually occurs in a play back system (at the end-user's system).

Here, let's delve a little deep into this. Here is the biggest reason why I would prefer to have unaltered studio digital master than re-mastered: When I get a re-mastered version, it is processed (which usually involves down-sampling). On my playback system, I again up-sample it (my system does, my CDP/DAC whatever my source). So, here is my reasoning. Why would I choose to lose the fidelity of digital recordings by down-sampling + up-sampling + other processing multiple times. I would much rather choose a smart player (JRiver/JPlay/Amarra) that analyzes a track prior to playing it back and feed it with the original studio masters.

End result? Pure-pristine-untouched digital samples of music recreating a holographic image of the original recording venue.

 

[Thad E Ginathom]

I can see your point on that now. i suppose that, int his day and age, remastering is likely to include a change in the sample rate etc, largely because "high-res" is what "they" are setting out to sell us.

I have a doubt, though. That is: exactly what are we buying as a "studio master?"

I don't know the answer to that question.

 

[ranjeetrain]

I am glad it helped.

The studio guys will always stay a step ahead of the consumer, regardless what becomes the consumer market standard. Even if they start selling (current) high-res to consumers as a standard product. This is a capitalist world; marketing machinery will keep provoking studio guys to get the latest and greatest products. They will keep coming up with even highly capable machines specifically meant for studio guys.

Currently high-res means (by and large) 24/96. Large percentage of audiophile community doesn't own equipment to do justice with higher resolution media. Due to that reason, 32/192 is serving as a reference for studio master as of now. I am not sure if more than a select few studios might be producing higher res masters.

Regardless, due to the limitations of recording equipment and other factors even 32/192 is an overkill right now. Studio guys still need to have the need to exceed that resolution. They may at best want to have the capability to process at 64/192 or 64/384 resolution. But even that might only be a requirement only for the processing session.

What is even more important is that DAC designers still need to learn to process higher amounts of data without making the final sound too-detailed (for which usually negative adjectives such as clinical/fatiguing are used). Till the time they learn to process higher amounts of data, 24/96 is the best bet.

Personally, I want 32/192 to become universal standard for digital music distribution. I am putting my money on this resolution due to it's suitability for consumer application.

 

[Thad E Ginathom]

Well, 32 bits could be useful if you want rocket launches on your home theatre to be realistic volume!

Hyperbole: I have no clue what the actual dynamic range would be. At least the disk makers will love you!

 

[Santy]

One hour of music at 32/192 will take up 5273 MB of disc space.
The Audio Archive, Inc. - Audio Bit Rate and File Size Calculator

 

[ranjeetrain]

5273 MB of storage today is cheaper and less obtrusive than 650 MB in 1980. Even at 32/192 we can store 1000 albums in one physical HDD. In analogue world it will take one wall full of physical media.


About the Dynamic range: Increased resolution shouldn't necessarily be linked with increased dynamic range. At least that shouldn't be the only objective to increase resolution. The biggest reason for increasing resolution is to have the headroom to be able to process samples without losing anything due to rounding/truncation.

If I take an example: 99.99 OFC is a sufficient start point for making great speaker cables and interconnects. But people buy 6N/7N/8N material. The objective is to reduce the loss. With increasing purity the loss will decline. Similarly in digital, with increased resolution chances of losses would decrease.

 

[Thad E Ginathom]

The biggest reason for increased bit rate, in the studio, is to avoid clipping when summing different sources. 32-bit floating point is to do with more accuracy in applying effects/edits. (better, more fully, and more accurately explained here)

This information was spelt out in my Cool Edit Pro manual a decade or so ago, and explanation can be found on just about every site that talks about music recording.

 

[Santy]

5273 MB of storage today is cheaper and less obtrusive than 650 MB in 1980. Even at 32/192 we can store 1000 albums in one physical HDD. In analogue world it will take one wall full of physical media.

I do not disagree. It it is not very expensive these days, considering that I have unmuxed Bluray rips @35 GB per movie so going by that, it is totally justified to have 5 GB for one album of high quality music for those who believe.

But a 5TB HDD is not out yet, I guess!

 

[jls001]

If a single file is of the order of 5.x GB, does it means the RAM will have to increase substantially? A hi-res (Thad: please excuse that usage as "high bit rate high sampling" is too cumbersome and not exactly catchy) at 24/96 is typically 100 to 150 MB. 24/192 is bigger. But even 2 GB RAM on a regular PC or laptop, theoretically, is more than enough to buffer one hi-res audio file. So will we need larger memory? In any case, this is only a "nice-to-know" question, and hardly a deal breaker, as memory will become ever cheaper.

32 bit masters at the highest sampling rate is a good thing. For most modern recordings, there is no more analog master tapes which can be stored away in climate controlled vault under much security. Future generations will have only these digital masters with which to issue remasters by future mastering engineers.