When Does it Become About Samples?

[Turtle]

As in, as opposed to good production ability? Is there ever a point where you've learned what you can with sub-par instruments and hit a ceiling with how your work sounds? I ask, of course, because it's recently become relevant to my experience. I will be the first to admit that I am possibly the worst digital musician I know at the technical aspects of audio production, but now I'm also wondering if it's reached the time where I should shell (heh) out for Actual Instruments, the kind for which you have to pay Actual Money Dollars.

More importantly, how do I tell? Accurate self-evaluation is even harder than the work I'm trying to evaluate. I could just say "I'm in a rut and don't know how to EQ," which is true, but I'd be joining in the collectively whiny "how do i shot web" without being conducive to improvement (and by "web," I mean "good production values for my music"). In the interest of trying to make this a legitimate thread for advice on the issue - instead of simply voicing my frustration at not seeming to improve in the last year - is this a familiar phenomenon? Does anyone else find that it was better in the end to stop working out with crunches and pushups and just get the darn Bowflex?

And most importantly, have I just posted a sentiment that's come up half a dozen times already? Probably. Stay tuned to hear the answer.

[Argle]

Yes. Better virtual instruments can definitely help. Quality samples are easier to mix. Whether or not you indulge in them depends how much money you want to throw into the bottomless pit of music production. If you have the money and a sweet tooth, go for it.

[AngelCityOutlaw]

I'm not a sound engineer, but I've offered this opinion many times and shall do so again:

A Line 6 Spider III does not sound as good as a Peavey 5150 to my ears. If Eddie Van Halen played through a Spider, he would still be Eddie Van F*ckin' Halen, but the simple truth is that his signature 5150 that he used for years is still a higher quality amplifier and is capable of producing a much better timbre. Equally great music can still be composed and performed with the spider. An equally great "production" however, is a different story.

The same can be said of virtual instruments vs other virtual instruments, real instruments vs. real and virtual vs. real.

[SnappleMan]

Most OCR users don't realize that sample "quality" is relative. On here what most people consider to be a "good" sample is a "real sounding" sample. As your ear becomes more tempered and experienced you'll find yourself less and less convinced by the realism in all samples, and when that happens all you have left are the samples that truly sound good for whatever reason.

Personally I prefer to use older samples or samples from hardware synths because the developers really do a great job in mixing and programming those patches to be playable and feel good under your fingers. Today's newest samples are returning to that mentality, since for the past 10 or so years we've all been installing 100gb libraries that are terribly programmed and have no musical nature to them whatsoever, so the developers now are focusing on making those massive sample banks usable in a musical setting by creating very smart articulation engines and whatnot.

In the end all that matters is what you do with something, and a new sample will not fix your problem since you'll have to take time to learn how to properly use that too. If your goal is realism then you'll fail 100% of the time until you get a real performance recorded, but if your goal is style and character then you can use all the samples and effects you have to create something unique and cool without it trying to trick anyone.

[AngelCityOutlaw]

Snaps makes good points.

[timaeus222]

Most OCR users don't realize that sample "quality" is relative. On here what most people consider to be a "good" sample is a "real sounding" sample. As your ear becomes more tempered and experienced you'll find yourself less and less convinced by the realism in all samples, and when that happens all you have left are the samples that truly sound good for whatever reason.

...

If your goal is realism then you'll fail 100% of the time until you get a real performance recorded, but if your goal is style and character then you can use all the samples and effects you have to create something unique and cool without it trying to trick anyone.

I'd like to build on what SnappleMan here has said. I consider sample "quality" to be not only how realistic it is, but also the competency of the timbre. A low quality piano can still fit into a song if it's processed well, like, say, through a little chorus and some hall reverb, then integrated into an 80s retro remix (where's WillRock? ). A synthesized electric sitar (like mine, for example) can be considered good quality if it still evokes the middle-eastern atmosphere every time it's played, whether or not it compares well next to a real electric sitar.

The only time when sample competency really matters is when you've established a sound palette for yourself that you like, and your ears find your samples not quite satisfying enough yet. Right then, go find what you believe you need. Once you're satisfied, you're good to go. Just work with what you've found until you know how to use it well, and you've got your sample quality set in soft stone---still time to change your mind. Even if you're satisfied with your sounds, you can always add more to your palette; just depends on whether or not you get good enough to keep wanting more.

I wouldn't say the goal for realism is to trick people though. I'd actually call something real if it sounds convincing enough, but whether I'm convinced or not, I haven't been tricked (unless it's something I shouldn't know squat about).

Edited June 8, 2013 by timaeus222

[Vig]

It's been said, but a samples that sound good and samples that sound realistic are overlapping but not identical sets. I think people get into trouble when they try for the sample that sounds the most realistic but falls short in that regard, rather than the sample that sounds good, and perhaps not realistic. No samples ever sound quite the same as a musician miked up. Listen to commercially successful music. It's all full of samples, but they are rarely meant to sound "realistic." I love the sound of a saxophone, but I would never try to sequence a saxophone in place of the real deal. I might use a saxophone sample, but I'd probably affect it to the point where it clearly isn't supposed to be a saxophone.

[Moseph]

I'm right now seriously considering buying LA Scoring Strings and/or the Sample Modeling brass set to improve my orchestra template, so my thought process may be helpful to you.

IMO, the time to upgrade your samples is when you can discuss the specific ways in which your current samples fall short and the ways in which the new samples will address those failings, and you are at the point where spending money to fix the situation actually seems like a good idea to you. For example, some of my motivating issues are that my current VSL brass section has a lot of awkward velocity layer breaks that make it a pain to program, and I'm not comfortable with how the horns and trumpets sound when used for solo lines. And my VSL strings have neither divisi nor the lush Hollywood-style timbre that I want.

For discerning sample problems vs. production problems, it may help to step back temporarily from whatever samples you're currently using and try to write a track using, say, just several instances of one synth. Don't try to be realistic in any way; just try to make something that sounds good. If you find that you still run into the same sorts of problems that you do when using samples, then your sticking point is probably production technique and/or composing ability.

Edited June 8, 2013 by Moseph

[Turtle]

I'm right now seriously considering buying LA Scoring Strings and/or the Sample Modeling brass set to improve my orchestra template, so my thought process may be helpful to you.

IMO, the time to upgrade your samples is when you can discuss the specific ways in which your current samples fall short and the ways in which the new samples will address those failings, and you are at the point where spending money to fix the situation actually seems like a good idea to you. For example, some of my motivating issues are that my current VSL brass section has a lot of awkward velocity layer breaks that make it a pain to program, and I'm not comfortable with how the horns and trumpets sound when used for solo lines. And my VSL strings have neither divisi nor the lush Hollywood-style timbre that I want.

For discerning sample problems vs. production problems, it may help to step back temporarily from whatever samples you're currently using and try to write a track using, say, just several instances of one synth. Don't try to be realistic in any way; just try to make something that sounds good. If you find that you still run into the same sorts of problems that you do when using samples, then your sticking point is probably production technique and/or composing ability.

This is good insight. Especially the last paragraph. I should probably give it a try and simplify, simplify. Then if it's the production that's the problem, I'll have to find some excuse for not getting any better at that for a year or so, despite having taken at least two college classes in that time about digital audio. >_>

The reason I'm even bringing up that it might be my samples is that the only element of my package that I've paid for since getting FL Studio is FL Studio itself. I'd like to think I've done pretty well for myself on what I can find for free but the cheap stuff is cheap for a reason...

[dannthr]

As in, as opposed to good production ability? Is there ever a point where you've learned what you can with sub-par instruments and hit a ceiling with how your work sounds? I ask, of course, because it's recently become relevant to my experience. I will be the first to admit that I am possibly the worst digital musician I know at the technical aspects of audio production, but now I'm also wondering if it's reached the time where I should shell (heh) out for Actual Instruments, the kind for which you have to pay Actual Money Dollars.

More importantly, how do I tell? Accurate self-evaluation is even harder than the work I'm trying to evaluate. I could just say "I'm in a rut and don't know how to EQ," which is true, but I'd be joining in the collectively whiny "how do i shot web" without being conducive to improvement (and by "web," I mean "good production values for my music"). In the interest of trying to make this a legitimate thread for advice on the issue - instead of simply voicing my frustration at not seeming to improve in the last year - is this a familiar phenomenon? Does anyone else find that it was better in the end to stop working out with crunches and pushups and just get the darn Bowflex?

And most importantly, have I just posted a sentiment that's come up half a dozen times already? Probably. Stay tuned to hear the answer.

When you want to make money. That's when you start using quality samples. When you want to make a job of it.

Composition skill is a never-ending staircase--there is always up. In the beginning, each step to a new level is small and easy, you can make breakthroughs with relatively minimal energy and effort. But for me, it takes months or years to work to the next level. There is always up and always someone to be inspired by, always. This aspiration really has nothing to do with making money.

Higher quality virtual instruments are about saving time and reaching higher production value with less effort OR designing production sounds with more flexibility.

You spend money on virtual instruments when you want to make money. If you're not interested in making money, then the only reason to do it is for your own ear pleasure--and that's not a decision other people can make for you.

And let me tell you what: It doesn't end. I am constantly sacrificing entertainment opportunities and travel opportunities to acquire tools.

Even now I'm contemplating buying a bicycle or buying 2CAudio's B2 reverb plug-in. I live in LA and I don't have a car, so a bike would be nice--but B2 will let me do this specific thing I want to do with my existing tools that I haven't been able to do with any of my current reverbs (at least not to satisfaction).

Do you want to live like this? Like me? Constantly under pressure to sound your best as quickly as possible? Because this shit stresses me out.

But also, it SOUNDS AWESOME!

[dannthr]

I'm right now seriously considering buying LA Scoring Strings and/or the Sample Modeling brass set to improve my orchestra template, so my thought process may be helpful to you.

Both excellent libraries.

Though, the time you'll save with programming the Brass will be diverted into mixing it.

LASS is GREAT, but requires a lot of EQ massaging to fix some harsh phase issues.

[AngelCityOutlaw]

I agree with Dan.

I think that the other posts are coming mostly from a musician's point of view rather than considering the listener's point of view. I think both must be considered. How does it sound to your ears, but....

If you're making music for someone else or lots of people are going to be listening to it, you need something that sounds very good. It's like, if some indie game developer wants as close to Hollywood film soundtrack as possible, but you can't afford to book the Hollywood Studio Symphony, you should be able to sound as close to the real deal as possible. If you can't, you will probably lose out to the person who can.

So yeah, in that kind of a situation, you can screw around trying your best to make some soundfonts sound as real and "high quality" as possible, or you can just buy a sample library that makes it easier to sound like how the client wants the final product to sound.

The only magic "sounds super awesome AND real" button is a killer live performance and that shit don't come cheap n' easy.

[EC2151]

Well, you don't want to use SNES spc soundfonts if you have something "better-sounding" lying around.

I love the sound of synthesizers, warm analogue strings, and stuff like that, so I often eschew realism because I know I don't have the budget or samples to pull it off all that convincingly. You always have to play to your strengths.

My FL "mentor", whom I learned a lot of my starting-out stuff from, told me that 99% of your sound from the DAW will come not exactly from your samples or instruments, but from filter effects and how you use them. I have found that this is the case, and I tend to just, mix, match, and tweak the preset filter options to get things that sound interesting or good to my ear.

[Turtle]

Well, you don't want to use SNES spc soundfonts if you have something "better-sounding" lying around.

I love the sound of synthesizers, warm analogue strings, and stuff like that, so I often eschew realism because I know I don't have the budget or samples to pull it off all that convincingly. You always have to play to your strengths.

My FL "mentor", whom I learned a lot of my starting-out stuff from, told me that 99% of your sound from the DAW will come not exactly from your samples or instruments, but from filter effects and how you use them. I have found that this is the case, and I tend to just, mix, match, and tweak the preset filter options to get things that sound interesting or good to my ear.

Heh, as far as my strengths go, I often use chiptunes because it allows me to arrange to my heart's content with comparatively little focus on making the mastering sound "realistic." Synth is deeeefinitely my friend. That said, I do want to be a bit more versatile with "real" instruments and need to know what I'm missing in terms of those filters (even after a few years of this, I'm afraid it probably just comes down to inexperience) that make the base sounds perform better. This is because, down the line...

When you want to make money. That's when you start using quality samples. When you want to make a job of it.

It would be fantastic to make a job of writing music for games or film or TV; heck, even commercial jingles. Long shot though it may be.

Composition skill is a never-ending staircase--there is always up. In the beginning, each step to a new level is small and easy, you can make breakthroughs with relatively minimal energy and effort. But for me, it takes months or years to work to the next level. There is always up and always someone to be inspired by, always. This aspiration really has nothing to do with making money.

Higher quality virtual instruments are about saving time and reaching higher production value with less effort OR designing production sounds with more flexibility.

You spend money on virtual instruments when you want to make money. If you're not interested in making money, then the only reason to do it is for your own ear pleasure--and that's not a decision other people can make for you.

And let me tell you what: It doesn't end. I am constantly sacrificing entertainment opportunities and travel opportunities to acquire tools.

Even now I'm contemplating buying a bicycle or buying 2CAudio's B2 reverb plug-in. I live in LA and I don't have a car, so a bike would be nice--but B2 will let me do this specific thing I want to do with my existing tools that I haven't been able to do with any of my current reverbs (at least not to satisfaction).

Do you want to live like this? Like me? Constantly under pressure to sound your best as quickly as possible? Because this shit stresses me out.

But also, it SOUNDS AWESOME!

Are you saying there's no real balance to be had? Like, you can't get moderately good-sounding samples and then learn a moderate amount of production (honest question, coming from someone who has tried for as long as possible to go as cheap as possible)?

I do want to learn more of how to make something sound good given the tools already available. But I'm skeptical that, for example, some guy working with the 128 general MIDI instruments would never need to just plain "get better software" to improve past a certain point.

I agree with Dan.

I think that the other posts are coming mostly from a musician's point of view rather than considering the listener's point of view. I think both must be considered. How does it sound to your ears, but....

If you're making music for someone else or lots of people are going to be listening to it, you need something that sounds very good. It's like, if some indie game developer wants as close to Hollywood film soundtrack as possible, but you can't afford to book the Hollywood Studio Symphony, you should be able to sound as close to the real deal as possible. If you can't, you will probably lose out to the person who can.

So yeah, in that kind of a situation, you can screw around trying your best to make some soundfonts sound as real and "high quality" as possible, or you can just buy a sample library that makes it easier to sound like how the client wants the final product to sound.

The only magic "sounds super awesome AND real" button is a killer live performance and that shit don't come cheap n' easy.

Hits spot-on, because in the future I would probably like to compose for some smaller game projects to really learn what it's like to build a soundtrack.

It seems like it'd be a better, if more grueling, exercise in the long run to just push harder at making a better pot with lower-quality clay. Which in turn makes it a question of "how does one learn this?" Not really a helpful question, and one that gets asked here countless times a day, but nonetheless I find that even being no longer as self-taught as I was a year ago, something's not clicking... I may not want to hear it but maybe I'm just not applying myself enough?

[Moseph]

Are you saying there's no real balance to be had? Like, you can't get moderately good-sounding samples and then learn a moderate amount of production (honest question, coming from someone who has tried for as long as possible to go as cheap as possible)?

I do want to learn more of how to make something sound good given the tools already available. But I'm skeptical that, for example, some guy working with the 128 general MIDI instruments would never need to just plain "get better software" to improve past a certain point.

(TL;DR -- It's better to know what specific benefit a software purchase will have for you than to buy something with the abstract hope that it will make you sound "better." But if you don't know how to evaluate software in relation to your abilities, one possible solution is just to throw a bit of money around and see what happens.)

It really depends on what you mean by "improve." You can learn all of the important stuff about music production with free software -- by which I mean you can understand how all of the generally-used audio production tools function and can make music that sounds good in the sense that the composition is excellent and the production techniques (EQ, compression, balance, etc.) are spot-on. But you spend money on software when you find software that gives you a particular sound (or ease of use, or whatever) that you can't get otherwise and when having that sound (or whatever) is worth the price to you. The trick is to know enough about your abilities to be able to evaluate software and understand what its benefit will be to you if you purchase it. And so I think the time to purchase a piece of software is when you can both justify the expense to yourself and can articulate why you want to have it in a more detailed way than "I hope it will make me sound better."

As an example, I spent ~$700 maybe six years ago on an orchestral library -- my first major sample library purchase -- and have never regretted the expenditure. My justification for the purchase at the time was that I was not satisfied with the orchestra sound that free software could give me. My level of satisfaction in this case was based on my judgement that I could make the the expensive library sound more like a real orchestra, which I could tell because I had the listening/production experience to evaluate the software I had vs. the software I wanted to buy vs. a recording of a real orchestra. I didn't buy based on some nebulous notion that I needed to spend money to sound "better;" I bought based on the understanding that the specific purchase I was making would change my sound in a specific way.

I think maybe the problem you're running into is that you don't yet know how to evaluate your own abilities well enough to know how new software will affect those abilities. And if you've hit a wall with this and think new software may help, the best thing to do might just be to buy some software and see what happens (this assumes you're willing to risk the money on experimentation). And the first step in that case is to ask yourself, "What software specifically do I want, and why do I want it instead of all the other available free or non-free software?" And in answering this question, you may find that you've articulated something about your abilities that you only knew instinctively before -- or at least have learned something about your musical interests.

EDIT: And this whole discussion leaves aside the fact that each piece of software has its own learning curve, which in some cases can be very steep (I'm looking at you, EWQL Choirs). It's possible to be good at music production generally but still to find yourself at a loss for what to do when dealing with a particular sample library. Even professional composers create forum threads asking how to get X sound from Y library.

Edited June 9, 2013 by Moseph

[Rozovian]

Prob gonna read this thread properly, just not today. Quick thoughts:

Everything depends on your definition of good. Snap's well-programmed old samples example is one definition of good, realism is another measure of good, versatility is another, some other aesthetic qualities are yet another.

But it's not just the samples, it's also your idea of what makes good music. if you're the type who'll say that old video game music sounds good, you're gonna have to qualify that with how they're good for the time, the technical limitations, or just good compositions rather than wholly good. Or the stuff that's playing on the radio, which you'd also have to qualify, albeit differently. If the music you make sounds like it came out of a crappy speaker playing retro sounds, and this actually suits the project you're making it for (game, video, whatever), then that's good because it's a good fit.

So to actually get to the question - if you're making something that'll sound all right next to other rock songs, you'll need the things that make rock sound good. Whether this is fake or real doesn't matter as long as the result is good. Whether this is top-of-the-line equipment or a shoddy practice amp or anything in between doesn't matter as long as the end result is, in context, good. Likewise if you're going for a Hollywood orchestra sound, or need to display some clear Daft Punk influences or whatever.

In practice, short of doing AAA games or scoring any serious live action stuff (regardless of scope), you should be all right grabbing older and lower-tier libraries. A decent set of drum samples can cover for a more expensive, realistic library until you're making something where you actually have to fool people that know this stuff. Likewise the orchestra. Likewise the electric guitar. Likewise pretty much anything. Cartoons, most games except the most realistic ones, less than entirely serious videos... for those you can make do with the cheaper stuff without really causing a disparity between the expected sound and what you're actually delivering.

If you're taking on projects where you don't feel that what you've got works, you probably need better samples. Getting to that point with your clientele requires a fair amount of making do with what you've got and building skills first, so I think this is a decent measure of when you need better samples. If you're just doing it for yourself, not really taking orders and delivering tracks for anything, it's just a matter of how heavy your wallet feels.

tl;dr: Definition of good, and context/use.

[Moseph]

Also (and it sound a little humorous to say this), buying sample libraries is a skill that you develop by buying sample libraries. Over time, you get better at evaluating demos, reviews, and companies' descriptions to figure out just what exactly you'll be getting for your money. And there are nuances to samples (e.g. wet samples vs. dry samples, 5 GB libraries vs. 85 GB libraries) that you may never develop opinions on if you stick to free stuff. I hesitate to list "it will make you better at buying samples" as a reason to start buying samples, but if you anticipate building a collection of purchased sample libraries in the future and are just trying to decide whether to buy your first one now or wait, it might be a consideration.

Edited June 9, 2013 by Moseph

[V___]

It also depends what you want from your productions (note, I play synthetic stuff mostly so if you apply my comments to piano or orchestral sounds you're retarded.. plus I write for myself, I dont rely on it for an income)

Theres a song - 'Babara Streisand' by Duck Sauce - that I hold up in these cases. Its an entertaining song, its popular on MP3 players, in clubs, on radios... I'll dance to it when smashed off my nipples. But fuck me, as a musician, I hate it. Its made of four layers at best, and three of them are samples of other peoples work (Vengeance samples, Boney M, etc). How can you release that and say its YOUR track?!

For me, I make sounds because I love CREATING the sound - the detail in building the right sound from scratch, balancing 2-3 OSC's with the right envelopes and effects... when you get the sound you were working towards, after hours of work... Well, being able to take something completely intangible from your head and make it physical its... orgasmic.... and thats an understatement

But sometimes I just want to make a track to show a friend or something, and samples give you that extra freedom and speed. Its not as satisfying, emotionally, but its a Babara Streisand - its fun easy and accessible.

You can make a free violin sound like $1m if you want to spend the time processing each sample. Its a matter of whether the time and enjoyment you get from that process is worth the cost of a better library.

[dannthr]

It also depends what you want from your productions (note, I play synthetic stuff mostly so if you apply my comments to piano or orchestral sounds you're retarded.. plus I write for myself, I dont rely on it for an income)

Theres a song - 'Babara Streisand' by Duck Sauce - that I hold up in these cases. Its an entertaining song, its popular on MP3 players, in clubs, on radios... I'll dance to it when smashed off my nipples. But fuck me, as a musician, I hate it. Its made of four layers at best, and three of them are samples of other peoples work (Vengeance samples, Boney M, etc). How can you release that and say its YOUR track?!

For me, I make sounds because I love CREATING the sound - the detail in building the right sound from scratch, balancing 2-3 OSC's with the right envelopes and effects... when you get the sound you were working towards, after hours of work... Well, being able to take something completely intangible from your head and make it physical its... orgasmic.... and thats an understatement

But sometimes I just want to make a track to show a friend or something, and samples give you that extra freedom and speed. Its not as satisfying, emotionally, but its a Babara Streisand - its fun easy and accessible.

You can make a free violin sound like $1m if you want to spend the time processing each sample. Its a matter of whether the time and enjoyment you get from that process is worth the cost of a better library.

Yeah, but even for synth stuff, I would argue that you still get a lot more for your money when you spend it: Zebra 2 and Diva by U-he, and to some extent Omnisphere by Spectrasonics and Tremor from FXPansion. I think they're worth it and I would prefer to use them over the free stuff that comes with the various DAWs.

Also, Kontakt and Reaktor from Komplete are ridiculously important for creating and designing your own custom sounds.

If I were recommending a synth and effects bundle for creating awesome stuff:

Instruments:

U-he Diva (beautiful analog synth emulation)

*U-he Dark Zebra (flexible, do-everything modular synth)

*FXPansion Tremor (drum synthesizer for creating custom drum sounds)

*New Sonic Arts Granite (granular resynthesizer)

*Native-Instruments Komplete

FXPansion Geist (amazing drum sampler/beat slicer/drum machine/pattern editor)

#Spectrasonics Omnisphere

#iZotope Iris

*Essential

#Neat but not essential

Effects worth considering:

FXPansion Etch

iZotope Trash 2

FXPansion Maul

2CAudio B2

FXPansion Bloom

Fabfilter Saturn

PSP Audio N20

Virsyn KLON

Virsyn MATRIX

And more! Woooo.

That's a pretty penny there.

[Argle]

You people are rationalizing too much. It's fun to buy new gear, and it's healthier than a crack addition. So go support the economy.

[dannthr]

That may be, but it's not healthy FOR your crack addiction. That rock isn't going to buy itself.