how much exp to the next lvl?

[mickomoo]

Sorry for the lame RPG joke, but it's early and I'm too tired to think of another title. So anyways, I've been composing and playing around for about a year, maybe a year and a half and I have about 16-20 "compositions." On paper, it seems like these would be solid pieces, but for some reason my execution seems off or something. I'm working on remixes of video game music (wips are in the remix section) but they don't really sound good at all and I can't completely tell what's wrong. I know that I have trouble with timing, I'm working on that, but it seems like I might have other problems as well, I'm just having difficulty assessing where to work on.

Below are my best recordings/compositions, could you give me tips or any common problems these might have:

I'm know my other recordings have issues too, but if I look at problems that might have occurred in my "better" songs, I'm assuming it'd at least give me a starting point to assess areas of improvement for most of my songs.

[Tensei]

Lack of rhythmic/melodic structure is the main problem here. Try writing something simple and melodic in 4/4 before you move on to crazy dissonant stuff.

[Rozovian]

Timing. Some samples have a long attack. Either cut the attack (sample start parameter, or cut it manually in the sample itself (on a copy of it)) or move the notes back so they play earlier and their attacks line up better with other instruments. (also what Tensei said)

Dynamics. Orchestral musc is all about dynamics, and most music benefits a lot from having softer and louder passages. Even within notes themselves you can build or break down a passage just by how loud a note is. When you're faking a performance, a human player, don't forget that human performers tend to change the pressure of sustained notes. Whether we're talking brass or woodwinds (or organ).

Mixing. The more the tracks bleed into each others' space in the frequency range, the more muddy it sounds. Decide which instruments you need up front, which ones can be further back, and use reverb and EQ accordingly. Write and eq the instruments you have so they all have their own place in the frequency range. Don't indiscriminately boost the bass' lows or the hihat's highs.

And most, importantly - LISTEN. Learn to listen. Listen to stuff on the wip board, listen to random crap on youtube, listen to professional bands and artists, listen to old remixes here on ocr, listen to the new stuff... listen and try to find stuff that could be better.

And read stuff. There's thousands of threads and articles online about this stuff, and should you not find the right guides, you can learn the right principles from the "wrong" guide too.

[cfx]

Composition and arrangement. The other posters are dead-on, so take heed. I do want to mention that your problem is not your instruments. It's how you're using them.

When you have corrected the things already mentioned (structure, timing, dynamics, and ear-training/listening) THEN worry about upgrading your gear. I made that mistake when I was starting out, and it set me back because I wasn't paying attention to the fundamental skills I needed to build.

[Moseph]

Lack of rhythmic/melodic structure is the main problem here. Try writing something simple and melodic in 4/4 before you move on to crazy dissonant stuff.

Elaborating on this, I think the biggest problem you're running into, mickomoo, is that you have some fairly complicated things going on in the music from a harmonic/melodic perspective, but the performance of the music (the programming of the instruments, the timing problems that others have mentioned, the dynamics, etc.) isn't up to the level that it needs to be for music this complicated to sound really good. For the listener to parse music like this, there needs to be a very clear sense of phrasing and overall direction. When there isn't, the music tends just to sound noisy rather than musical.

To work on this, I think it might help to step back a little from what you've been writing and write some really simple stuff and work primarily on constructing a good performance. Once you've got some more experience in the performance aspects of the music, you should be able to come back to the tracks in the original post and make significant improvements.

Here are some things to google or ask people about that you'll probably have to learn at some point: expression control, quantization (I know from other threads that you're already aware of this one), tempo maps, and possibly filter automation and velocity layer crossfading if your sampler supports them.

[mickomoo]

Hey, thank you guys for the quick feedback, I sincerely appreciate that!

Timing. Some samples have a long attack. Either cut the attack (sample start parameter, or cut it manually in the sample itself (on a copy of it)) or move the notes back so they play earlier and their attacks line up better with other instruments. (also what Tensei said)

Dynamics. Orchestral musc is all about dynamics, and most music benefits a lot from having softer and louder passages. Even within notes themselves you can build or break down a passage just by how loud a note is. When you're faking a performance, a human player, don't forget that human performers tend to change the pressure of sustained notes. Whether we're talking brass or woodwinds (or organ).

Some of my instruments have key switches, would that be sufficient enough to humanize the timing and dynamics?

Mixing. The more the tracks bleed into each others' space in the frequency range, the more muddy it sounds. Decide which instruments you need up front, which ones can be further back, and use reverb and EQ accordingly. Write and eq the instruments you have so they all have their own place in the frequency range. Don't indiscriminately boost the bass' lows or the hihat's highs.

I'm reading on mixing but I'm still having trouble assessing how much EQ lead instruments need, and what low end instruments need. Frequency spectrum analysis programs would help? I'm thinking at the very least I can prevent instruments from sharing the same ranges.

When you have corrected the things already mentioned (structure, timing, dynamics, and ear-training/listening) THEN worry about upgrading your gear. I made that mistake when I was starting out, and it set me back because I wasn't paying attention to the fundamental skills I needed to build.

I was thinking newer instruments would help me out with either my timing or mixing, lol. I at the very least wanted to get a trumpet and some other brass.

[eilios]

I'm reading on mixing but I'm still having trouble assessing how much EQ lead instruments need, and what low end instruments need. Frequency spectrum analysis programs would help?

1. There's no definitive amount. It's different for every mix, and every instrument.

2. It does.

EDIT: They are right about learning to make use of what you have being better than trying to upgrade. At first, I thought that if I had killer synths like Massive, Zebra 2 and FM8 it would instantly improve my sound quality(as opposed to Synth1 and 3xOsc). While they are quality synths, it's still superior to learn how to use what you have than to constantly try to "upgrade".

[Rozovian]

Some of my instruments have key switches, would that be sufficient enough to humanize the timing and dynamics?

Nope. Some articulations available with keyswitches might line up better, but then it's a different sound. If you want long string notes, you can't solve the timing problem by switching to pizzicato.

With humanizing, it might help, depending on what the keyswitches actually do, but it's still just another tool in the toolbox. You still gotta learn to edit the notes so you get a more human sound (velocity, expression), and you gotta learn to tell whether the problem can be solved with technique or if you need different sounds (no amount of sequencing technique in the world is gonna make a single-sample guitar sound like a real guitarist). Listen to stuff.

I'm reading on mixing but I'm still having trouble assessing how much EQ lead instruments need, and what low end instruments need. Frequency spectrum analysis programs would help? I'm thinking at the very least I can prevent instruments from sharing the same ranges.

This comes from listening. Listen to stuff, compare it, open it up in your DAW and see how it looks on the spectrum analyzers and whatever tools you've got. At this point, you only need to worry about two things, frequency and amplitude, and neither requires any special software to work with.

I was thinking newer instruments would help me out with either my timing or mixing, lol. I at the very least wanted to get a trumpet and some other brass.

Brass is difficult to sample because the instrument behaves differently depending on how it's played. If you use it a lot, see if you can find a pro package/instrument for brass. A good one, I mean. It might be better to focus on styles and sounds that you have the resources for. Then again, you might be stuck only doing electronica that way, as I am.

But generally, better instruments only gets you so far, and then you might run into hardware limitations instead. While better tools means a better sound, learning to use the tools you have is way more useful than getting new tools all the time.

--

edit: go look up screenshots of the synths eilios mentioned (as well as FreeAlpha, TAL-Elek7ro and some other simpler synths), look at the many knobs and windows they have. If you don't understand what they do in basic synths and can't adjust them to make the sound more suitable for the track you're working on, you're certainly not gonna be able to do that in a more advanced synth. That said, synth1 has a terrible interface.

edit2: I agree with Moseph, except in the case of cheap samples that weren't recorded together and therefor sound different together. Sometimes, you need a bass boost and other stuff to make a more hall-y sample fit in with more dry samples, or un-dry them with more than just a reverb. But more often than not, an orchestral mix has problems because of how it's written, any technical problems are secondary.

[Moseph]

I'm reading on mixing but I'm still having trouble assessing how much EQ lead instruments need, and what low end instruments need. Frequency spectrum analysis programs would help? I'm thinking at the very least I can prevent instruments from sharing the same ranges.

I may be in the minority in thinking this, but IMO EQ'ing sampled orchestral instruments is not something you should consider as a first line of defense in getting a good mix. (Note that I'm talking specifically about sampled orchestral instruments and not about synths or things you've recorded yourself.) In my general experience, orchestral libraries don't sound bad if you don't EQ them. EQ can be used to great effect in polishing up the sound -- if you know what sound you're going for -- but level balance and orchestration tend to be bigger issues for orchestra music than EQ.

Part of the reason that I take this view is that traditionally orchestral composition has been for live performance on acoustic instruments and EQ as such hasn't existed as a way to modify the orchestra's sound. In these cases, it's orchestration techniques that determine whether the frequency balance works. If you focus on using digital tools such as EQ to change how the orchestra fits together without any clear idea of the effect you're trying to achieve, you can end up with something that doesn't sound realistic and might have been more properly addressed through orchestration.

As an extreme example, if the problem is that you can't hear the flute solo over the loud brass, the solution is not to kill ~1.2 kHz in the brass and boost ~1.2 kHz in the flute -- the solution is to recognize that from a performance perspective what you've orchestrated doesn't work, and trying to make it work without reorchestrating it will only make it sound unrealistic.

[SonicThHedgog]

what are you using for your trumpet?

[eilios]

edit: go look up screenshots of the synths eilios mentioned (as well as FreeAlpha, TAL-Elek7ro and some other simpler synths)

FreeAlpha is really great, but why TAL-Elek7ro? There's a newer version out.

[Rozovian]

FreeAlpha is really great, but why TAL-Elek7ro? There's a newer version out.

They look pretty much the same. " II" was just, apparently, too much for me to type.

[mickomoo]

what are you using for your trumpet?

The default trumpet that comes with Kontakt 4 in Komplete 6. Either that or DSK brass, or some soundfonts

[Rozovian]

The default trumpet that comes with Kontakt 4 in Komplete 6. Either that or DSK brass, or some soundfonts

If you have Komplete, you have what you need for now, and then some. Learn to use it better.

Also, don't try to learn the synths in Komplete before learning some simpler ones first.

[mickomoo]

If you have Komplete, you have what you need for now, and then some. Learn to use it better.

Also, don't try to learn the synths in Komplete before learning some simpler ones first.

I was really afraid to even touch the synths, esp absynth with the patch mutations and whatnot. I've not really branched out composition styles so it will be awhile till I use FM or absynth. I just wish I could EQ these vsts to make them sound better

[Rozovian]

Wrote way too much.

Best way to learn a synth is to start from an initialized patch, everything turned off or set to its default or minimum value. This is a lot easier with simpler synths, like the aforementioned one.

It helps to understand each kind of synthesis before using it. Additive and subtractive synthesis are fairly straightforward, one is a combination of waveforms, the other filters out frequencies (like an eq) from waveforms.

Massive has a lot of waveforms to work with, but is mostly subtractive. The interface is pretty simple, 3 oscillators, two filters, some effects, and a box with the envelopes and lfos and some other features. If you know your way around a more basic synth, it shouldn't take long to understand Massive.

FM8 deals with frequency modulation, which is when you use an oscillator to control the pitch of another. FM8's default patch, the one that's there when you open it up, is pretty straightforward - just a single oscillator. You can turn on and route more oscillators into it to see how it changes the sound. Furthermore, change the pitch and waveform of them, see how that changes the sound. It also has an easy editing window where you can control everything from just a few knobs. Useful for when you sort'a have the right sound, but it's not quite there yet.

Absynth uses a wide range of synthesis and processing techniques (additive, subtractive, fm, granular, ring modulation, comb filter etc), so it's not a good place to start. On the other hand, if you find a patch you sort'a like, you can mutate it into something more fitting.

Reaktor might seem really intimidating, but it's probably the best way to understand this stuff.

So if you wanna learn synths, start with something simpler, and move up. Jumping headfirst into Absynth isn't a good idea

Here's the "short" version:

Get FreeAlpha. Turn everything off in it. Learn what everything does. Use presets, learn to tweak them. Make your own patches with it. Move on to TAL-Eke7ro II, repeat process. Move on to Massive, repeat process. Move on to FM8. Move on to Absynth.

Going into Reaktor and spending a few hours on making a simple synth thing there (one oscillator, one envelope, one filter) can teach you a lot about how synths create sound. It's not necessary to make music or to use synths, but it's interesting, fun, and useful. In that order.

As for everything else, just listen and compare, imitate or deviate. Listen to newb stuff to learn to tell what's wrong with their tracks and avoid that in your own works, and to tell what they're doing right then try to do it yourself; listen to pro stuff... for the same reasons. Take a mix of yours, throw out all the effects, and mix it again. What will you do different? How will you EQ it this time?

There's no magic "improve" knob in an EQ or any tool. Everything requires some knowledge of what it does and the ears to tell if you're making it better.

Also, read my guide. Apply everything you learn.

I suck at writing short.

[Anorax]

[Off Topic]

I suck at writing short.

You did good compared to your other posts

[End Off Topic]

[spacecowgoesmoo]

Also, don't stress too much about memorizing every single piece of (very useful!) advice posted. Try to use them to improve, of course, but as long as you keep on writing you'll get loads better on experience alone.