timaeus222

A bit conservative in the beginning, but the solo and the bass licks help a lot. Yeah, I'd say submit it! Not the most interpretive in the compositional sense when it comes to how the source itself is interpreted, but I think it's above the bar due to the textural dynamic variation and the added solos and bass phrases.

Definitely agreed with the crits & comments; the drums were Groove Bias though - pretty old, so probably not much else I could do there on the snare. Would definitely do it though if I could. By the way, something cool I remembered is that I spent like 2 hours rendering individual stems, so if you want, you could go here to find this ReMix, except you can isolate every single instrument. Could be good for folks on ear training! (Unfortunately it takes so long that it's the only ReMix that I have on there!)

@Furorezu Huh, sorry, I didn't notice this when you posted it. Whatever you did, it worked; I think you didn't add your post-mix reverb, because it doesn't sound nearly as distant, and it has a decent bass presence. I'm not on my good headphones so I can't say much more than general advice for comparison to what you're doing, but it's definitely an improvement in the perceived audio fidelity. ----- General advice (not necessarily directed toward your mix, but to you in general, so you know what to look for): One thing to watch out for is adding what you call your "post-mix reverb". If you lower your dry mix (the input signal), then you get a more distant sound since you retain more of your reverberated output (the wet signal) than your dry signal. It's like taking the instrument out of the room, but keeping its "reverberated essence", which isn't exactly realistic. Also, it sounds to me like you're adding it on the Master track, which is not necessarily how you should do it. In a digital environment, you're free to add reverb onto each instrument separately. Technically, even a bass guitar would have natural reverb in a room, but in a mix it adds muddiness (frequency overlap in the 100~300 Hz range, more or less) since the low frequencies are being reverberated somewhat as well. So for example, if you begin with identical reverb settings for a bass guitar and rhythm guitar, you should raise the low cut within the reverb plugin that's assigned to the bass, so that the wet signal is high-passed (has its low frequencies filtered out). The wet signal is just the processed (reverberated, in this case) output due to the given input. For a bass, generally a low cut frequency of ~200 Hz is about right, since it cuts out the low end reverberations, but not the midrange + treble, and your bass still mostly has its natural reverberation behavior. For a rhythm guitar, a low cut frequency of 300~400 Hz, more or less, would accomplish the similar goal. This way, the kick and snare's low frequencies, which are collectively around 40~250 Hz overall, should be clearer.

huh oh well Super creative. Few remixers would think of doing it in this style for a battle theme, and incorporating the source into the bass and e. piano. Great work!

@Nase has a good point; For example, depending on the sample, I actually sometimes like stiffly-sequenced chords when I write electric piano chords that are processed through a wah pedal, since it creates a spikier transient that has more presence, whereas having slightly offset notes softens the transients a little. Or, sometimes I like to gate big chords for effect, which is a way to "sequence" stiffly.

Ah, okay. That was AngelCityOutlaw's atonal drone.

Good, I was going for realistic crackling.

Looks to me like this says it: You're given a selection of sources, and you work it out with your teammates which one goes to who. And here, sounds like in the boss-battle rounds, you can pick amongst the three sources your three-person squad has picked.

No problem; they're not major things, and this actually sounds good already. It's just advice for further improvement.

Oh, also, in terms of having overlapping treble frequencies, it may just appear to be that way because the saw chords are very detuned. Not necessarily a bad thing, as soundcloud encodes in 128 kbps, which can degrade your upper-treble frequencies anyhow. I think it sounds sufficiently well-mixed up there. What I would examine are the drops at 1:00 and 2:13. To me, they kind of don't live up to the hype you create in your buildups. I'm not sure if a cymbal is just buried in the mix, but an audible cymbal at those transition points should help improve the drops. I also think at 1:43, the lead could be more different in tone from the panning arpeggio. Right now it sounds like the same waveform, which blends together "too well" and decreases the clarity of the mix. What if you added a filter LFO for motion, and some distortion for harmonic strength? Other than those nitpicks, I think this is pretty close to what you were going for.

I do too, but I never use it, because it's not an insta-solution, per se. No computer algorithm quite matches the "randomness" of a human's rhythm, sustained-note length, and playing intensities (especially if the same humanization preset is applied each time), so having a MIDI keyboard helps me get over those barriers. Yes, I do have 8+ years of piano experience, but honestly, I could be much better at improvising. I still have enough ability to play chords (for the human rhythm), solos/melodies (hoping for that happy accident!), and short passages that I can touch up later. However, I'm not one of those guys that plays things in perfectly the first/second time, and I could probably play what I play today, just as capably with 1 year of piano experience or less. I'm not bragging, by the way - I actually mean that in a humble and encouraging way. You can feel free to apply the advice to your own music or just discuss it, but I think many people (including you) can benefit from applying what was said in this discussion. IMO, it's one of those skills I think you or others can learn early, and it can be a nice tool to have while you work on other related things, like hearing chords in your head, picturing soundscapes before you write them, imagining melodies before writing them, etc.

In that case, it would help to do so, because it may be what @germanjazzguy is hearing. It's not exactly obvious, but I hear it too: if you don't sidechain the kick with the bass, since your kick happens to have a long meaty tail, the frequencies at 20~100 Hz can clash with the bass at nearly the same frequency range, creating what can be called muddiness (clashing bass frequencies). If you try EQing the bass down where the kick's bass frequencies are, it sacrifices power for clean bass, which you probably don't want. Sidechaining the kick with the bass allows the bass to duck as the kick plays, and there should be a bit less muddiness as a result, and you don't have to sacrifice some power for a clean bass frequency blend.

Are you sidechaining your kick with your bass? You seem to be doing it with your synth pads.

i don't think any good composer wants something to sound completely mechanical (100% perfectly-quantized rhythm, zero variation in velocities), because it sounds stiff and lifeless. It's not an "obsession"; it's the product of a trained ear. Mechanical notes are the equivalent of a real person playing with zero emotion, and evidently, a person cannot have 100% perfectly-quantized rhythm, nor can they have perfectly even note intensities in real life - that would be extremely unusual. For pianists as an example, a real person does not plunk down notes at the exact robotic intensity or rhythm on every single note (unless they truly try to play badly by slamming a finger onto a piano for every single note, which is not how anyone is supposed to play piano, even beginners who don't have experience using all their fingers on a piano). When we on the forums ask for humanization for a track to be submitted to OCR, we usually aren't saying, "go make it so no one, not even the best digital music composer in the world, can tell it's a sample." We're often saying, "make it so it's not so stiff in rhythm and intensity, so that there are actual dynamics in the song." The point of humanization is to give a sensation of dynamics, and that doesn't always require realistic rhythms and intensities. As long as the rhythm is not 100% quantized, and as long as the note intensities are varied, that's a step towards humanization - at that point, you should ask yourself, "is this how a human being would play it [approximately]?", and adjust further until at the time, you think it sounds good. It doesn't have to be convincing to everyone, but the "general audience" shouldn't be able to tell that much. Sometimes the rhythm and intensity adjustments are subtle, or sometimes they're kind of obvious, depending on your experience in rhythm. For example: Mechanical rhythm + intensity Mechanical rhythm Mechanical intensity Fairly realistic You should also consider the context - is the instrument in question exposed, or mixed down pretty well that it becomes harder to hear the qualities that tell you it needs humanization? So, for a solo piano performance, humanization is absolutely crucial; for something that features a piano but not at the forefront, probably not as necessary to be uber-realistic... With certain OSTs, people don't necessarily shoot for "man, I can't tell if this is a real band or not." As long as it fits the mood, era, etc. for the game, and the composer and other participants are happy with it, I think it's fine. It's what they were going for, probably. But I would say those standards are associated with the fact that the relevant OSTs are probably pretty old, like 8-bit, 16-bit, stuff that's pre-2000s - early 2000s. Or, it's based on dated styles. I don't think anyone expects something crazy-realistic from, say, Pokemon Sapphire's OST, but it fits for a GBA game. If it "holds up song writing", then it just means you need more practice working faster. It honestly doesn't make me slower, because I'm used to it, and I adjust velocities and rhythm as I go, instead of putting it off until the end. That way it doesn't feel like a lot of work left to do later. You don't want to grade 48 lab reports the day before they're due; you want to grade a few every day until you have to turn them in later, so that it's not such a pain for you.

If you know the principles behind how to make the lead, you should be able to make it. The main difference between synths is the approach to making certain sounds, except for sounds that are particularly complicated (such as some FM sounds, or metallic sounds). Try going for a detuned saw lead with multiple voices, and a slight envelope depth on a low-pass filter's cutoff frequency. Further tweaking would be needed to get it exactly, but that should get you most of the way there.

Basically, it should sound like one cohesive song if you aren't familiar with the sources, rather than multiple songs pasted together. IMO, key things to focus on include (but are not limited to) good transitions, smart source overlapping (for "bonus points"), that the arrangement doesn't meander (since it's a long piece), and that the sources aren't just inserted for the sake of insertion, but that they make sense that way. And besides that, the usual expectation of solid production and creative+valid arrangement.

huh, I feel like the smiley was changed. Or maybe it's just because of a different computer. Anyways: Yeah, I definitely agree with Gario that having just textural additions on top of relatively conservative source isn't inherently insufficient. There may be more merit in changing the form of the music, or not, depending on what the textural additions actually are. I don't mind that it's basically textural and production-based changes happening around the source arpeggio; but I would have considered: does it make sense, and does it make a large enough improvement to the original that it counts as a separate interpretation? In this case, I think the production would have to be pretty solid to carry the composition. That's my take on it anyhow. I don't inherently mind the textures chosen (they make sense to me). I don't have my good headphones with me right now, but I think the production's still a bit rough, with a few muddy (EX: 1:11 - 1:32) or resonant spots (EX: 2:16). I think you also have some sub frequencies (check your kick and bass, and maybe your sidechaining) that are adding some overcompression in spots with drums. In terms of a reference mix for an (IMO) esoteric track like this, maybe this is suitable, since it has some fairly abrasive textures, and unusual processing.

About 0:22, yeah, the crackling was an aesthetic choice that was aimed to create some unease; what I did was increase the bitcrushing depth on the entire mix, which could sound like your speakers' wiring got loosened. Mostly, it was inspired by Clem/Redg. Ah, okay, that would be a zourna (although I see how it can sound like bagpipes in its nasality); as for what I think you're taking issue with (the playing style?), it's actually a common way you could play a zourna in real life, so if that trill is what it is that's kinda bugging you, it's something that the instrument can inherently do. Either way, thank you for the feedback! I had a lot of fun writing this, and AngelCityOutlaw was a great help in keeping the arrangement ideas flowing, making the foundation that was to be fleshed out, and brainstorming what to do next.

Sí señor. And for collabs, you got my portion of the YESes.

hah, yeah, it's because I did my track in literally two days.

Yeah, fair enough; for the Torvus bog melody, it's really just a product of the way I wrote the CC11 + CC1, not because of the library I used; I just need more practice with that kind of emotional writing. If I had more time (and I weren't at WSU already), fixing the legato leads would be on the list. But for OCR I think this is sufficient.

Should totally pin that.

FWIW, you can freely monetize remix videos for my mixes (though with my collaborations you may have less luck getting the a-ok).

I'm not saying they don't matter. I'm saying few of them are rooted in anything more than built-up resentment or some sort of personal hate. I already read all of the ones here, and I honestly think that only a few of them are productive to listen to, and many of them are literal hate comments hid under slight euphemisms. There is value in noting that people have reacted this way and accounting for that in OCR's actions, but there is hardly any value in taking them as unbiased truth.