timaeus222

Harmonically, this actually sounds really good. Harpsichord can be tough to make sound humanized. The reverb helped, but careful velocities help even more. The only thing I'd say is to work on working around free orchestral material some more to make them sound as realistic as possible, and that involves volume swelling automations, using more articulations, and sequencing with regards to the attack of the instrument (for example, strings). You've improved a lot with what you're using, from what I can hear. The sounds you used here actually work very well together. Bring this level of dedication to the WCRG!

Really? I already said please use careful and nice wording. That just makes it seem like you hate me. I'm not trying to come across as a know-it-all, I'm just giving advice on what I already know. I've already tried and tested everything I advised on that post and I know it works. I just didn't have time to synthesize everything there and provide an example on every single thing there. SnappleMan would say "squashing the transient is not the goal here", and it really isn't, in my opinion and in his opinion. In fact, while doing parallel compression it's really important not to squash your transients overall, in the end. You can lose really crucial power characteristics in your drums if you do. This is assuming you have a wet mix knob in your compressor. Yes, you want glue, but too much and it breaks up the frequencies in the top end a bit. Remember that it's going to stack up in your mix. By the time you add in other instruments, it'll make the drums boost a little bit more for every instrument that plays at the same time. That's why it's important to have a make-up gain slightly lower than what you would prefer, and it wouldn't be a bad idea to adjust later if it's not quite there. I was recreating something the other day that involved really heavy drums, and I noticed my snare breaking up a bit, so I lowered the gain until it stopped, and it just made a huge difference in energy for me. However, that was probably the only time I ever really felt like the transients were squashed to an unfavorable extent; one of the drum layers I used had a relatively long decay already, and it was specifically a drum sample I'd never used before. That said, the ratio really doesn't have to be that high every single time. It should only be high enough to give the snap you want, but not so high that you lose top end information like in the example I provided above. I tend to just do 2:1 or 4:1, even on drums. And in the end, you'll be turning the wet mix knob down to ~50% to help even more on that aspect provided you have a compressor that can do that, such as The Glue.

Thanks. That's pretty much what I was intending to convey with regard to the brass/vocal compression. Those waveforms tend to be very wonky. When you first start out with compression, though, it's not a good idea to "just do it" on brass and/or vocals "just because". Know why, and know when it's good or bad. Also, yay, you spelled my non-numerical name right. And yeah, these days I try to give as accurate advice as I can. If I don't know it 100%, I might say "-esque" or "you could try this and see how it goes" or something. It would be rather noticeable. I am rather glad I don't say "This is 100% true, and you can't deny it" unless it's totally, completely, undeniably true (like majorly heavy reverb on a blasting kick drum playing 32nd notes creating ridiculously blatant mud). And uh... What is "P.S.S.H"? xD

That's the important thing to consider. There are loads of technical aspects to composition and production that some people can describe as purely technical, and some things that can be understood via opinion, but I've seen many people meld those two ideas together and assume that it's just all subjective all the time. For example, very bright, annoying thin resonances that bug people's ears is not purely opinion. You can't say "It doesn't hurt my ears at all" unless you can't hear it at all (or you've heard it all through your life 24/7 for some apparent reason and got used to it), in which case it doesn't apply to you. It's a technical detail that really holds true for anyone who can hear resonances of that nature. Gario goes into some detail with this in his fourth paragraph. More examples: Saying something is too loud. If you say the whole song is too loud, it could just have really tight compression, like . If you call that song too loud while you listen to mostly quiet music, it's subjective because it's new to you and you aren't accustomed to loudness like that. If you say the whole song is too loud, and you're somebody like DarkeSword who knows what bad loudness truly is, then it's mostly objective with a hint of subjective. If you say a lead sound is too loud and upfront in a mix and it's actually dry and unprocessed, it's objective. If you say there's too much bass at 64Hz, it's objective because you pinpointed where it was. It's all about your experience and when you can say with accuracy whether you like or don't like something because of your opinion or because of technical details.

Yep. You don't have to necessarily agree with me, but all I'm doing is sharing how I do things. I like my results, and I want other people to share in the enjoyment and get results they like too, however they want to do it. Besides, I was constantly telling myself to write "might work" or "could work" instead of "should work", "works well" instead of "works best", "rather well" instead of "really well", etc. Anything to make myself come across reasonably non-biased.

Hey, I'm not trying to teach bad habits. And it's fine to compress brass or vocals if it's transparently done. I just don't recommend it heartily. With the assumption that you're a complete master at compression, if you hear literally no difference after compression (i.e. no transients have been ruined) but you see the difference you want to see in a spectral analyzer, then it isn't detrimental to a significant enough extent to worry about it. How else would you process vocals with a completely wacky dynamic range, then? Nobody sings with a perfect, relatively flat dynamic range, not even Jillian Aversa... I'm giving advice that I believe is truly accurate. I'm not trying to mislead anyone. I'm not going to say anything that is blatantly wrong and off-base. What I said was NOT 100% subjective. I wouldn't have said all that if I wasn't truly confident in what I currently know, and I'm not aiming to brag or act arrogant. If you don't believe I'm qualified to give advice, just look at some examples on my website and decide for yourself. Besides, many people were still learning and at their early stages of experience in 2007-08. You can't expect everyone to be masters then. I'm pretty tired of people thinking I don't know anything simply because I'm giving advice. I wouldn't be giving advice if I thought it was wrong outright. You've heard my remixes, you know what I can do. Believe me if you want, but I'm trying to be a nice, helpful guy. Just roll with it. If you're going to criticize me or dispute me, at least use careful and nice wording.

Composition and Arrangement: 1) Another thing would be to make sure they're in the same "mood". It won't help you if you have the same notes for a while, then suddenly a clash because one source is using a major scale and another is using a Mohammedan scale. 3) To build on that, you should be picking instruments that make your EQing job easier. i.e. Don't pick two supersaw leads that are doing a duet throughout with no panning. 4) Any reverses are good for transitions, but percussive rolls, sweeps, and risers would work too. Mixing: 1) Building on this, you should be EQing in such a way that your leads come out sufficiently. Don't do isolated (while the instrument is soloed) boosts just to make your instrument sound good if it clashes with the lead, but don't do strictly narrow bands all the time for peaking or notching EQ either. EQ in context. They don't just "affect" the frequency range; they shrink it. Low pass filters cut out the high end as you lower the crossover frequency, and high pass filters cut out the low end as you raise the crossover frequency. The crossover frequency should just be wherever the band token currently is. You can automate these band tokens to do compensation EQ in certain spots if you are going to do something that requires filter work in only that section to resolve inevitable clashing. Remember to reverse the automation by the time you get out of that section if the new section doesn't require it. An example would be to automate a high pass on a bass sound that has good treble content but weak low bass content in order to make room for another bass with little treble content but better low bass content to layer it and make a heavier section. However, "louder bass content" doesn't necessarily mean "better bass content". 2) Compression is mainly to tighten and strengthen drums, and to tame wild waveforms on non-percussive instruments such as brass sections and vocals, depending on the settings. Try to do it transparently if it's on non-percussion. Sidechaining is typically done with a limiter which can act as a compressor as well. FL Studio's limiter, for example, is a compressor, noise gater, and limiter. 3) Automating waveform changes is interesting; if your synth works with wavetable synthesis, then they likely have a Wave knob to automate via MIDI CC. It's a rather complex concept though, as not many free synths can do it. ------------------------- Selecting and writing the "right" lead sound (appealing to as many people as you can): 1. First, think about the mood of the section you're writing. Is it high energy? What drum rhythm are you using? If it's Drum & Bass, then fitting leads might include timbre-morphing leads and filter envelope leads (long attack if a negative depth, long decay if a positive depth) with relatively late vibrato. If it's Half-time (Dubstep, for example), fitting leads might include portamento-abusing leads and leads with really heavy vibrato (assign an LFO to the pitch with a depth of a half step or greater, then lower the mix level as necessary). If it's chillout/downtempo, then fitting leads might include mallet instruments, soaring saw leads with subtle morphing timbres, various types of sine waves (pure sine, dual AM sine, quadric sine [not quadratic], etc.), and so on. If it's something completely unclassifiable (like Drum & Bass in 6/4 shifting to 5/4 but using a modified 4/4 rhythm), just look at the energy and go with that. 2. Next, think about the class or purpose of sound you want, narrow down the technique you would use to create them (or just find them somehow), and further craft the tone. If it's aggressive, sync leads, supersaws (to an extent), thick leads (vibrato at a really high LFO rate), detuned saw leads, and similar could work rather well. If it's emotive, moog leads, less aggressive (semi-tame) sync leads, distorted pulse wave leads (think Sytrus's Hyper resonant preset), and detuned saw leads with slightly slow attacks and heavy vibrato (like a mourning lead) should work rather well. If it's calm, slow low-pass-filter-envelope leads, simple sine wave leads (again, many different types depending on the synth), and maybe moog leads would work here too. 3. Now, look at the tone in an objective manner. What do you want to do to perfect it? Is it too resonant? If so, find the reason and fix it. Usually it's a resonance knob turned up too high, other times it could be an FM knob turned up too high creating an unwanted thin resonance, and other times it could be excess treble as a result of too much waveshaping and distortion work. Are the frequencies going to let it fit into the song reasonably well? If it's not quite there, don't EQ with your DAW yet. Try to see if there are internal methods of EQing first. What if the sound is too thin? You can't "add" frequencies that were never there in the first place. Does it fit the mood at all or is it completely out of place? Use other techniques to create a completely different lead sound if it doesn't seem to fit. As you start out, you might want to ask for someone else's opinion. Eventually you can tell for yourself whether it fits or not. Is it expressive enough? What can you do to fix that? Portamento/Glide, Vibrato, Reasonably long Filter Envelopes, Wavetable Shifting, Oscillator Sync LFOs, Frequency Modulation Depth (FM Depth), etc. Loads of options. 4. How is the lead supposed to sound in terms of phrasing? Think of a lead sound as an organic instrument, and pretend it has a "perfect" way for phrasing, and create that "perfect" way. Should it be staccato sometimes and legato other times? Do you want portamento in some spots to accent those notes? Does it have a level of glide/portamento that won't sound weird when you do grace notes? Does it sound "right" playing faster here and slower there? Do elongated notes work better or should this be a solo? What is the pitch range, and in what way should the lead move within that range? Is it legato or retrigger? Most of the time, legato works better, but retrigger could work I guess. You're better off sequencing in legato mode, and then not overlapping notes whenever you want retrigger. It gives more flexibility, but it really depends on how you're modulating your filters; envelopes or LFOs? Selecting bass tones and writing "good" bass parts (appealing to as many people as you can): 1. What mood are you writing for? Is it supposed to be ambient? Sustains should work well, but for longer notes the tone should preferentially be slightly evolving, or a pad should be evolving if you don't have access to an evolving bass. That way it draws attention more than just a simple static sustain. Is it supposed to be funky? Try a 50/50 split or simply a 50% mix level on a wah pedal (or a wah pedal emulation/imitation) on an electric bass sound, or find a clean/relatively unprocessed slap bass sound. Not the only way, though. Is it supposed to be intense or heavy? You might need basses that use rich, deep-sounding filters. Vintage filters will give you bass tones that can really nail those low notes and not "fail". I've heard bass sounds fail below A3 before, and it sounds pretty bad. You should hear the tone breaking up if it's "failing". Is it dubstep (super specific here)? How aggressive is it? You probably need incredibly gritty, unforgiving tones if it's aggressive, and if it's not intense, the basic LFO linked to a slightly resonant cutoff on a thick oscillator tone should be fine. Aggressive tones can be accomplished using waveshaping/distortion, FM synthesis, bitcrushing (I showed that earlier), and rarely, granular resynthesis (not synthesis, necessarily). 2. Think about the class or purpose of sound you want. Narrow down the technique you need to use to create them, then work on it. Ambient basses could be rich FM basses, low pad drones, etc. Something soft to go with the calm atmosphere. Funk basses could be wah fingered electric bass, slap bass, Yamaha CS-80 squelchy basses, Roland Juno pitch envelope basses, etc. Anything oldschool could work very well. Intense basses could be vintage filter basses, heavily waveshaped/distorted would-have-been-a-dubstep-wobble basses (without the LFO, basically), etc. Very detailed work is recommended. Dubstep basses could be FM basses ("eeyeah", vocaloid, etc.), bitcrushed basses ("yah"), basic LFO-to-cutoff-on-slightly-resonant-filter basses, etc. Very detailed work is recommended. 3. Now, look at the tone in an objective manner. What do you want to do to perfect it? Is it failing at the lower notes? If so, just find another filter model to use that can handle lower notes better. Vintage filters can, for example. Sometimes you're just playing way too low in general. Anything lower than C3 is difficult to make work unless it's a sine wave or a patch that is tuned really high already. Is it too resonant if it's a dubstep bass? If so, lower the resonance. Is it terribly weak and too forgiving? Try more waveshaping and distortion; more detuning; put filters in parallel to louden them. Increase the volume reasonably. That should help things. Is it too thin but still accomplishing the filter work you wanted? Thicken it with some detuning, external chorus, etc. More voices helps too. One voice can't detune effectively, obviously. 4. How is the bass supposed to sound in terms of phrasing? Think of a bass sound as an organic instrument, and pretend it has a "perfect" way for phrasing, and create that "perfect" way. Is it supposed to be playing fast notes or slow notes? Write accordingly. Is it emulating a real bass? Sequence or keyboard it in realistically and reasonably. Is it in retrigger mode or legato mode? Retrigger works well if the filter doesn't change much, and sounds terrible if the filter does change significantly. If the filter changes over the course of a long time, retrigger won't let that shine. Legato mode works decently if it's "true legato" rather than glide/portamento legato. i.e. Some sample libraries have "true legato" for their bass patches which crossfades two notes sometimes, and other times it fades out the first note, fades in a real hammer-on/pull-off sample, fades that out, then fades in the second note, all in a short time frame. The Shreddage II manual talks about this. Note range? Is it outside the range of a bass, or is it emulating the which is more than just a bass? I speak from experience, and I'm not simply spouting nonsense or misleading material. I'm just saying what I learned. There is NOT 100% subjectivity in this, nor 100% objectivity in this. These are things I advise, not "rules". I do things this way and I get results I love. That's all. Good day all!

This source could really go any way in a remix. I've heard a shorter source go for a longer remix, so it's possible for this to pass. I recognized this all the way through, but I'd say for the arrangement, add more source recognizability in there. It might ring in as ~40% source. I really liked the gating and the bitcrushing sweeps. 2:51 = iZotope Stutter Edit?

Yeahhhhhhh! I love VirtualDub! I also use Sony Vegas, and it's practically something I can do in my sleep now. =D I'd like to get After Effects someday, but it's not a necessity for me.

You could probably just try a bunch of notch EQs on other instruments at the snare fundamental and that should let it come through enough.

Yeah, there we go! Probably the last thing would be to see if you can layer a stronger snare, and that's about all I'd suggest. It currently sounds almost like a strong rimshot. I actually really like that low gritty bass (1:08).

For me, my round 2 was 29 hours and 18 minutes, and my round 4 was 21 hours and 5 minutes. I tend to overdo things and make them really intense though.

Usually it takes me 20~30 hours including arrangement and mixing/mastering. It's held true so far too.

DAAAAMN. I love that pseudo-sync bass (It's not made with typical oscillator sync, but with waveshaping and distortion on a pulse wave) and the intense saw lead. <3 Fun wahed EP too.

Sounds great! I hear a little overcompression but it's not bad. The Gratifier/Ultrasonic-esque guitar tone could have used a little less distortion, but other than that, intense mix.

I can say with confidence that my Grado SR-60i's ($80) are not going to fall apart anytime soon; I've had them for at least 6 months now and they look the same. And yes, headphones can't get sufficient bass necessarily, but you can get the best bass possible from them, IMO, if you use the Beyerdynamic DT-880's ($300~400). I don't have them, but I know they're awesome and I'd like to get them someday.

What is this?! xD I think you had fun with this.

This is lacking low end power below 65Hz, but the dubstep wobbles are okay at their octave IMO. The wobbles just need to be a tad louder (in mostly volume rather than EQ, from what I can hear) in the end if you end up making other parts louder, as the track overall, from what I see in my spectral analyzer, is not as loud volume-wise as it can be. The phasered bass by itself doesn't quite pull off the apparent aggressive nature of this track due to its lacking low end. Here's a bass sound that I believe could really nail the intro, for example. Listen on speakers and subwoofers to really hear and/or feel the sub-65Hz frequencies. Try to create something similar to that in aggressiveness if you think you need it, or ask me to collab or something. The drums also lack power and some truly over-the-top tight compression from what I can hear. Here's an example of a completely jacked-up version of the intro (in a good way). I only did stuff to the main bass and the kick/snare though. The off-sync glitching on the bass kind of confuses me a bit, so you might want to look into that and revise it a little to see if you can do some more rhythmically flowing gating and retrigger. I didn't know what you did with that rhythmically, so I just did something that was close. I think the wobbles fit fine as they are, but they wind up leaving me wanting them to be thicker than they are now. Preferably, they could have more waveshaping and distortion for a grittier tone, and slightly less resonance for less of a thin "wah" sound and more of a darker "wuh" sound. It doesn't have to be complicated necessarily; just a bit thicker would be nice. Aside from that, like Argle pointed out, some of the leads are a bit buried sometimes under the bass instruments (which doesn't really have much bass frequency content like I said above). The leads need to be louder in volume, could be less thick (thinner than a supersaw, thicker than a sine wave) to help them fit into the soundscape more easily, and some mids would preferentially need to be scooped mildly to bring out the leads. Scooping the mids might leave a person less prone to boosting the leads too much (if frequencies were freed up for something to come through, you don't have to raise the volume so much of what you want to come through). As for the lead tones, I'm gonna show you two synth leads that I think would fit somewhere in the mix. I'll timestamp where I think each one might fit best, and that might give you ideas, timbre-wise. 80's Acid House Bass as Lead (Loads of internal Waveshaping and Distortion on a resonant pulse wave, LFO on the cutoff) ~ Could work at 0:13 Detuned Sync Square Lead (Detune a square lead and put on some mild oscillator sync) ~ Could work at 0:34

The only real downside that I found was that the really low bass (20~80Hz) is basically mildly high passed, but you don't really notice it until you're in a situation where you have to mix bass rigorously. I'd say the ATH-M50's have better bass but slightly worse treble than the Grados from the frequency graphs.

You still have plenty of time to update the file before the judges see it. I think the 0:07 sawtooth lead could have less treble. Not too much, just maybe around 0.4~0.8dB less. At 1:48, the low pizzicato could be more bright. Other than that it sounds good.

You know what, I'm gonna dig up my supposedly finished Yu-Gi-Oh remix and give another pass-through on the bass mixing.

I'd actually say that humming ideas is just brewing ideas. You have the compositional idea in your head, but it's not a composition yet.

Submission Standards Last I recall, a medley consists of many songs. Medleys can be mashups too, but in that particular Zelda case it was a cohesive medley. However, it was not a traditional mash-up, as it was not two or more songs playing at the same time the whole entire time, but a song that introduces brief cameos of many songs. From the same site: If you plan to submit to OCR, then that's what I would suggest you do if you want to go through with this medley, as the traditional medley is too predictable, IMO.Also, your track is private.

Yeah, I can definitely make that.

The Grado SR-60i is actually really good. I find it to have really clear midrange, excellent treble (maybe 90% of the way towards sufficient), and defined bass. You still need subwoofers though to cover the whole spectrum.