*NO* Myth 1 & 2 "Victory at Madrigal"

[Liontamer]

Artist Name: Zanezooked

Original Decision

RESUBMISSION of https://ocremix.org/community/topic/53085-no-myth-1-2-victory-at-madrigal/

My original submission received kind words for the arrangement but was criticized for the mastering and sample quality. Hopefully this revision addresses that. I also took the opportunity to make some other changes.

List of changes:

• Quality string libraries are now used: Spitfire Abbey Road One Soaring High Strings and Legendary Low Strings for violins, cellos, and basses; Audio Imperia Areia for violas; and EastWest Hollywood Strings for tremolos and other special dynamics. (The brass, by the way, is Aaron Venture Infinite Brass, and the brief flute and piccolo are from Aaron Venture Infinite Woodwinds. The solo cello is the Audio Modeling SWAM Solo Cello performed on an Akai EWI USB.)
• Mixing and EQing was completely redone, track by track, using Izotope Neutron
• Mastering was redone using Izotope Ozone.
• The solo cello has been carefully EQed using reference recordings of real solo cellos, and placed in the mix better to make it feel more like an organic part of the whole. It no longer blares over the whole orchestra.
• In the mid-track crescendo leading into the triumphant stings, the female choir now sings a falling sequence, while the male choir still sings the original rising sequence. This signifies Shiver's fall into despair and Rabican's growing power over her. Also, a falling female choir sounds better—less shrill.
• The female choir has been redone using six AI voices from ACE Studio, layered together, and then layered with the original EastWest Hollywood Choirs. I had asked about potentially doing this in one of the threads about AI and no one objected, so hopefully this is okay. AI did none of the compositional work for me. The process of crafting the AI voices is exactly the same as the process I used originally with the sample-based EastWest Hollywood Choirs: I placed all the notes, assigned them phonemes, and then tweaked expression parameters to try to get something that sounds natural. The difference is that you can actually hear the consonants and syllables with the ACE Studio singers. This ad hoc choir wouldn’t work if the AI singers were exposed, but as part of a mix I’m pleased with how it turned out.
• At the end of the track I added an extra bar between the brass statement of the theme and the return to the piano and cello, to provide an extra breath.

Original submission notes:

 

I’ve been a regular visitor to OCRemix since 2002. A significant portion of my music library comes from here. So it’s with some trepidation that I send my first submission. And not only am I daring to submit to OCRemix, but I also have the hubris to remix the great Michael Salvatori and Mr. Halo himself, Marty O’Donnell!

“The Siege of Madrigal” is a piano piece by Michael Salvatori. It’s one of two simple but gorgeous melodies Salvatori composed “in the style of Marty” which Marty O’Donnell loved so much he integrated in multiple ways into the soundtracks of their respective games (the other is “Unforgotten” from Halo 2, which saw a return as “Never Forget” in Halo 3). Marty loved “The Siege of Madrigal” so much he even hid it inside Halo, so for many people it’s known as “the Halo Easter Egg Song.”

I heard it in Myth, though, years before Halo, so I know it as the song that plays as the war-weary, anonymous narrator describes a last-chance plan to strike the armies of the Dark and their leader, the Fallen Lord Shiver, and break the siege that is strangling the city of Madrigal.

Myth is a game about small, desperate battles fought by a single battalion on the fringes of a greater, losing war, so we don’t directly see what happens at Madrigal; the player spends the mission running a diversionary raid to try to draw some of the enemy away from the city. But in the following mission briefing we are told what happened:

“The battle for Madrigal lasted four days without pause. Shiver fell on the first night in a spectacular dream duel with Rabican, one of the Nine. No one expected this. We have never before challenged one of The Fallen and won. . . . They say that The Head had an old score to settle with Shiver, and told Rabican that her one weakness was vanity, and showed him how to exploit it.”

This sudden, unexpected victory in the face of certain defeat is what Tolkien would term a eucatastrophe. It captured my imagination. I tried to tell that story in this remix.

We start with the Siege of Madrigal theme in piano in its original C major, but with the rising quartet of notes between phrases removed to make it more hesitant. The addition of harmonic high strings underlines the uncertainty.

At 0:30 a solo cello picks up the theme. Marty O’Donnell made the low cello the featured instrument in Myth, so this seemed appropriate. (Also, I love the cello.) The rising quartet of notes is restored, making the theme a little more hopeful.

At 0:59 we get a sweeping string statement of the Madrigal theme, but the peak of the melody is lowered from A to F, making it more somber and unresolved.

At 1:28 the piano comes back along with high string tremolos, but instead of finishing the melodic phrase with a D, it modulates into an unresolved and uneasy D#, taking us into the key Db, while we also switch from 3/4 to 7/8 time.

The next part represents the “dream duel” between Rabican and Shiver. What’s a dream duel? It’s never specified. I imagined each duelist attempting to invade the other’s mind and plant thoughts that undermine their strength and open them to attack. Shiver’s weakness was vanity, which seems strange for a 1000-year-old undead crone who looks the part. But she wasn’t always an undead crone. So perhaps Rabican’s strategy is to remind her of what she has lost. She may be immortal, but is it worth it if this is how she spends eternity?

Rabican’s persistent attacks on Shiver’s psyche are represented by a male choir and a creepy, guttural chant in Irish, with Shiver’s thoughts are represented by a female choir also singing Irish. Irish names and words are often used in Myth, though not very accurately—more for flavor—so this seemed appropriate. I used Google Translate and then found an audio pronunciation for each word, and I’m sure I mangled it, and it’s not like it’s understandable anyway, but hey, it gave me something to work with, and I thought it was cool.

At 1:31 the male chant begins:

D’aghaidh, a cloigeann (Your face, a skull)

Do chuid gruaige, baile do lucha (Your hair, a home for mice)

D’aghaidh, a cloigeann (Your face, a skull)

At 1:45 the low cello starts an aggressive, grinding ostinato. This is the "Great Library” theme from Myth II: Soulblighter, by Marty O’Donnell.

At 2:01 the female choir sings a fast, 7/8 version of "Fallen Lords" theme from Myth: The Fallen Lords, by Marty O’Donnell:

Cad é an aghaidh seo? (What face is this?)

Cén craiceann, cén gruaig é seo? (What skin, what hair is this?)

Cá bhfuil na leannáin a ghlac chugam? (Where are the lovers who embraced me?)

2:22: the men’s choir rises ominously, doubled in horns to give their line punch as Rabican pushes the blade as deep as he can:

Níor bhain tú bás (You have not defeated death)

2:30 the women sing,

Is fear céile éad é bás (Death is a jealous husband)

Is fear céile éad é bás (Death is a jealous husband)

2:37 men and women sing together:

Tá a bharróg síoraí (His embrace is forever)

The section builds to a frantic crescendo and then the clouds part at 2:56 as the Madrigal theme returns with its high A restored. The low cello is still prominent but now rhythmically supports the triumphant tone of the strings.

At 3:23 we get a final, driving reprise of the theme in full glory, with horns and trombones taking the melody, a soaring counterpoint in the trumpets, and a bodhrán (Irish drum) pulling us forward.

At 3:51 we end with a final statement of the Madrigal theme on cello and piano, made hopeful by borrowing a high sequence from the earlier trumpet counterpoint. There’s an image from Myth I was thinking of here, a soldier standing on a battlefield, arms spread, face to the opened heavens. The battle is won. Weariness and joy mingle like the rain.

(Madrigal would be utterly destroyed months later. But don’t tell our soldier that.)

I hope this is an enjoyable remix!

---

Games & Sources

Myth: The Fallen Lords

• The Siege of Madrigal, by Michael Salvatori
• The Fallen Lords, by Marty O'Donnell

Myth II: Soulblighter

• The Great Library, by Marty O'Donnell 

Edited 19 hours ago by Emunator

[prophetik music]

my original vote criticized the instrumentation but loved the arrangement and concept.

this has 4.6db of headroom.

opens with piano (straight from the Siege track, to my ears) with some nice light strings over top. cello carries the melody after that and sounds great, really rich tone. it might be a touch loud, and the vibrato's a little aggressive in parts. 0:59's strings sound better than last time for sure.

1:32's change is still abrupt but not quite as weird as it was before, i think. cello and subsequent vocals sound a lot better. there's a highly artificial artifact in the waveform through this section that i don't know is intentional - it looks like you managed a volume shift by the gain knob on your limiter, which is kind of wild. it coincides with the entry of the choirs.

this section with the two choirs sounds good, but i think the men's choir is a bit loud as it's hard to hear things beyond it when it's going full-on.

2:57's idea is super fun. from an execution standpoint, there's nothing in the middle of the chord stack besides some flute runs. there needs to be something supporting the higher strings so that the bass doesn't feel disconnected. when the brass drums come in, it fleshes out a lot. the drums really run over a lot in the mix though, so some more EQing may be needed there.

the ending is a call-back to the Siege track, and feels pretty similar to the opening.

from an execution standpoint, this is a lot better than last time. the strings overall are gobs better, the instrumentation has a lot more individuality to it, and the choirs sound great. i think right now though that the overall instrumentation and use of samples is right on the edge of where i'd consider it to need to be to be posted. there's some missteps in the mastering (like the density of the end section around 3:25) that i think keep it on the NO side for me, but i'm honestly going to wait on a few more votes before i make my decision. i think this is pretty borderline but the arrangement is really fun.

 

 

?

edit 10/8: i am going to NO this. i believe the execution ultimately falls down in too many places. some major elements that come to mind is the rigid piano scoring, hard vibrato on the cello, 2:20ish's seemingly disparate elements, 2:57's emptiness under the string tune (just the cello and flute runs, but nothing supporting the melodic line strongly enough to count), and the compressor engaging at 3:21ish thanks to the drums having no room tone to them and the emptiness in the scoring at that same section. again, i love the idea - it's just not there on the execution side.

NO

Edited October 8 by prophetik music

[Chimpazilla]

Still way too much headroom on the mastering side.  Why?  YouTube will normalize it, but people downloading the track will need to turn up their volume.

Strings still don't sound completely natural, but this type of arrangement is very hard to get sounding completely real.  The samples are much better than last time.  The choir is much better.   The cello sounds uncanny because it seems only one articulation has been used, a legato articulation with vibrato that enters very quickly.  Each note is the same, and on the longer notes the entry of that vibrato is very evident.  A keyswitched patch would have been significantly better, but I think this works well enough.  The cello lead gets severely drowned out during the dense section, I can still hear it but it becomes part of the heavy soundscape at that section rather than riding above it as a lead instrument should.

The mastering numbers indicate to me that the track is bordering on overcompression, while the peak value remains low.  I'm seeing -4.4db peak which is odd.  The RMS is showing as -10.6db which is quite loud.  There are points in the track where I see a swell in the sub-30Hz range, often it is the timpani doing that.  Those sub swells, along with any stray low-lows on other instruments that didn't get EQ'd out, are causing a sub/low pileup of frequencies that are stealing your mastering headroom and causing this situation.  You'll have more mastering headroom without having to smash things, if you EQ low-lows and subs out of anything that doesn't need that range playing.  And your peak value should be hitting more like -0.5db.  Even so, I'm not hearing overcompression artifacts so I don't find this to be a dealbreaker.

I still think this is a lovely arrangement of these themes.  I'm borderline here due to the samples still not sounding natural, and the strange mastering, but this is enough of an improvement for me to pass it.

YES (borderline)

[Emunator]

I'm coming in blind to this one, having not heard the original version. I can see why this is divisive - the highs are very high, and the lows are definitely noticeable and have an undeniable impact on the overall listening experience.

On the positive side, the arrangement here is really beautiful, emotive, and dynamic. It feels like you've gone to a lot of lengths to integrate past feedback here and push your sample quality up a notch, and overall I would say that you stuck the landing in terms of programming orchestral instruments, at least within the scope of what we can expect from a hobbyist community. It's not perfect, and I'll get into the specifics later, but this is not an easy task to pull off so you've got my respect for getting as far as you have. The string programming sounds pretty nice across the board, and the choir fits great. It's AI, but it sounds like it was utilized in a way more similar to Vocaloid software so I don't have an issue with that specifically.

Now for some of the issues I'm still hearing with the present version, which I'll bullet point to make it a little more digestible:

• The piano sequencing is quite rigid in terms of timing. It sounds like every note is quantized pretty heavily to the grid, which is not how it would be played in by a real pianist. You'd have subtle timing variations to your chords and arpeggios that would fall off the grid and give it a more natural flow, even when you're playing block chords. This is unfortunate because of how exposed the piano is in the intro (on the plus side, the tone and sample quality is fantastic, so this should be very fixable with your existing resources.)
• The cello tone sounds great, but I agree with Chimpazilla that the articulation feels awkward. It's funny, you mentioned that you controlled it with an EWI, and I actually can hear how that line would sound better on a woodwind instrument the way it was played. Do you have control of the vibrato or sample start on the SWAM cellos? If that's not already baked into the performance, I'd look into that and see if you can vary the way that your cello notes trigger with each note. Not a dealbreaker but worth noting.
• The overall mastering/mix is interesting - if I had to guess, you used a streaming-based mastering preset, which sets your peak headroom quite a bit lower than CD quality. Ideally, in order to sound good within the OCR catalog, which is based on CD mastering standards, it should be peaking anywhere between +0.0 and -.5dB, so if nothing else, normalizing your volume to CD standards is a must here.
• Brad picked up on the use of a master volume automation to control dynamics, which is really not the way you want to do this. It sounds very unnatural even without looking at the waveform. You'd want to control these with individual volume automations or, since you're using high quality libraries with dynamic control, use MIDI CC1 or the mod wheel to control the built in dynamic layers with your instruments for much more realistic dynamic swells

I've listened to this about 5 or 6 times and can't help but be struck by the quality of the arrangement in spite of these issues above. At the very least, I think it's worth going conditional over a track that's mastered with such a drastic headroom dip compared to the rest of our catalog, especially when "quiet" doesn't seem to be what you're going for intentionally. I think the rest of the issues are "nice-to-haves", but I would love to share this feedback with the artist if this eventually does pass so that they can have a chance to integrate some of the other feedback, since this is very borderline as-is and I don't think it has to be.

Curious to hear how this shakes out!

EDIT 11/23: Changed vote below.

NO (resub)

Edited 19 hours ago by Emunator

[XPRTNovice]

I mean, yes, we're all in agreement here on the mastering side of it. Needs to be louder, I'm turned up quite a bit just to hear everything.

But I have some other production/arrangement critiques. The repetitive cello lick that starts at around 2:00 goes on forever, and is the most prominent part of the arrangement. The strings often sound out of place, with the entrances/exits of them being really abrupt. The cello dominates for a huge part of this arrangement, but doesn't really have much to say.

I'll be honest, I love this arrangement. I think it's absolutely beautiful, and I feel sad that I have to NO this based on what I'm hearing. This is really ambitious, but the strings overall kept kicking me out of the arrangement. The attacks and releases need to be smoother, more carefully planned, specifically because you have them as such a prominent part of the arrangement. The highs are often really high, and the cello/dbl bass sticks out like a sore thumb through a lot of the arrangement. The real issue here is overall blending. The 1:33 violin sustain is very loud compared to the rest of what's going on, and hardpanned to my left ear so it's just sitting there out in the open. The drums and vocals, to me, are mixed the best, and they're the feature of the piece, but the foundation sustaining them isn't working.

I want to emphasize how much I love this, and how ambitious this is, but we're falling really short in the mixing department for me on this one. 

 

NO (resubmit)

[Hemophiliac]

So, I've been waffling back and forth on this one for a while now.

The arrangement (aside from the sudden transition at 1:31) is good. All themes have been incorporated and have added plenty of your own interpretation to them. The macro-structure tells a story and the piece goes places, bravo.

Where this needs improvement is the production and sequencing. Most of the time the samples feel like they have been handled ok (but could have been better with more attention to detail and articulation switching) and you get by with what you have. However, the sore spot is the lead cello.  A few times it is too forward in the mix when it should be more blended (1:46-2:24, gets better when the male choir enters). 0:46 on the cello has a strange flanged quality to it for a moment.

I think the arrangement is strong enough to overlook some of the production/sequencing issues here.

YES (borderline)

Edited October 10 by Hemophiliac

[Flexstyle]

The mastering issue here is that it's brickwalled like crazy, but then the master output is volume-automated turned down. Really weird. Everything sounds squished. Ease up on the master compression and let the whole thing use all its headroom with some actual dynamic range, maybe? Other Js cover good points so I'm not going to go in depth on samples, arrangement, etc. Call me a single-issue voter if you must, but I'm voting....

 

NO

[Rexy]

I haven't heard the original submission, but even with the resub, the arrangement is completely on point with the shaping of the main source, integration of smaller source sections for the sake of drama, and the overall shaping of it all - so for once, I don't have a reason to break this down. :-)

However, my most significant issues while listening were with the sequencing and production.  Credit where credit is due, the choice of samples is solid, and hearing that you've performed the cello part with an EWI makes me understand the overall intention of its placement, velocity variation, and all.  Wes already touched upon the main body of attack being something that can get touched up with envelopes away from the EWI to make it feel more authentic, rather than touching upon the uncanny valley, so that's one aspect to shape that up. 

But then you have parts like your piano that sound like mouse inputs with no velocity or note length variation, and your backing instrumentation with the strings and brass has similar issues.  The percussion is way more expressive in comparison, though once that leaves for the final theme repetition at 2:57 before the hand drum returns at 3:21, there's a whole bunch of exposed mid-frequency space that makes it feel empty.  I recommend transposing some of your higher strings down an octave so that you can fill that mid-space more efficiently and that your flute writing also gets some breathing room.

This idea is brilliant.  I understand that you've gone to great lengths to work with better samples now, but some extra attention to articulation and frequency space would get this over the bar for me.  Please send over a third version!

NO (resubmit)

[Emunator]

I listened back to this and see the argument for the NO more strongly this time. I was obviously very borderline at first, so there's no need to hold this up - let's get this polished up and hopefully the next version will jump over the bar rather than barely eeking over it.