Chameleon Jazz Connection promo album "Highway to Hell" - mastered by Studio Compyfox

[Compyfox]

My newsblog is currently in "overhaul mode" (I lost my backend admin, have to redo everything myself). I didn't want to post this in the music section, since it's not my own music, but I was doing the mastering.





I want to present you a nice little Berlin local jazz combo called "Chameleon Jazz Connection".

I had the chance to help them finishing the editing and doing the mastering for their promotion album "Highway to Hell", which acts as showcase for their live shows.

Since 21st of April 2013, you can find a new song every following Sunday on their page - free to download. They also have a SoundCloud and YouTube account, if streaming is more your thing. Though the YT songs a bit cut off. Probably due to encoding bugs.


You can follow the band at:
http://www.chameleonjazz.de/

And also on Facebook (which leads to their YT and SoundCloud Channels):
https://www.facebook.com/chameleonjazzconnection

You need to be registered however to follow their news entries. Those of us who are not on Facebook (I count to those), are a bit left out. But you still get access to the songs regardless. So please enjoy.





Some info's about the editing and mastering (for the tech interested):

The songs were recorded by Hannes Rackow. According to bandmember Normen Goltz (who did the mixing of the album), most of the tracks were recorded in a training room with limited-on-purpose recording gear. Among them, the infamous Glyn Johns drum micing technique. The vocals were later overdubbed at the home studio of Normen Goltz.

The mixes were done in Cubase, at 88kHz/24bit. A ton of tricks were used to get more "oomph" for certain instruments. For example: heavy parallel EQing/signal processing.

The mastering was then handled in my studio in Cubase 7, upsampled to 96/32 with a stand alone SRC tool, and later bit-reduced in Wavelab 7. I heavily relied on ToneBoosters EBU Loudness for aligning the tracks with each other. Though we decided to go for K-12/AZ+4 as absolute maximum loudness for fortissimo passages. Average it's K-12 (v1) - best of modern day loud mixed while still having enough dynamic in the songs.

Since the mixes were already great to begin with, I only fixed certain flaws and muddyness of the songs (read: I applied the infamous "fairy dust"). The wish was to have the mixes run through a tube device (initially the Fairchild 670, though we decided on VCC RC-Tube since the effect was more prominent) and the songs to be as loud but still as dynamic as possible. The rest was up to my personal preferences fitting with the production.

Since I work completely in the box, it was "VST land" for the production from this point forward. I used Melda Production's Dynamic EQ, VoS's Thrillseeker XTC, some M/S tricks with plugins like Brainworx', saturation was mainly handled with Slate's VCC and VTM. For loudness sake, MWaveShaper (or MCompressor I think) as Clipper and ToneBooster's Barricade as Brickwall Limiter setup to ISP mode with -1dBTP as headroom. The now available MP3s were bit-reduced with the trusty Apogee UV22HR, and encoded into MP3 with LAME.

Yes, I am oldschool! *cough*





If you like the production, feel free to leave a comment in here, and also over on the band's pages. I'm sure Chameleon Jazz Connection will love to get some feedback.

And if you have the chance, go to their concerts. Or simply book them.



Thanks for reading.

[zircon]

Very nice! Love the clean, smooth sound. Jazz deserves this kind of clear mastering without too much loudness. Thanks for the explanation too!

[Compyfox]

I wasn't much involved with the editing (mixing) of the whole production. I only gave feedback (read: I did some consulting work) prior to providing me the files to master. So the credit completely goes to Normen Goltz in this case.

I was mainly handling the stereo mixdowns. But I do remember asking for two slightly altered mixes for the set. In one track, the vocals were a bit drown by the music, in another track I kind of wanted a less prominent bass.

I'm surprised that not more people gave feedback so far. I know that there are some jazz/pop fans around here on KVR.



Oh and, since the SoundCloud link is not on FaceBook (yet) and rather hard to acces through the official page:
https://soundcloud.com/chameleon-jazz-connection

(the songs are not cut off here!)



I also got a mail by the band, they always love to get more feedback.
So please do that.

[Compyfox]

The next song is uploaded. This time it's a cover version of The Smashing Pumpkin's "Tonight, Tonight".

You can find it on YouTube now, and eventually on SoundCloud and on the official page as MP3 download within the next hours.




A little bit of info on this one:
This might sound a tad "muddy" compared to the rest of the bunch. But I had to put a lot of effort in there in the bass range to not let it go haywire. Read: ton of dynamic EQing (there's another track where I had to do this).

Of course I could have raised more highend frequencies, but... this would have resulted in quick and tiring ear fatigure. Stereo mixdowns are limited like that.

And to be honest, with the "drowning reverb" on the drumset, the vocals and the slide guitar, it has a feeling of being performed "live" (and recorded at a concert as well, just listen to the proximity effect at the beginning) rather than having a pure studio production.



Even though I'm not a Smashing Pumpkins fan, I like this interpretation (out of the CD, I favour this one the most). It's sounding more sad than the original upbeat-sounding version. In my opinion, this particular interpretation fits the lyrics better.


So please enjoy the next free release.

[Compyfox]

The next song is online on all platforms (SoundCloud, YouTube, official page).
It's "Should I stay or should I go" by The Clash.



Again, this track has a massive (intentional) live feeling. For example, the drums mixed more in the back. And to feature certain instruments in parts of the song, FX usags was massively reduced (again, most noticeable on the drums).

A lot of work was put into the bass (as you can probably hear). Prior to the stereo mastering, I was in the studio of Normen Goltz to take a closer look at what's going on. The bass didn't want to sit right, so we put some hand on it, threw out all initial plugins, ran it through a LA2A emulation, changed the EQ a bit (to let the bass stand out more) and suddenly the Double bass (Contrabass) sat better int he mix. We also changed some vocals settings for a better listening experience.

This (and a cover version by The Cure, which will be eventually released) was the only track where I actually had a closer look prior to mastering.


I do remember that I used a lot of dynamic EQ in the low frequency section. Especially the parts where the notes slide up. Even though the track might(!) sound a bit dull on certain stereo's (I checked several environments myself: TV speakers, living room, kitchen, Senheiser Neckphones, Studio Speakers in uncorrected room/without ARC), I think I still did a good job in taming the double bass/contra bass.

The higher frequency content also caused some trouble. Since the high hats have a focus, I actually rolled off a lot of content there (highshelf, I think around 8k to 10k), prior to running through a tube console, tape and then added the "fairy dust" with parallel EQ again. The end result was a much less ear piercing frequency response.

Though I wish I had the chance to work on the shouts in the background. But... you can only do so much with the stereo mixdowns.



Anyway, like last time - please enjoy the next free release.
Still four tracks to go (AC/DC, The Cure, The Cardigans and Coldplay).

Let's see who guesses the remaining songs that will be released over the course of the next four weeks.

[Sounddigger]

Some nice work there! The sound is generally quite open, dynamic and quite fresh for this kind of group. There are some interesting approaches production wise that result in a well-dosed richness of sound. Performances are nice too. Also, I think we can be confident that during mastering, the dynamic range has been preserved as far as possible this definitely lets the sound be pleasantly accommodating to the ear and lively.

I feel that the Bass on the track "Should I stay or..." is still problematic. This is in the first place due to a problematic mix. The Bass is not focused enough, a touch too loud, and will probably be unpleasantly boomy on a number of systems. You already mentioned it was a tough one. If this version should be an official release, it might be good to look at it again. For this kind of group (Jazzy), it's quite alright to have the bass a bit below what would be typical for rock, pop, EDM, etc. So if you take that as a basis, seeing you have plenty of Bass to spend in this mix, you could be somewhere in between, not being afraid to substantially tame it. On the other hand though, this can also be seen as on the limits of personal taste...

The track "Tonight, Tonight" is the one I like best. Good mood, nice rich sound while keeping a live feel. Pleasant on the ears.

[Compyfox]

Yes, it's a two edged blade: personal taste vs technical correction.

There is only so much possible with stereo mixdowns. Even with M/S techniques. An alternative would have been to mix the track from scratch, or edit it accordingly. Though the client felt that the bass is too low, and on my rig it revealed "it's too strong". It was actually a tad worse before we changed certain aspects in the mix prior to the mastering process. And even more overdone on initial release before I was involved.

Of course a complete remix (edit and mix) would sound different, but the slightly washed out sound was intentional. And for someone that usually doesn't mix, it was an outstanding job.

Then again, this is a promotion release as appetizer for live shows, not an official CD (yet).



I'm happy that you like the cover of The Smashing Pumpkins. I still find it a tad dull (mix technically), probably due to the live feeling and the limited possibilities while mastering. At least if we talk about technical aspects, not performance and general sound.


What do you say to the loudness, btw?
We agreed on shooting for K-12v2 (SLk).

[Sounddigger]

Yes, it's a two edged blade: personal taste vs technical correction.

There is only so much possible with stereo mixdowns. Even with M/S techniques. An alternative would have been to mix the track from scratch, or edit it accordingly. Though the client felt that the bass is too low, and on my rig it revealed "it's too strong". It was actually a tad worse before we changed certain aspects in the mix prior to the mastering process. And even more overdone on initial release before I was involved.

Well, THIS is precisely a case where a good multi-band compressor can do wonders.

I quickly loaded that track into my DAW (Studio One) and tried a few things.

Due to the limited instrumentation it's definitely possible to tame that Bass.

The main problem occurs at and around a frequency of +/- 240 Hz, where Bass and voice are clashing in a brutal way.

The way I would resolve this is as follows (but please note that I only took a quick listen, so these are just a couple of pointers, even though fairly precise):

Actually, much can already be resolved with a surgical frequency reduction at 175 Hz of about 1,75 octave (wide enough to also cover the clash at 240 Hz), of around -4,75 dB.

I found adding another band at 113 Hz of 1 octave, with reduction of -3,75 dB cleared things up even a bit better. That's using the NCL 10-band parametric EQ. With another EQ, you might have to set those dips a bit differently.

Then get a nice multi-band compressor. Put it AFTER the EQ. I tried the free GranComp3 from Buzzcomp and the quite affordable Tri-Comp from Minimalsystem. The Tri-Comp is modeled on the 1176, but then in 3-bands (which doesn't exist in the real world), and seems a bit of a better choice for this track, due to its warmth, although the GranComp3 is wonderfully transparent and will also give great results with EQ tweaked right.

Set the high frequency of the lowest band at around 275 Hz. This can vary from compressor to compressor, depending on the effective frequency slope. Ideally, I would set it at around 185 Hz, but due to the slope 275 Hz worked better. Play around with it. Check SPAN (or other frequency spectrum analyzer).

The low frequency of the highest band could be set at around 4600 Hz.

On the Tri-Comp I set the low band at a ratio of 4:1, with the mid and high band at 2:1. For the GranComp3 I settled at 2:1 for all bands.

An attack of 3ms for all bands works just fine. Tweak to taste. The Tri-Comp doesn't give read-outs for this, so go by the ears (and rough visual indications).

The treshold/input should be set so that the amount of compression is just a squirt (in Swedien's terms ) There should barely be compression for the upper two bands, but the peaks of the lower band should be hit enough to tame the Bass.

So far, this should already sound quite a bit better. Now, to really tame things, simply lower the gain on the lowest band on the multi-compressor to taste (or vice-versa, raise the gain of the upper two bands), so that the Bass gets at a level that is more balanced with the rest. On the GranComp3 I got good results by turning the gain of the lowest band down by (no less than) 6,6 dB.

And voila!!

To round this off, I would suggest that the voice has a bit of sibilence that can hurt the ears in this mix, and it might be useful to try out a squirt of de-essing (which I didn't try), or simply do a dipping EQ curve at around 2,7 to 3 kHz of about 1 octave, around -3,75dB.

All this needs a lot of tweaking to really fine-tune the results, but you should definitely get things a lot tighter this way. It worked here! (If you want me to, I can of course get you the result I got).

Have a ball with it!

What do you say to the loudness, btw?
We agreed on shooting for K-12v2 (SLk).

Didn't check it. I trust you on that one!

[Compyfox]

Hm... I might check that one over at my environment. Though I usually try to stay the hell away from multiband compression. And 6dB gain reduction sure is a bit much IMO. I actually like to put the bass on top (again, personal taste).

Come to think of it, with this mastering job, I only used dynamic EQ, parallel EQ, M/S stereo widening (and narrowing below a certain frequency), then trough a saturator (tube console), a tape machine - and for loudness, 2dB gain reduction max (peaks, full spectrum treatment) with a comp, then a clipper to tame strong rouge peaks and have it easier for the limiter.

Compared to the source files, it's still a drastic improvement IMO.



But I definitely appreciate the class A feedback.

[Sounddigger]

And 6dB gain reduction sure is a bit much IMO. I actually like to put the bass on top (again, personal taste).

Actually, it is a bit too much reduction in this case. As I mentioned, I only took a quick look at this, and the 6dB was a result of initial settings. It does sound good, good enough for Jazz that is, i.e. with Bass not overbearing. But on second glance, -2.3 dB seems to do it just fine (that's on the GranComp3, for the Trip-comp there's no exact read-out).

In any case, even if these seem like massive modifications, the multi-band compression tends to even it all out.

However again, follow your own taste. It's the approach that counts here.

Come to think of it, with this mastering job, I only used dynamic EQ, parallel EQ, M/S stereo widening (and narrowing below a certain frequency), then trough a saturator (tube console), a tape machine - and for loudness, 2dB gain reduction max (peaks, full spectrum treatment) with a comp, then a clipper to tame strong rouge peaks and have it easier for the limiter.

Compared to the source files, it's still a drastic improvement IMO.

Well, it ALL adds up. Good stuff adds up to something good, the other stuff not so much, duh...

But I definitely appreciate the class A feedback.

You're welcome

[Compyfox]

I can check tomorrow soonest. Maybe even with a demo of Tri-Comp (which can be built with modular tools and the right 1176). But I now fear, that certain tracks (that are still in the queue) might suffer the same issues. Even though I invested several hours on the low-frequency realm of each track.

It's still a promo release, now drastically enhanced. One way or another.



Each track sounds totally different mix wise either way. For a complete CD production (if I had the individual tracks), I would have taken a focus on having a better consistence in terms of sound. Granted, way more work (and I don't fool anyone - this is not always possible unless you copy over settings rather than go by feel!), but it can pay off.

It's still a stereo mix editing we're talking about. And not even the most craziest M/S processed can fix everything. Doesn't mean that I'm not open for feedback. So that's greatly appreciated.



I still have my issues with multi-band compressors though. They always did more harm than good for me - and I'm not one of those that slams it on everything during mastering. But in this case (bass), they can indeed be a lifesaver.


I'll perform the test. If it's really drastically better, I might ask the band to do a re-release of that particular track. I then literally crapped out on that behalf.

[Compyfox]

I just did a test with T-Racks Linear Phase EQ (however not with Q's of 1,7 and 1 octaves, but more narrow), and it does indeed clean up the low end frequency range for this particular track.

The compression is then a matter of preference. I tried Tri-Comp and an own setup with a modular environment (only the low frequency was compressed with MCompressor, 4:1, 3ms Att, 150ms release, threshold -15,5dB, Soft Knee 25%). Turns out, that even with a gain reduction of -6dB, it doesn't add much more to the production.

So it's only the EQ.


Then again, it's not an uber-drastic improvement. But it does reduce certain ringing in the lowend areas. And I found one at the same spot in another track. So yeah - again, subtle but there.

Thanks for the find. Still, with ARC active, it's barely noticable.




EDIT:
Some info about the loudness (to the geeks among us):

The album was mastered in K-12v2.

The average loudness hovers between -3LU and 0LU SLk, mezzoforte passages are usually around 0LU SLk, forte fortissimo passages (chorus, breakdowns, etc) at +3LU SLk maximum.

Due to the weighting filter and differences in terms of ballistics, the K-12v2 meter registers tracks at K-12 AZ+4 (sometimes AZ+5) maximum, with an average signal around -13dB to -11dB. The K-12v2 SLk meter goes up until +6 with certain tracks.

The mixes are all within K-System v2 specs. Measured with ToneBoosters EBU Loudness.

[Sounddigger]

I just did a test with T-Racks Linear Phase EQ (however not with Q's of 1,7 and 1 octaves, but more narrow), and it does indeed clean up the low end frequency range for this particular track.

The compression is then a matter of preference. I tried Tri-Comp and an own setup with a modular environment (only the low frequency was compressed with MCompressor, 4:1, 3ms Att, 150ms release, threshold -15,5dB, Soft Knee 25%). Turns out, that even with a gain reduction of -6dB, it doesn't add much more to the production.

So it's only the EQ.


Then again, it's not an uber-drastic improvement. But it does reduce certain ringing in the lowend areas. And I found one at the same spot in another track. So yeah - again, subtle but there.

Thanks for the find. Still, with ARC active, it's barely noticable.

Glad you find the suggestions for EQ correction useful.

Keep in mind that even if you find the original ringing only "minor", this kind of phenomenon can take on very big proportions on some systems. Maybe your monitors are a bit flattering to these frequencies. Something to keep a sharp eye/ear on.

The decision to add a bit of multi-band compression is also a conceptual one, in the end. Although with the settings I suggested it doesn't seem to add much more, that is precisely the idea. I wasn't suggesting to get into the loudness war, after all In fact, even with these very moderate settings, it can make a big difference for certain listening conditions. Especially when listening on small speakers, such as PC speakers. I find it improves the sense of detail, as well as (moderate) punchyness.

On the other hand, if you experience little difference, then maybe you're not driving the compressor hard enough, and maybe you should raise the input level (on the Tri-comp) or lower the treshold some more.

Anyway, there are always so many options. What matters is that you end up with something you like listening to. So if these suggestions have helped at that, then that's a good thing

[Compyfox]

I was just in touch with the band, and they plan to release the whole set as "pack" after they released every track individually. This set will feature the slightly fixed tracks.



I decided to go for the EQ suggestion for the "Clash" cover. It did indeed clean up the lowend, but I tried to stay away from drastic compression on just the low frequency band. I fiddled around for over half an hour and I was like "meh, doesn't work for me". I still use compression in that particular track, but only certain peaks are treated with a max GR of 3dB. And no, I did not go for the Tri-Comp, I used MCompressor with 7ms RMS, 25% Soft Knee and a Threshold of -16dB. No gain compensation or cut. I felt that it didn't work out any further, so I left it as is.

I didn't take the suggestions as any form of futher "loudness war" battles. The K-12v2 mixes are fairly hot already, and my client actually agreed on not going any higher (like constant -6dB RMS)!



BTW:
ARC on my end with my Behringer Truth are sharp as hell and fairly flat. I even had to turn on the high frequency roll off with certain songs. So they are rather revealing than flattening.

I am currently checking the whole set on Laptop speakers (EeePC 900) and Sennheiser Neckphones again. It transfers well. And yes, the "sound" is part of my style. So I'm glad you people like it (as can be seen on YouTube and Facebook especially).




And to get back to actually talking the promo set rather than pure technical stuff:
The next track release has something to do with the "theme" of the album.

[Compyfox]

The next track by Chameleon Jazz Connection is online on their page (MP3) and SoundCloud (stream). The Youtube version will probably online later that evening.

This time it's a cover version of AC/DC's song "Highway to Hell" - with a twist if I may add.



Some (slightly) technical stuff:
This some gave me some headaches (volume wise). The first 2/3rd of the song is fairly low in terms of volume compared the last 1/3rd of this track, and especially the rest of the CD. Automation was essential, else the listener would have thought at the end of the song "WTF is going on?! My ears!". Even though that was somewhat intentional.

This song also features an ultra wide stereo field, which works excellent at the last part of the song (vocals, ad-lips, etc). Else, no rocket science with this particular track.



As usual, please enjoy. Three weekends and three songs to go.