Print

Please login or register.
1 Hour 1 Day 1 Week 1 Month Forever
Login with username, password and session length

[Ged Leitch]

ok ok i know the deadline isn't "officialy" in yet but most of all the entries are, and i think nows a good time for this thread...

Comments about entries can be posted in the other "discussion" thread


Guys,
this is the thread for us to post on how we done our masters and their objectives/ problems/ equipment settings etc...

Anyone wishing to start off please do...

And be as detailed as you want

Feel free to post away

[Ged Leitch]

ok, would you rather I went first guys???

[Ged Leitch]

Right i'll start it off then...

how I done mine.

Signal chain - in order ...


URS Mix EQ

Hi pass @ 25hz gentle slope

- 1.5db @ 80 hz narrow Q

+ 1.5db @ 800 hz medium Q

- 1.9db @ 10Khz narrow Q

+ 0.5db @ 16 khz medium Q
_________________________________

Spitfish De-esser

Detector set around the 6 khz mark for the vox sibilance.
__________________________________________________________

Voxengo Marquis Compressor

 T3 mode - Feedback - soft mode
 15ms attack - 100ms release
 Ratio - 2.00
 Knee set at 30db

 Threshold set to get about 2db max GR.

 Also used the key input filter boosting a pronounced bass resonance from the kick drum around 95 hz.
As well as boosting the 3 khz region to soften some of the vocal parts that sounded harsh.
Both were very narrow band boosts.
____________________________________________________________

Timeworks Mastering compressor/limiter

Used this to get a decent enough level for the mix, I find it seems to have more headroom than the software L2 IMO!
It doesnt clamp down on the signal to fast as my experience using the L2 plugin showed.
____________________________________________________________ _

In conclusion,
I'd say I liked this mix.
I thought what i neded to do was to smooth out the highs circa 10khz and upper mids from the vox.
I also wanted to "gel" the kick drum into the rest of the track which I thought was sitting too much in front.

I tried my best to keep it "natural" and didn't try to take the mix somewhere it didn't want to go.
I basically went for the best balance I could acheive while at the same time giving it some smoothness in areas which it needed.

So, there you go guys,

I'm sure I'll get lampooned for the techniques I used but hey ho!



Cheers,
Ged.

[Pingu]

Thanks for revealing Ged.


I cant be as generous.

If anyone is interested in what i did pm and i will reveal.

I dont expect a pm though.

[bblackwood]

I think all should be required to post how they achieved their results.

Or is this simply a steak-swinging contest?

From now on, it will be required for all WUMPs. I would hope that those who had the opportunity to participate in this one would be appreciative enough to at least post how they did what they did...

[Ged Leitch]

bblackwood wrote on Mon, 15 May 2006 03:51

I think all should be required to post how they achieved their results. Or is this simply a steak-swinging contest? From now on, it will be required for all WUMPs. I would hope that those who had the opportunity to participate in this one would be appreciative enough to at least post how they did what they did...

Cheers Brad,

thought there might have been more posts ...sigh...

[Pingu]

bblackwood wrote on Mon, 15 May 2006 10:51

I think all should be required to post how they achieved their results. Or is this simply a steak-swinging contest? From now on, it will be required for all WUMPs. I would hope that those who had the opportunity to participate in this one would be appreciative enough to at least post how they did what they did...

Im not revealing out of snobism, what do i have to be snobby about.


Ill be laughed out of town if i reveal.



Here goes, dont hate me guys.

First i did DC offset removal.
then i went into

Har-bal, yeah i know use your ears.

then high pass filter at 35 hz

T-Racks compressor (sidechain set at 218 hz, attack 50.5 ms release 504 ms stereo 0.5 ratio 1.71 input drive 5.1).

Then Voxengo Elephant 4X sampling

EL-3
Lim speed = slow
In 2.0
Out 0.0
Stereo linking 90%

[mbruce333]

Well, here is what I came up with.

Crane Song Hedd for my DA/AD loop, processing with Triode at 3, Pentode at 0, Tape at 3.5.

Manley Vari-Mu, about 3db of GR, Med attack and med slow release.

Millenia NSEQ-2, SS mode, +1db@56 shelf, -1.5db@125, -1.5db@16k.

Then through a DBX Quantum for a little M/S EQ.  I don't have the exact settings becaused it crashed before I could write them down (this seems to happen alot.....hmmm).  From what I remember, for the mid EQ, a little cut around 400 or so, and a little boost around 5-6K.  I feel like this is where I should have put some extra attention as I didn't ever really nail the midrange.
I don't think I did much if anything to the Side EQ.

Finally, a little L2 to bring up the level, around 3db of GR on the meters.

I ment to clip the crap out of the HEDD, but I forgot  


This has definatley been a cool thing to do!  I'm just getting into my first room, with gear that until a couple of months ago I've only read about!  So I'm definately learning something new about the system everyday.  At this point, I have a day job, but would like to eventually move into mastering instead.  It'll be a lot of work, but it all pretty exciting.  

Thanks to those who took the time to post some comments, they've been very helpful!  I was hoping to get time to do some listening and post mine, but Mothers day got in the way...maybe tomorrow!

Later!
Mike Bruce

[cerberus]

parallel busses (cubase sx3)
==========================================

m/s bus (sent to src, reverse copy bounced and sent to limiter)
waves lnmb - (de-ess)
waves q-clone sampling eq- (neve suitcase i.r; flat on mid, hp on side)
waves ssl echannel   compression and eq
lnmb low level expansion


c4 crush bus (stereo; sent to src)
q clone (calrec pq 1061 HPF + ampex 351 line amp )
mda delay 18 ms delay with 1% feedback
waves c4 heavy crush


harmonizer bus (flipped polarity, R-L cross-panned, sent to src)
q clone (ampex 351 line amp i.r. + HPF i.r. of undetermined origin)
mda delay 11 ms delay with 1% feedback
waves doubler 4 voice with custom "ged" preset

crosstown bus  (sent to src)
straight through

limiter (sent to src)
=============================================
waves L3

src (in-> 88.2/32 float ; out -> 44.1/32 float, sent to dither)
=============================================
spark xl

dither (in 44.1/32 float)
=============================================
mbit+, no noise shaping.

+++++++++++++++++++++++

the settings i used.. do you really want to know?  it's an additive system, some processes do partially null each other's effects such that adding their signal to the chain can reduce the overall peak level compared to rms...that would be the best case for all the parallel busses....but also they perform other tasks, particularly with focusing the stereo image. also the q-clones with "flat" i.r.s seem to work as allpass filters, i believe they help affect phase rotation, but also i am using reverse panning, reverse polarity, and delays to affect that too.

one important thing i measured and fixed was a hole around 4k, i filled it with boost from the waves ssl e channel eq...but of course the multibands are working a bit harder bring down the overall brightness.

how does this really work?  imagine you have 3 color faders, red, green and blue.. you could use them to adjust color balance, but all 3 together also could be used to change the brightness.   this, in a nutshell is the reason i set up such a wacky chain.. in the end, i have these faders that allow me to control this and that.. and the result of all this complexity can also emulate non-linear behavior, something computers are not especially capable of without some prodding.  

and then the real result comes down to aesthetics, not the gear.

+++++++++++++++++++++++

this took me half the time as my best work has taken me because i rushed to match chrisj's suggested 5 hour time limit.. and in doing so i made several mistakes which i will try never to repeat:

1. i did not take the time to select appropriate reference tracks. instead i imagined some elvis costello and robyn hitchcock in my head....too bad i didn't think of green day, but at least i didn't think of belle and sebastian..

2. i did not  work through the limiter..(to save time and cpu)... i prefer to set the limiter first and then tweak it once again after everything else is set. the l3 has a lot of parameters to check... so i cut this corner, only set it up once.. by then it is too late to think about making it intentionally loud, i thought i could get away with 1980's rms levels... i guess not. i will change the way i do limiting from now on,

3.  there should be at least another parallel bus just to help make the middle freqs seem  punchy and to bring them forward...i only had time to make the vocal sound euphonic, i short changed it....not enough time...normally i use at least 3 parallel busses with processing, here only 2.


jeff dinces

[Viitalahde]

And here's my set-up, way way more than I usually do..

1) Voxengo MSED, reduced side gain by -1dB to have more balls in the middle

2) Gained L/R +1,5dB on the way out to ride the analog loop more

3) DIY parametric EQ (based on Sontec MEP-250A)

..65Hz -1dB, bell, Q=0,7
..350Hz -0.5dB, bell, Q=0,4
..1khz +1dB, bell, Q=0,4 (this must be the honkiness..  )
..11,2khz, -0,5dB, bell, Q=0,4

4) DIY passive EQ with Neve BA283 output stage

..100Hz, +2dB, shelf (notice the 65Hz dip in the previous stage that shapes this!)
..5k, +1dB, broad (fixed) bell

5) Gyraf G10 vari-mu (total ~1,5dB)

..Input stage gain 8dB
..Full output
..Threshold two-o-clock
..Ratio eight-o-clock
..Attack two-o-clock (I have slower attacks than the stock version)
..Release ten-o-clock

6) Back in the DAW

7) PLparEQ (Refine Audiometrics)

..400Hz, -1,2dB, bell, Q=0,47, only in the side channel

8] Voxengo Elephant

..Input +4,6dB
..Output -0,2dB
..Mode=CLIP, linking off, 4x oversampling
..DC filter @ 16Hz, bessel type

9) Dither=POW-R1 => 16bit.

[Luke Fellingham]

If Brad requires it........

Sony Oxford eq:
High pass @ 28Hz
Shelf @ 150 Hz +0.6
400Hz -.9
7.1KHz -1.9
13KHz -1.4

Spitfish De-esser:
Set to about 7KHz in soft mode

Avalon 747, compressor in tube mode
Low ratio giving between .5 and 1 dB of gain reduction
Slight boost with 5KHz shelf

Analogue eq
48Hz wide bell small +1.5
180Hz narrow -1
2.1KHz wide +1
14 KHz side +1

Back into computer
UA Precision Limiter:
2dB gain reduction

Sony Oxford Inflator:
Effect 100%, Curve 0

Pow-r2 dither

[Ged Leitch]

Excellent guys, amazing how differently folk approached this.




keep em coming!

[F. Davis]

Mines a bit wierd now when I look at it but I think I got a result.

Parallell Compression @ -18dB
  I squashed the parallell by Eqing first , cut 430Hz -3.9dB Q1, 7924Hz -3dB Q6.5, 4048Hz -2.5dB Q6.5 , 5512Hz -2.6dB Q6.5, 12489Hz -4.1 Q6.5
  Then I Sqeezed the living daylights out of it with an L3 with a long release time and a flat curve.

Then with the parallel mixed in  I did the following

Waves Renaissance EQ ,60Hz-1.5dB Q1.2,279Hz-2dB Q6.5, 859Hz-3.9dB Q6.5,4kHz -3.9dB Q6.5

I used three Waves Linear MBs to do the following.(this is hard to describe)

Between 2.5kHz and 4kHz about -2 db really fast compression at 19:1 (Mad!)
About 6db compression below 90Hz mainly on Kik peaks.
About 2 dB compression between 520Hz and 1kHz (fairly Fast attack and release)
As much as 5 dB compression between 250 and 600 Hz with a medium attack and release time (I know this overlaps the last one but so what)
Then as much as 4 or 5 dB limiting on everything above 8kHz (Oooooh)

Then I used a Waves de-esser at 9.2kHz using the narrow band

Then anothe Renaissance EQ
80 Hz -2.9dB Q6.5, 324Hz -3dB Q6.5, Right Channel 630Hz -5db Q3.46, Left Channel 841Hz -6 dB Q5.27, 6974Hz -4.7dB Q6.5, 11561Hz -2.6dB Q6.5

Then I used a Waves Linear Phase EQ to tackle the Cymbals mostly

5254Hz -4.7dB Q6.5, 6632Hz -2.3 Q6.5, 9130Hz -2.5dB Q6.5 and 13867Hz -2.4dB Q 4.56

From there I went out through my bog standard digidesign 192 I/O to a Massive Passive where I boosted about 4 dB at 1.5 kHz (wide Band) and set my High pass at 39Hz.
Then into a Vari-Mu
Thresh Max, Attack Slowest setting, Medium Release.
I turned up the I/P until the gain reduction meter was barely moving then turned up the outputs until my 192 converters were really distorting and backed off a bit.

(I'm sure it's still distorted at this point even though I backed off but I don't mind cause I think it suits the program material)

From there , to add insult to injury, I inserted a hardware L2 with about 1.7 dB limiting and automatic release. I also used the IDr dither from that Type one with normal noise shaping.

I then bounced down using Protools.

P.S.
The entire deal took just 45 mins. I don't think it's healthy to spend much longer than this on any one track.

[nmw]

after a quick once through to decide what the mix seemed to be asking for and needed....

dealt with the imaging first using waves s1 (id often use my MS unit with lundahls for this however i wanted to use alogorithmix red instead of my outboard EQs for this one so this meant a switch round). closed down to about 0.85 with this.

using algorithimix EQ red (this is sadly from memory as the EQ reset when i reopened project this morning and had defaulted to flat), boosted bass at about 80hz by approx 3db with a Q of about 0.8 to add some body to the kick and also the bass itself, for the highs used a rather extreme setting with a high shelf of -12db cut with a Q of 1.35 at about 15k. this seemed to dull the harsh portion of the highs but work the actual body of the hats well. i then used a bell at 250hz with about 1.5db boost and a Q of about 2.3 to try to add something here to even out the balance a little. finally applied a bell at 4k with a boost of 1.5db and a Q of 2db to try to just bring out the vocal and snare top end a touch to balance out the sound some more.

i then put the signal through my Thermionic Phoenix vari-mu with  a lot of input gain to warm it up a bit, medium attack and release ( i found the kick to be especially sensitive to attack settings with it losing or going boxy very easily) threshold was set to give it about max 3db of gain reduction in places. my attempt here was to leave it as dynamic as possible as it already felt somewhat badly compressed in sections but to try to gel it altogether and add some dimension and depth to the sound.

then used the ouput of the comp to drive my lavrys into clipping to get some volume onto the signal.

once inside sequoia again i felt that there could be a benefit to a touch more shaping now things had been smoothed out and i applied algorithmix red again with two boosts. one at 2k with a Q of 1.5 and gain of 1.8 and the second at 250hz with a Q of 3 and gain of 1.1

decided there would be room for some more volume and boosted the sequoia track volume by approx 2db (and clipping this). it seemed that any use of L2 dulled the sound badly and this worked far better. there was room for about another 2.5db on there before it turned to crap using this method but i didnt feel there was any need with this to drive it to the limits, there had also been muchmore give on the lavry input too but again same thoughts applied.

dithered using TPFD to 16bit.

im sure on a listen today ill hear much more to do to fine tune but given i only came to this very late on sunday night and did the whole deal as one pass in 45mins including down and uploads without time for a break and reassesment it turned out ok.

[OldSonic]

OK, here Russians go  

  Pyramix internal EQ
 82Hz, Q11, -3.8dB
 125Hz, Q2.8 +2.4dB
 210Hz, Q2.8 +2.9 dB
All this was just attempt to change kick-bass balance.
Then Apogee DA1000, and go to the analog world...

   Focusrite Blue 315EQ
LoShelf 33Hz +0.6 dB
LoMid    560Hz, bw1, -1.5dB
HiMid    1.8kHz, bw0.8, +1.5dB
HiShelf 22kHz, +1.5dB

    Manley VaryMu
Minimum treshold, max. release, attack set to 12 o'clock, unlinked. Gain reduction 1.5, up to 2dB

    Maselec MLP
WideBand limiter section ON, zero gain reduction flashing sometimes. HF limiter section ON, MS-mode, hf gain reduction up to 4dB on some peaks.Prism AD-1, go back to digital (no any clipping on AD).

     HEDD
Pentode - around 3, Tape - around 3.5 - 4.

     Hardware L2
ARC - ON, Link - OFF, GR - 3 dB, output set to -0.3dB.
Back to Pyramix, 24 bits. Fade-in fade-out in Pmx, and last step Pyramix internal dither to 16-bit. That's all, takes about 40 minutes.


     


 

[Ged Leitch]

amazing diffrences in set ups/ approaches there...

Jeff, thats a WAY too complex chain for me to even begin to toy with, but the result sounded great!





more!



i want more!

[Bob Boyd]

Ged Leitch wrote on Mon, 15 May 2006 13:10

Well Bob, i'm glad you managed to get the mix downloaded cause this sounds great! Great top end Kick a bit too pronounced overall great stuff!

Thanks Ged,

My main goal was to warm this mix up and try to make the cymbals sound not so "drum machine like".  Interesting you should mention the kick.  I ran a version of this before the one I posted and went back and added a touch of the kick back in.  Althought the mix has some obvious issues, I felt like the more reserved version lost a bit of the original intent.

For anyone interested, here's a quick breakdown on my signal path:

- Playback - ProTools HD|Accel
- Lavry Blue D/A
- Crane Song HEDD 192 - a touch of Pentode and Tape
- Weiss EQ1-LP running in Liner Phase and M/S mode to address the panned cymbals (take out brightness), panned guitars (add mid definition), and vocals seperately (add some mid/lower mids).
- Maselec MEA-2 - took out some annoying kick and added 6dB at 48Hz to bring some extension to the bass guitar.
- Massive Passive - took out some 3.3k in the vox.
- API550m - engaged, no EQ (transformers)
- Chandler LTD-2 - light compression so I had a touch of dynamics that worked as a whole, + 3dB of gain (the LTD-2 are warmer as you gain up the output stage)
- Lavry Blue A/D - no clip, no saturation engaged
- Weiss DS1-MkII - dessing around 9.4k, Q - a little over an octave wide, 10.0:1, no knee
- System 6000 - multiband dynamics to control the kick issue and some vocal annoyance, touch of de-essing here (some digital clipping occured here as well)
– Waves Trans X Multi - to make the snare transients snap more evenly with where I left the kick
- (if I was going to apply Weiss SRC, it would occur here with a input signal below full scale but the mix came in at 44.1k)
- limiting to -.3dBFS (on pop/rock stuff, my target is usually -9 to -8 RMS)
- pow-r 2

Between the Maselec low end gain and then the low end compression with the 6000, I decided to try to round out/extend the bass guitar and then control most of the kick dynamically.  I don't always use multiband but I felt that the kick problem was not an EQ issue as much as a dynamic one.

It ended up being quite a path but it's not uncommon for me to string something like this together if I think the song needs it.

All in all, a fun exercise!

[Ged Leitch]

Bob Boyd wrote on Mon, 15 May 2006 20:27

Ged Leitch wrote on Mon, 15 May 2006 13:10 Well Bob, i'm glad you managed to get the mix downloaded cause this sounds great! Great top end Kick a bit too pronounced overall great stuff! Thanks Ged, My main goal was to warm this mix up and try to make the cymbals sound not so "drum machine like".  Interesting you should mention the kick.  I ran a version of this before the one I posted and went back and added a touch of the kick back in.  Althought the mix has some obvious issues, I felt like the more reserved version lost a bit of the original intent. For anyone interested, here's a quick breakdown on my signal path: - Playback - ProTools HD|Accel - Lavry Blue D/A - Crane Song HEDD 192 - a touch of Pentode and Tape - Weiss EQ1-LP running in Liner Phase and M/S mode to address the panned cymbals (take out brightness), panned guitars (add mid definition), and vocals seperately (add some mid/lower mids). - Maselec MEA-2 - took out some annoying kick and added 6dB at 48Hz to bring some extension to the bass guitar. - Massive Passive - took out some 3.3k in the vox. - API550m - engaged, no EQ (transformers) - Chandler LTD-2 - light compression so I had a touch of dynamics that worked as a whole, + 3dB of gain (the LTD-2 are warmer as you gain up the output stage) - Lavry Blue A/D - no clip, no saturation engaged - Weiss DS1-MkII - dessing around 9.4k, Q - a little over an octave wide, 10.0:1, no knee - System 6000 - multiband dynamics to control the kick issue and some vocal annoyance, touch of de-essing here (some digital clipping occured here as well) – Waves Trans X Multi - to make the snare transients snap more evenly with where I left the kick - (if I was going to apply Weiss SRC, it would occur here with a input signal below full scale but the mix came in at 44.1k) - limiting to -.3dBFS (on pop/rock stuff, my target is usually -9 to -8 RMS) - pow-r 2 Between the Maselec low end gain and then the low end compression with the 6000, I decided to try to round out/extend the bass guitar and then control most of the kick dynamically.  I don't always use multiband but I felt that the kick problem was not an EQ issue as much as a dynamic one. It ended up being quite a path but it's not uncommon for me to string something like this together if I think the song needs it. All in all, a fun exercise!

PHEW!
thanks Bob...

That signal chain costs more than I earn in a 3 year period!!!

Glad you gave me the chance to hear the TC6000 in action

[bleen]

I lost my original session, so I recreated it this morning using a couple of different approaches (still don't like the sound of it - I need to address the acoustics in my room more than anything..).

Mine was done ITB (and it sounds like it....) using Waveburner with the following chain:

Waves DeEsser (to take a bit off the vocals and cymbals) -> DUY Valve (for some warmth) - > Logic Linear EQ -> UAD-1 LA2A -> DUY Width -> UAD-1 Precision Limiter

For your viewing pleasure:

[Pingu]

Good one Bleen.


Great way to sum it up too.

[Pingu]

Damn amazing to see the different approaches.



Do i win a prize for the most pathetic path.

[Ged Leitch]

Pingu wrote on Mon, 15 May 2006 20:57

Damn amazing to see the different approaches. Do i win a prize for the most pathetic path.

heh heh

i know what you mean Pingu

i'm feeling the analog envy all over!

[Bob Boyd]

Ged Leitch wrote on Mon, 15 May 2006 14:44

That signal chain costs more than I earn in a 3 year period!!! Glad you gave me the chance to hear the TC6000 in action

 

For whatever it's worth in 1993, I had a Mac, Sound Designer 2, Waves Q10, and $350 Tannoys.

[Ged Leitch]

Bob Boyd wrote on Mon, 15 May 2006 21:27

Ged Leitch wrote on Mon, 15 May 2006 14:44 That signal chain costs more than I earn in a 3 year period!!! Glad you gave me the chance to hear the TC6000 in action   For whatever it's worth in 1993, I had a Mac, Sound Designer 2, Waves Q10, and $350 Tannoys.

Bob, what do you think of my "techniques post?

Am i doing anything strange?

Have you ever used the SC input filter of a comp for freq dependant compression or do you usually default to  the TC600 for that if needed???

Cheers.

[Ged Leitch]

calling Steve berson!

could you at all post your equipment and techniques spec for the track in the WUMP?

be interesting to hear how you approached this Steve, and thanks for participating!

[Bob Boyd]

Ged Leitch wrote on Mon, 15 May 2006 15:41

Bob, what do you think of my "techniques post? Am i doing anything strange? Have you ever used the SC input filter of a comp for freq dependant compression or do you usually default to  the TC600 for that if needed??? Cheers.

Looks fine to me.  Based on your EQ notes, we were going after similar things.  

I don't play with sidechains that much.  If I'm trying to fix an area dynamically, I'll usually grab the 6000 or if it's one particular band, I will reach for the DS1.  If I'm not using it for broadband compression, it's all about the sidechain.  You can have it focus anywhere in the spectrum.

On another note, I love the PSP MasterComp on some things.  If I need something light, I can dial that in quick and if I just "kiss" it, sometimes it sounds really good! (Note to the software-only guys!)  I have gotten good results adjusting it's sidechain from time to time as well.

[chrisj]

Looks like I lost my whole post so let's see how much time I have to recap.

Input 'reconstructed' into 64 bit float by smoothing off dither noise (HF info at around 1 LSB)

Treble boost from hell that extends down to around 5K, too much

Slight cut below 130 hz, boost upper midrange, pull back extreme highs again- every effect done with attenuation-boost combined with saturation/anti-saturation and slight time shift (a few ms max) going along with the boost/cut

Slight compression, nothing very unusual about it

Every gain/EQ operation is followed by a slew limiter (aka de-ess) but it's set so minimally that it's barely noticable, except by comparison with what you would've got without it

Peak limiting uses saturation as make-up gain to simulate peaks of the same RMS/density that it would've been if there wasn't any limiting

Oh yeah, and the homebrew dither.

I'm sorry that I can't be more informative- I have no time left and the previous post covered more technicalities. I'll answer any questions if there are any. Next time my entry will inspire more in the way of 'how did you do that?'. For now- cheers

[Ged Leitch]

Bob Boyd wrote on Mon, 15 May 2006 21:56

Ged Leitch wrote on Mon, 15 May 2006 15:41 Bob, what do you think of my "techniques post? Am i doing anything strange? Have you ever used the SC input filter of a comp for freq dependant compression or do you usually default to  the TC600 for that if needed??? Cheers. Looks fine to me.  Based on your EQ notes, we were going after similar things.  I admit I don't play with sidechains that much.  If I'm trying to fix an area dynamically, I'll usually grab the 6000 or if it's one particular band, I will reach for the DS1.  If I'm not using it for broadband compression, it's all about the sidechain.  You can have it focus anywhere in the spectrum. On another note, I love the PSP MasterComp on some things.  If I need something light, I can dial that in quick and if I just "kiss" it, sometimes it sounds really good! (Note to the software-only guys!)  I have gotten good results adjusting it's sidechain from time to time as well.

Sir Bob,

Do oyu mean turning the sidechain "Boost" on or "cut"?

I also have had good results boost the hi or low shelf in the SC of Mastercomp.

[MT Groove]

I'm posting from home so I don't have my settings in front of me.  But here's what I did:

TC Electronic MD3 EQ
- Cut at low mid freq.(This was the only EQ processing I applied)

TC Electronic MD3 Multiband compressor
- Xover @ 100Hz and 4kHz (if I can remember)
- There was compression on all 3 bands with more on the high band for deessing and smoothing of the harsh cymbals.
- Slight manual make-up gain with varying amounts for each band. (I sometimes like this better than broad EQ boosts/cuts)
- Expander and limiter bypassed

Sony Oxford Inflator
- Effect @ 100
- Curve @ 25 (this created the intentional distortion Cerberus noticed)  
- Band Split Mode
The Inflator process give it some loudness and some warmth.

TC Brickwall Limiter
- Upsampling Mode
- Auto Release
- Stereo Linked
This is for catching all the overs which the Inflator does produce while in Band Split Mode.

Waves IDR Dither
- 16Bit

If anyone wants more specific settings I'll post them when I'm back inthe studio.  But I doubt anyone would be interested.  

Reading all these setting and techniques from everyone is such a learning experience.  This is great stuff!!

[Ed Littman]

Bob Boyd wrote on Mon, 15 May 2006 15:27

- Playback - ProTools HD|Accel - Lavry Blue D/A - Crane Song HEDD 192 - a touch of Pentode and Tape - Weiss EQ1-LP running in Liner Phase and M/S mode to address the panned cymbals (take out brightness), panned guitars (add mid definition), and vocals seperately (add some mid/lower mids). - Maselec MEA-2 - took out some annoying kick and added 6dB at 48Hz to bring some extension to the bass guitar. - Massive Passive - took out some 3.3k in the vox. - API550m - engaged, no EQ (transformers) - Chandler LTD-2 - light compression so I had a touch of dynamics that worked as a whole, + 3dB of gain (the LTD-2 are warmer as you gain up the output stage) - Lavry Blue A/D - no clip, no saturation engaged - Weiss DS1-MkII - dessing around 9.4k, Q - a little over an octave wide, 10.0:1, no knee - System 6000 - multiband dynamics to control the kick issue and some vocal annoyance, touch of de-essing here (some digital clipping occured here as well) ? Waves Trans X Multi - to make the snare transients snap more evenly with where I left the kick - (if I was going to apply Weiss SRC, it would occur here with a input signal below full scale but the mix came in at 44.1k) - limiting to -.3dBFS (on pop/rock stuff, my target is usually -9 to -8 RMS) - pow-r 2

here's my chain......

Wavelab 5
mytek stereo 96 DAC
into dangerous S&M box Sum(mid), Minus(side) width set to +6 (first time I went that far)

the Sum,
went into the left side of the IBIS with the eq settings of
+1db, 52hz tight q,
+2db, 554hz wide q
+3db, 148k wide q
+2db 19k wide q
I used the chandler ltd-2 lightly with fast attack & release & a bit of makeup gain.

the Minus, went to the right side of the IBIS with the color source at 5 mainly for the guitars(thats alot)
+3db, 220hz wide q
+4 db 1769hz mid q
+6 db 19k wide q (i know, to bright, but cought some problems at the end with the desser)
I used the right side of the stc-8 for a bit of makeup gain (between the 2 compressors outputs i can vary the mix to taste)

ADC- HEDD  no tape or tube effects
trans x wide for the extra snap in the snare & kick (Bob, great minds think alike...)
Algorithmix red eq
+11.9 db, 112hz tight q ...very transparent eq's that can be pushed,got a little bass tone from that boost
+1.2 db, 2.33 hz wide q
Waves de esser @ 6979k
+3db of master fader clipping
inernal wavelab dither

Ed

[cerberus]

Pingu wrote on Mon, 15 May 2006 15:57

Do i win a prize for the most pathetic path.

fwiw, i know a very respected mix engineer with a current billboard top 10 single who responded to a disparaging comment i made last week about t-racks:
"the soft clipper is killer, i love that plug"

i did not ask him about har-bal.
not just a path, also a technique...an entire school of thought...

jeff dinces

[aivoryuk]

Okay here is my signal chain and my work ethic on the track.

First things first i listened to the track as my impression was that the all the bass energy was in the kick drum which was too upfront for my liking and was at the obscuring the bass guitar slightly.
the high hat was also slightly forward but at the same time did drive the track along.
guitars seemed a bit muddy
the vocals seemed fine to me.

one of the things I felt was impressive about it the track was the performance. It was pretty tight IMO
so although the piece had a few flaws techniqually I felt the performance of it was the most important thing to concentrate on.
I did do take where it was slightly more processed in terms of Eq but when comparing to the original I was totally happy.

so I did the version I uploaded and here are my settings In order.
First I did a manual gain change of -0.25dB on the point of where the vocal says 'Patronise me' in the first line. It seem to send the chain in to overdrive. You can notice on most of the masters
Waves Lowband Eq
86Hz -2 narrow q
43Hz +1.4 narrow q

URS Fultec (I know this is primarliy a mixing eq but i do like this for minor cuts/boots)
300Hz -1.5 narrow bandwidth
1Khz +1 fairly Narrow bandwidth

Dynamically I felt the track was good so i went out digitally into my Tc Electronics triple C in Fullband mode and used
ratio 1:12-1 (the lowest ration
attack 70 ms
release 500ms
Just to glue it to glue to make it more solid.

I then added level 3 DRG from the triple C which is my version of clipping and digitally out again recorded realtime in 16 bit
There is no peak limiting on my version

Overall I felt it came out ok as I had preserved the performance.
For the sake of a dB or 2 extra mid or whatever i felt it wasn't worth it


sorry for the bore and thanks for all the comments and thanks to ged for arranging this i really enjoyed it.

Ps Pingu: don't be soo hard on your chain i really enjoyed yours it was one of my faves

[PieterS]

My approach was "in the box" using my own software along with RML software in the SAWStudio environment.

The setup:

1)
Sonoris Equalizer in upsampling mode
* high pass @ 16 Hz 12 dB/oct slope
* +2.25 dB @ 47 Hz BW=1.3 oct
* -1.5 dB @ 94 Hz BW=2.0 oct
* -2.75 dB @ 840 Hz BW=1.6 oct
* +2.0 dB @ 2.3 KHz BW=2.1 oct
* -2.75 dB @ 16.3 KHz BW=2.5 oct

2)
Sonoris Compressor in opto mode
* ratio = 1.2
* threshold = -25.5 dB
* attack = 100 msec
* release = 200 msec
* knee1 = 20%
* make-up gain = 2.5 dB
* sidechain high pass filter @ 100 Hz

3)
RML Levelizer (brickwall limiter)
* gain = 11.1 dB
* ceiling = -0.26 dB

4)
Sonoris Dither
* TPDF type noise
* noiseshaping type 2

[Pingu]

aivoryuk wrote on Wed, 17 May 2006 01:03

Ps Pingu: don't be soo hard on your chain i really enjoyed yours it was one of my faves

Cheers man.


As yours mine.

[OTR-jkl]

My entry sounds like crap. I knew better, but I did mine late at night (1am) after a long day AND at low volume (that was the killer)......and it shows....... Excuses I know, but it was then or never.

Analysis:
To my ear, this track is quite thin, highs are harsh and the mids are honky.

Goal:
Smooth the top but still leave some room up there, get rid of the boxiness in the mids, beef up the bottom and (obviously) boost the level - w/o spanking it. I wanted it tight yet still bouncy.

Chain:
Playback
- Samplitude (1st instance)
- Original file (44.1/16)

Process (within Samp1)
- PSP MasterQ (FAT On)
  HPF; 23Hz, Knee 0.67
  Lo shelf; -0.37dB @ 86Hz, Q=0.50
  +0.25dB @ 86Hz, Q=3.54
  -1.57 @ 167Hz, Q=1.54
  +0.35 @ 217Hz, Q=2.67
  -3.57 @ 489Hz, Q=3.26
  -0.95 @ 6.57kHz, Q=2.00
  -1.05 @ 9.5kHz, Q=1.52

- PSP MasterComp (mode: Soft, RMS)
  Ratio=1.4
  Thresh=-2.62
  Attack=657ms
  Rel=262ms
  Link=100%

D/A-A/D
ART DI/O loaded with a NOS Telefunken 12AX7; tube saturation at about 95%; not quite +4dB analog gain

Capture
- Samplitude (2nd instance)

Limit (within Samp2)
- PSP Vintage Warmer (mode: Single band)
  Drive=+5.30
  Knee=2.0
  Speed=55.0
  Rel=0.28
  Mix=100%
 
Dither
- Samp TPDF; bit depth set at 2.0


Conclusion:
Missed the mark. Bottom is still thin and the kick is not defined well. Top is a little too edgy. Mids not too bad but instead of boxy are now maybe a little canny. Level not too bad, could be slightly louder. Punch & bounce decent. Overall, not thick enough - especially in the upper bass/low mid region.

[Pingu]

OTR-jkl wrote on Wed, 17 May 2006 03:12

My entry sounds like crap. I knew better, but I did mine late at night (1am) after a long day AND at low volume (that was the killer)......and it shows....... Excuses I know, but it was then or never. Analysis: To my ear, this track is quite thin, highs are harsh and the mids are honky. Goal: Smooth the top but still leave some room up there, get rid of the boxiness in the mids, beef up the bottom and (obviously) boost the level - w/o spanking it. I wanted it tight yet still bouncy. Chain: Playback - Samplitude (1st instance) - Original file (44.1/16) Process (within Samp1) - PSP MasterQ (FAT On)   HPF; 23Hz, Knee 0.67   Lo shelf; -0.37dB @ 86Hz, Q=0.50   +0.25dB @ 86Hz, Q=3.54   -1.57 @ 167Hz, Q=1.54   +0.35 @ 217Hz, Q=2.67   -3.57 @ 489Hz, Q=3.26   -0.95 @ 6.57kHz, Q=2.00   -1.05 @ 9.5kHz, Q=1.52 - PSP MasterComp (mode: Soft, RMS)   Ratio=1.4   Thresh=-2.62   Attack=657ms   Rel=262ms   Link=100% D/A-A/D ART DI/O loaded with a NOS Telefunken 12AX7; tube saturation at about 95%; not quite +4dB analog gain Capture - Samplitude (2nd instance) Limit (within Samp2) - PSP Vintage Warmer (mode: Single band)   Drive=+5.30   Knee=2.0   Speed=55.0   Rel=0.28   Mix=100%   Dither - Samp TPDF; bit depth set at 2.0 Conclusion: Missed the mark. Bottom is still thin and the kick is not defined well. Top is a little too edgy. Mids not too bad but instead of boxy are now maybe a little canny. Level not too bad, could be slightly louder. Punch & bounce decent. Overall, not thick enough - especially in the upper bass/low mid region.

Wow thats intersting.

Thank you for posting.

FWIW i likes yours.

2 things i would appreciate a little elaboration on if you could.

1. Your attack time on the mastercomp.
Do you commonly set late attack times?
It might not be late for some but to me this seems late.

2. Could you please explain how exactly you are using two instance of Samp, and why.

Im presuming its cause you dont trust the bouncing 100 percent.

Also what driver are you using in order to achieve this.

I tried a while ago to do this with the 002 rack but couldn't figure it out.

Cheers.

[OTR-jkl]

Pingu wrote on Tue, 16 May 2006 14:23

FWIW i likes yours.

Thanks for the vote. I don't like it because I know I can do much better...

Quote:

1. Your attack time on the mastercomp. Do you commonly set late attack times? It might not be late for some but to me this seems late.

It varies. I try to keep the original transients intact as much as possible; only clamping down on them if needed. Since its still too bright & edgy, I probably should've used a faster attack. Remember though, I did this at low volume so I wasn't clearly hearing what was going on...

Quote:

2. Could you please explain how exactly you are using two instance of Samp, and why.

I'm using the 1st instance for playback and to add any VST plugs such as EQ & Comp. I route the output of Samp1 to Dig output - thru any outboard processing - back to Dig input. Samp2 (in Live Input mode) is used to add s/w limiting and capture the signal to the HD at the current Fs and at 32bit. Capturing is done by utiizing the MixInFile function. I then build the Master EDL from the captured 32bit files. Dither is added when Samp creates the image file from the Master EDL.

The Dig I/O card is set to Digital Input (external) clock source and the D/A-A/D acts as the master clock.

This method really is no different than if I were to use Wavelab or some other app for playback and Samp to capture. The main reasons I work this way are so that I can introduce h/w into the chain + it allows me to hear everything working together. There is a small amount of latency involved but it has no bearing on anything other than a slight delay in monitoring processor adjustments.

BTW, I totally trust Samp's bounces. I did the null check when I first started using it...

Quote:

Also what driver are you using in order to achieve this.

WDM drivers. I've tried w/ASIO & MME drivers but have never been able to make it work.

Did I explain well enough...?

[Pingu]

Yep.

Thanks heaps for that.


Probably why i cant get it to work is cause im using ASIO which i think im stuck with cause of the 002.



Thanks for the explanation.

[Bob Boyd]

Ed Littman wrote on Tue, 16 May 2006 01:27

here's my chain...... Wavelab 5 mytek stereo 96 DAC into dangerous S&M box Sum(mid), Minus(side) width set to +6 (first time I went that far) the Sum, went into the left side of the IBIS with the eq settings of +1db, 52hz tight q, +2db, 554hz wide q +3db, 148k wide q +2db 19k wide q I used the chandler ltd-2 lightly with fast attack & release & a bit of makeup gain. the Minus, went to the right side of the IBIS with the color source at 5 mainly for the guitars(thats alot) +3db, 220hz wide q +4 db 1769hz mid q +6 db 19k wide q (i know, to bright, but cought some problems at the end with the desser) I used the right side of the stc-8 for a bit of makeup gain (between the 2 compressors outputs i can vary the mix to taste) ADC- HEDD  no tape or tube effects trans x wide for the extra snap in the snare & kick (Bob, great minds think alike...) Algorithmix red eq +11.9 db, 112hz tight q ...very transparent eq's that can be pushed,got a little bass tone from that boost +1.2 db, 2.33 hz wide q Waves de esser @ 6979k +3db of master fader clipping inernal wavelab dither Ed

Really enjoyed reading through that Ed.  Good stuff!

[cerberus]

i have begun to put some of the data on a chart.  i think there are trends and commonalities we can spot that we all might learn from... but it will take me a while to sort through all this data, especially to unwind feargal's parallel processing.

does anyone who did all in-line processing think the parallel stuff is weird? ineffective? or just beyond the point of diminshing returns?

jeff dinces