Upward Compression Excercise

[Cary B. Cornett]

I was less than happy with the bass sound on one of my mixes, and a comment I read in an interview regarding applying "mastering" processing to individual tracks before mixing tempted me to try an experiment. I have decided to share the results with this group for your reactions. If anyone thinks the difference was worthwhile and would like to know details of the process used...

So here is a segment of the mix with the "old" and "new" bass, and the soloed tracks, all as processed in SawStudioLite, for any comments.

Mix A
Mix B
Bass A
Bass B

[Dave Labrecque]

Cary,

Noticably fuller and deeper in B. OK, what's the secret?

BTW, is it a Rickenbacker? That's what I have; not an easy bass to get a good sound out of, in my experience. I don't know how McCartney, Geddy Lee, Chris Squire, et al. did it.

[DannyV]

Yes! .....what Dave said. I noticed a little more high end
sizzle in the "a" Bass track. Wellllll What did you do?
Later,

Danny

[Pedro Itriago]

BTW, is it a Rickenbacker? That's what I have; not an easy bass to get a good sound out of, in my experience. I don't know how McCartney, Geddy Lee, Chris Squire, et al. did it.

At the risk of hijacking the thread (as if I ever...), could you expand on why you find so difficult to get the rick sound to your liking?

For Chris I remember an interview with Bill Bruford. When asked why he had his particular snare sound, he replied that he was forced to tune it in a higher pitch to be more noticeable in the mix because Chris was fonf to play his rick in the upper side of the registry.

As for Geddy, I think I read somewhere that he got tired of the ringing of the strings with the frets. That's why he travelled to Steinberger, then Wahl and now Fender J again.

[Pedro Itriago]

Sounds like a rick and that they went too "happy" with the bridge pick-up in both amplitude and frequency.

What did you do, multiband (2 bands) compressed it for b?

[Mark Stebbeds]

As for Geddy, I think I read somewhere that he got tired of the ringing of the strings with the frets. That's why he travelled to Steinberger, then Wahl and now Fender J again.

He's still searching for a bass sound that works with his annoying voice.

Mark

[Mark Stebbeds]

I was less than happy with the bass sound on one of my mixes, and a comment I read in an interview regarding applying "mastering" processing to individual tracks before mixing tempted me to try an experiment.

You mean using compression/limiting and EQ? Beyond that, what do you mean by "mastering" individual tracks ?

Mark

[Cary B. Cornett]

In answer to replies so far...

The bass used was a Gibson (I forget which model), and the player, Pete Bankert (who owns Rock City studio in Ann Arbor), always uses a pick. It was originally tracked to analog 24 track (a Soundcraft), using a Countryman DI through the mic pre and EQ of a Flickinger N32B console, and compressed to tape with either an 1176 or an LA-3.

I don't have a 24 track machine, so tracks were transferred at a friend's studio from an MCI JH-24 into PT HD, burned to CD, and from there imported into my SAW rig.

The "A" bass sound went through a Sonoris Leveler (normal, compress, -10 dB, 45 Hz, with 3 dB gain makeup), Then a Studio Levelizer using only the Peak Limit set for 29% with 3 dB gain makeup, and last an Ultrafunk Equalizer R3 with 3 bands active: 40 Hz HP with Q of 2.1, 150 Hz bell with Q 0.8 and up 6 dB, and a 3872 Hz bell with Q 2.0 down 3 dB.

As said before, I was not quite satisfied with the results...

I decided to try a "mastering" approach on just the bass track using upward compression. First, there was a bit too much hiss on the original track, so I fired up the Diamond Cut DC6 continuous noise filter. My first attempt to filter suffered from a slight buzz-crunch artifact triggered by the fundamental tone, so I decided to process only the HF stuff on the bass. I used Saw's Build Mix to process the original file through the PLParEQ3 plugin. I ran it in Linear Phase mode at Q5 with two HP filter sections to kill any subsonic thumps. I took the "de-thumped" file and processed it twice, once to create just an LF file rolled off at 800 Hz, and again to run an HF file (thru an 800 Hz HP). Next I filtered the HF file thru DC6 to cut down the hiss.

I then set up 4 tracks in MT for the bass: two for the LF file and two for the "noise reduced" HF file. One track of each was left unprocessed, with the other track heavily compressed and limited. Each of the "squeezed" tracks first went through the console compressor, with attack 200 ms, release about 700 ms, ratio 30, and threshold about -30 dB for a sort of "slow ride" leveling action with about 4 dB gain makeup. Next was an Ultrafunk compressor, attack about 30 ms, release about 250 ms, threshold about -26, ratio 6.0, knee 10 dB and about 3 dB gain makeup. Finally a Levelizer, compressing at ratio 2.0, attack 8 ms, release 200 ms or so, threshold about -26 dB, with a limit set at about 20 % and gain makeup about 3 dB.

I played all 4 tracks together with the mix, starting with all at unity gain and pulling down the "HF Squeeze" track about 3 dB to tame the "snap" a bit. I then did a build mix of these bass tracks, results of which you heard in the "Bass B" file.

By now you are thinking "too many steps and too much work". Unfortunately , the latency of the PLParEQ3 forced me to pre-process the tracks so that I could slide them back into place. I felt that the use of linear-phase filters for the crossover was important to avoid phase cancellations when the tracks were re-combined. DC6, of course, is strictly stand-alone, thus forcing another "off-line" step.

I almost forgot to mention that along the way I tried using multi-band compressors on the HF and LF files, but these did not give me the clarity of sound that I wanted.

After all that work, I personally felt that the "B" version of the bass stood up better in the track withouth having to be too high in peak level. Using the upward compression trick allowed be to "fatten up" the track while still preserving transient detail in a way that the usual "downward" compression can't do.

I have a feeling I will be doing a lot more of this...

[bcorkery]

Cary,

B really does stand up. I hope you're being paid by the hour!
Whew!

[Cary B. Cornett]

You mean using compression/limiting and EQ? Beyond that, what do you mean by "mastering" individual tracks ?

Mark

I am not entirely sure what the guy whose interview I saw this in meant, but my take on it is that these days some of the processors used in mastering are sufficiently complex and/or process power hungry that they can be difficult to use in real time on multiple individual tracks of a mix.

Even if I could afford the Weiss processors, for example, I could not easily drop a bunch of them into a mix. The PLParEQ plugin is VERY power-hungry at top quality settings. For example, just one PLParEQ3 applied across a stereo piano track jumped the MT Load in Saw up from a few percent to about 27%. It is one SMOOOTH EQ, but that quality definitely comes at a price.

If I wanted to apply that kind of quality and/or complexity of processing to every track in a mix, I would almost certainly have to process each track "off-line" so as not to run out of processing power.

Just wanting to use upward compression on several tracks individually can force a significant increase in the number of tracks I have to use, because most of the dynamics processors available do not provide the option of an upward compression mode. In time past I have petitioned on this forum for some kind of "accessory" plugin to make this possible, but I guess I just wasn't persuasive enough.

hope this answers your question...