Elektron-Users

Ah! thanks for the answer! I understand that in some cases you don't want to bring the unwanted noises up (like in the case of some guitar amp where you hear a constant "SHHHHHH") but the MD and R-51 (well, depending on the settings) have very few of that.

I'm far from a knowledgable person when it comes to mastering. I don't know what the difference is between parallel compression and straight compression, but my understanding of compression is that it reduces the distance between peaks and quiet sounds (it reduces the dynamics), effectively bringing up the low noises you don't (or do) want.

An expander would do the opposite; augmenting the amplitude difference between the quiet and loud sounds, and thus could act as a noise reducer.

Anyway, the demo was intended to show the difference between a straight pattern from the MD and the same pattern going through the R-51, so I didn't want to compress or process anything... I only normalized so the sound would be loud whitout peaking, but yeah I could've just exported the file without normalizing too.

BTW, thanks for the book suggestion. I'm going to buy that! I read a couple of Mr Katz's articles and it was very interesting.

And while the subject of mastering and post processing wasn't my initial intention when posting this, I always find interesting to read and discuss on the subject. So if any of what I said isn't accurate, feel free to correct me!

Syl Kougaï wrote:

Because when you normalize, you raise the level of the loud input passage BUT also the low level input, where there is a lot of bad things like noises, lose resolution of the soft excerpt due to digital interpretation, and +3db on that is a lot on certain material...
Some common sense is can also be applied. +3dB isnt a lot at all when you are doing mixing for example. You raise the faders, you lower the faders as you want, you use EQs etc.

Syl Kougaï wrote:

I recommand this book, witch is on mastering but speak about a lot things that we all need to know for mixing :
Mastering audio: the art and the science by Bob Katz

Yes Katz book is good and the information it contains should be interpreted in context of mastering. Normalizing stuff which is sensible recorded at 24 bit does not ruin the sound or make it any more noisier than it is. After all, after recording sound gets treated in mixing which is, by its definition, altering sound levels.

Toni wrote:
Syl Kougaï wrote:

Because when you normalize, you raise the level of the loud input passage BUT also the low level input, where there is a lot of bad things like noises, lose resolution of the soft excerpt due to digital interpretation, and +3db on that is a lot on certain material...
Some common sense is can also be applied. +3dB isnt a lot at all when you are doing mixing for example. You raise the faders, you lower the faders as you want, you use EQs etc.

Syl Kougaï wrote:

I recommand this book, witch is on mastering but speak about a lot things that we all need to know for mixing :
Mastering audio: the art and the science by Bob Katz

Yes Katz book is good and the information it contains should be interpreted in context of mastering. Normalizing stuff which is sensible recorded at 24 bit does not ruin the sound or make it any more noisier than it is. After all, after recording sound gets treated in mixing which is, by its definition, altering sound levels.

Agreed, normalising is fine as long as your not doing it to cover up a poor signal level. Best off trying to record a loud signal with sufficient head room, if your not sending it to anyone for mastering then normalising is pretty sensible.

On Topic:

The Metasonix sounds really nice.

+3dB isnt a lot at all when you are doing mixing for example.

And if you want to re-use a track on an other project, and for example in solo...

Yes Katz book is good and the information it contains should be interpreted in context of mastering.

There is a good session inside the book about mixing.

Normalizing stuff which is sensible recorded at 24 bit does not ruin the sound or make it any more noisier than it is.

It does "ruin" the sound because you multiply a whole number by a number with comma to get a whole number. So if you want to process +3db on a file, It is better to calculate if you can applied +3 o +6 and then apply it with the gain changer.

There is a session about it in the book of bob...

Heya guys:

I'm far more an engineer than I will ever be a musician so here's my thoughts:

Normalisation is often a poor choice, on clean single hit samples it's ok, but in an ideal world you want to be raising your gain stage without raising the noise floor which is something normalising can never do.

As for parallel vs. standard insert compression, parallel compression rocks! It's also commonly known as New York compression, and in some ways acts in the opposite way as a normal compressor.

The benefit of putting a compressor across a group (or sending the thing you compress through a send with a compressor on it) means that the higher value signals (the loudest parts) will remain untouched, and the quieter signals will be raised. So instead of "squashing" the loud signals, you're "raising" the quiet ones. This works particularly well with drums (send them all to a bus and compress the hell out if it, then feed it back in. In addition to making some very punchy drums (think Jon Bonham, Dave Grohl etc) you can control the amount that is fed back in via the aux returns.

To the OP, how are you processing it? Are you just running the audio into the audio input of the module or the CV? I'm saving up to buy a modular ATM and Have been wondering how to integrate my Elektron synths. Can you use GND machines as a CV/Gate source? This is probably more suited to a thread on it's own, but any light you could shed would be much appreciated!

Pretty new to this stuff, but I've been clocking my modular stuff with the gnd-impluse machine out of output C, works great. Had to learn to p-lock every thing that can be p-locked on it (obviously do this once and copy/paste the notes a bunch of times), in case I use a ctl-all or something, because that will mess the pulse up. I'd asked about this in an earlier thread and thought it wasn't working for me, but I was failing to also p-lock the effects and routing. Have started a little utility pattern kind of thing where I have 16 beats of that set up, that also has rs click, and other meat and potatoes stuff to quickly grab....

Wintermute wrote:
To the OP, how are you processing it? Are you just running the audio into the audio input of the module or the CV? I'm saving up to buy a modular ATM and Have been wondering how to integrate my Elektron synths. Can you use GND machines as a CV/Gate source? This is probably more suited to a thread on it's own, but any light you could shed would be much appreciated!

I took the audio out of the MD and connected it into the R-51's input. Since the R-51 is mono (like most modules), I took only one channel out of the MD (the left main). The R-51 output was connected in RCA on my crappy low end m-audio sound interface (I got a new MUCH better audio interface since).

If you would like to get stereo, you'd need 2 R-51.

Like bsmith said, you can use the GND machines as CV triggers on the modular.

An envelope follower before it would be interesting too: there's an LPG compression patch for the up at MW somewhere that lead to interesting results with the maths and QMMG... Id imagine it could get real nasty real quick on the metasonix stuff. I like the sound of the metasonix stuff, Im not overly keen on some of the noises coming from the manufacturers though.