From the same session as per the “thin” acoustic guitars I was asking about the other, I have a drum clip for you to listen to and hopefully help me with.
I need to stress (as I failed to last time), that this cannot be re-recorded. It is what it is, I just need to make the best of it.
The drums are an electric set, so were DI’d using four channels of our mixing desk. Specifically the four tracks are
Snare
Kick
Overhead
Overhead
Honestly I can hear very little difference between the two Overhead tracks - so I’m not sure what he (the bassists who plugged it all in and makes these decisions) was thinking* - except that the splash cymbals are less ‘splashy’ on one than the other.
Including that I don’t know why he didn’t record the midi events either instead of the audio, or as well as the audio?
In this mix I’ve hard panned them, one left and one right. Kick and snare are dead centre.
Because it’s an electric kit, DI’d in to the desk there’s zero bleed between the four tracks.
My questions then are:
What can I do to calm down those splash cymbals? and
What can I do to bring out the snare rim shots a bit more? I’m considering sample replacement (augmentation), so that option is definitely ON the table!
This sample is raw (other than the conversion to MP3), so no EQ, no reverb, nothing.
So I’m guessing it’s kick, snare, and 2 overheads for the 4 tracks?
In your example, other than the fill, I’m only hearing hihat and crash/splashes… was it just because of the chunk you sampled?
Anyway, in terms of kick/snare, you could always utilize a drum replacement software to get some realistic samples. On the overheads, you won’t be able to resemble (that I’m aware of?), but some tastefully used room reverb could add a touch of realism. Those are my initial thoughts
There’s a bunch I haven’t mentioned lol. Honestly, I think automation for the big changes and maybe compression to smooth it out a bit more. I think if you rely on compression to fix it, you’ll end up with a possible mess. I’ve been wrong a lot… but that’s how I’ve learned the most
Ok, I downloaded both your files above and had a bit of a think about this. There are a number of ways you can approach this - some of which are covered above. But I just worked with what I had, and I came up with a solution I think is pretty cool… It basically equates to automating the overheads.
…But since I didn’t have separate kick and snare tracks, I decided to use phase cancellation. I simply ran the “cymbals only” track in parallel with the “kick snare and cymbals” track. I cut out the sections where the crashes were that I wanted to turn down and faded each crash down to nothing before the next cymbal event. Then I flipped the polarity on the cymbal track. At unity gain it completely deletes the cymbal crashes, but as I gradually turn down the polarity inverted track it brings the cymbal crashes back until it sounded at a more even volume. I was able to tame the loud ride cymbal at the end too.
Of course, you have the cymbal, kick and snare tracks separate, so you can just go through and automate them… but it was a fun exercise! Here is the result.
@ColdRoomStudio Andrew, your end result sounded very good, but I have failed to understand what you’ve written!
Cut out from which track?
Here’s why I asked from which track you cut them out, because if you cut them out, surely there’s nothing left to fade down!
What does this mean?!
It does sound like a fun way of doing it, except for the fact that I’ve not understood it.
What I have managed to do is to improve it using the Cubase built-in Multiband Compressor (picking the ‘Drums 4’ preset). I’ll be reviewing that this morning as just after I applied the Multiband Compression it was bedtime (my “studio” is in our bedroom, so when wife wants to go to sleep, the music stops!)
Sorry Jeff, I’m probably just confusing matters by saying I did it that way.
I only did it that way because I didn’t have access to the separate cymbal, kick and snare signals.
You could do the same thing with simple volume automation (ie setting your DAW project to automatically adjust volume) - just turning down the crash cymbals when they initially hit and then gradually bringing up the volume to normal for the next hihat or ride hit. It is quite a bit of work for a whole song, but it is what mixing is all about.
The way I did it relies on the fact that if you have two identical tracks in your DAW playing at the same volume and you invert the polarity on one of the tracks, the two tracks will cancel one another out completely - you’ll hear complete silence. In your DAW there should be a button on the track that has this control - it’s also known as “phase” and represented by a circle with a slash going through it at a 45 degree angle…
Now, because you posted a cymbals only track above, along with a complete mix of all the drums, when I put both tracks in my DAW together, I could actually subtract the cymbals completely from the mixed drum track by using the principle of polarity inversion (or phase reversal, as it is sometimes called). Try it for your self - It’s amazing how it works!
…So rather than using volume automation, I just cut out the cymbal hits that were too loud and placed them on the polarity inverted track. Now I could control the amount I was able to turn down the cymbal hits because the polarity inverted track was actually subtracting their signal from the original mixed drum track.
That’s about the best I can explain it - it is probably easier to demonstrate than explain! I hope that makes sense!
Just a side note : in Superior Drummer 3, you can drag and drop an audio drum track and it will generate a midi file for you. Except when the track is full of subtleties and ghost notes, this is a very straightforward process that can help tremendously with sample augmentation/replacement.
Without SD3 you can do that manually as well, it’s just longer.
Oh! That is pretty cool! In Cubase 10.5 I had to select the ‘Audio’ menu, then ‘Processes’ -> ‘Invert Phase’, and voila! the duplicated tracks cancel each other out! Silence! It’s…odd.
You need to use a Windows audio mixer, something like Voice Potato for instance. You could also try to switch to the Wasapi driver instead of your interface’s, if it isn’t too glitchy.
People have suggested most things I would, which would namely be; automation, eq and some compression. I use a plug-in called drum leveler which has been a game changer for me. You could, with that plug, narrow in on the frequencies of the toms and or cymbals and kind of self automate them up or down respectively.
Probably not the most cost effective solution, but I do find that I use that plug in on all my sessions. Good luck.
