How to remove snare drum ring from a recording?

I’m kind of surprised that you couldn’t capture the profile from one ring and use that over all the recording, but you may be running into the vocal keying part of the tool.

If the “ring” is slightly different every time, you’re pretty much dead. Everyone assumes there are different frequencies associated with each instrument. What’s actually happening is each one is a large collection of sounds each with a bunch of frequencies — and they overlap.

These are the frequencies associated with one single piano note.

If someone is playing a chord with other instruments, that display would be solid with frequencies and no way to tell which ones are associated with what. (sorry I don’t have a smaller pix yet.)

Nearly all posts that start the conversation with “I have a mixed recording and I want to separate something from something else” are doomed.



Thanks for the quick reply. However, what would you suggest I do in a step-by-step manner? From what I can hear, the snare ring is pretty much uniform. When I listen back, it really is surprising how much ring is in the sound. I wonder if this had a lot to do with how the sound engineer placed the mic over the drum. Who knows…these were some recordings I did back in '99.

Would it help if I uploaded a sample of what I am talking about?


Yes. Totally.

I straightened out the pictures.

This is one piano note.

And this is two different piano notes.

The tall humps on the left are correct for G1 and G2, but everything to the right of that is a pseudo-random mixture of all the harmonics and overtones, the stuff that you pay Steinway a lot of money for, and nearly impossible to split.


I want to know what you’re calling snare ring. I think it’s the first time I’ve heard that phrase. I can imagine overloading a sound channel and getting odd sounds. Drums are all about impulse sound and those are terrifically hard to record.

Maybe close micing a drum and picking up sounds that normal people can’t hear.

If you had the drum on its own track that kind of surgery is difficult. Doing it in a mixed show…



Attached is a 6-second sample of what I am talking about. I remember back when we were doing recordings the engineer told me to take the O-ring off the drum for some reason. I guess he had some specific thing he was doing with the recordings. That said, I never received the Master recording.

I can provide longer and other samples with different types of tunes we did on the session.

Any help is appreciated!


Welcome, Kman:

Are you aware that you can upload audio clips (and images) in these forums? Just click in the lower tab (“Upload attachment”)?

Before I discovered Audacity, I made an investment on several commercial tools, and am very curious, always making comparisons of features and performance.

See for instance the post below, where I processed a clip, posted the before and after, and other forum participants replicated the Noise Removal processing using an upcoming version of Audacity:



Please do so. I heard and saw the first one and still can’t grasp the meaning of the term “snare”.


Kman: If you have access to Usenet (the granddaddy of all Internet forums), go to the newsgroup comp.dsp (over there, we deal with Digital Signal Processing) and check out a thread entitled

“The Human Voice has been widely characterized, correct?”

The fundamental problem is that a “noise” (a man’s garbage is another man’s richness) is in 99% of the cases, spread across valuable frequencies.

The “remove somehow” that you mentioned is the easy part. The difficult part (theoretically impossible, for the general case of all possible noises) is characterizing the noise.


It seems to be an industry term. The full term would be “Snare Drum Ring”, which is a sound, not to be confused with “Snare Drum O-Ring” which is a circular part:


This is your original:

Is this what you mean by “Snare Ring”??

If that is the case, I subtracted that sound, and this was left:


Good evening. Thanks for the interest in this subject. What I mean by “snare ring” is the snare drum ringing/vibrating every time it is hit. In the sample the snare drum is on beats 2 & 4…just like most straight forward rock or smooth jazz. The sample you provided was actually the hi-hat opening, which is fine. That said, what you did helped slightly. Hopefully by clarifying what I mean by snare, you may be able to provide some other advice. Thanks again.


I’m lost. You’re going to have to give it to us in timeline seconds. Or compare it to a performance that doesn’t have the problem. I think you have just officially joined the group with specialized knowledge of a very tiny problem and the rest of us are going to be little or no help.

If you don’t have the studio split tapes, there’s no way you’re going to dig out one artifact of one instrument in the middle of a mixed performance. Anything we do is going to degrade the other instruments.

We can’t split a mixed performance into individual instruments and that’s basically what you want us to do.



The Snare hits (with accompanied ring) come at 2.25-2.40 secs / 3.85-4.0 secs / and 5.4-5.55 secs into the clip.


It’s very hard to spot in the sample file, although one is aware of something wrong. However, there is too much going on with cymbals, rides, trumpet and so on.
You’ve most likely picked up some room resonances with your kit.
One would normally use a parametric EQ to filter out this boominess (bass drum) or ringing (snare).
Even better is an adjustable (multi-) band compressor. The drums can be made really tight in this fashion.

The mix lacks some foundation, I think, at least in comparison to the solo instrument, whereas the upper end is rather prominent (again, band compression could help to reduce the HiHat/Cymbals in order to make them more distinct in time).

Excluding the trumpet gives:

If possible, provide a longer sample (e.g. through a file sharing service).


I fully agree.

I respectfully beg to differ, Koz. See this post:


Is this the sort of thing that you are after Kman2015?


Since the futuristic solution is still being cooked, I have another potential solution for you.

(1) Play your snare drums, with the same sequence, as it is being recorded. In the middle of that, you will remove the O-ring (which presumably is causing the undesired artifact).

(2) Make 2 WAV files: one with the pesky “snare ring” noise and one without it. Use 16 bit depth/resolution. A short recording will suffice.

(3) Download this free (a gem!) program, Audio DiffMaker

It will tell us the exact difference between the 2 audio clips.


RamonFHerrera, you have previously told us that you are a complete novice with audio, and although you say that you have been studying noise removal techniques for the past couple of weeks, that does not make you an expert. While I applaud your enthusiasm for the subject, I don’t think that you are yet in a position to be able to advise other Audacity users. If you were fully familiar with Audio Diff Maker and understood how it works, you would realise that your suggestion to Kman2015 is nonsense and will not help him to achieve his aims. I would therefore respectfully request that you desist from misguiding other forum users.

Hence the word “potential”.

I may be a novice in audio -as a user of apps written by others- but I have a couple of engineering degrees. One in EE and one in CS. Have spent more than my share of time behind an oscilloscope and solving Laplace and Fourier transforms. I am the author of one of the first applications which solved analog circuits, with graphical input (for the Macintosh, circa 1984). That specialty is based on Fourier transforms, with which folks in the audience are (or should be) to some degree familiar. Additionally, I have a stack of research papers that I am reading as we speak.

In these 2 weeks I have managed to get in touch with some pioneers of state-of-the-art research, plus THE authority in Noise Reduction (as a user), the teacher of teachers, my current hero, this guy:

Who has offered to advice me in my search for automated Noise Elimination (as opposed to inferior Reduction)

At this point I am trying to find the difference in performance between:

(a) Yours Truly (whom I have maintained in the most deliberate manner a novice, for the purposes of my humble study)
(b) A “gun for hire” with decades of experience, who cleaned up an audio clip for me. It took him several weeks.
(c) La crème de la crème in NR techniques (the above mentioned guy who reviewed the work done by level (b)).
(d) Automated Noise Elimination software.

Your “cease and desist” suggestion has been duly noted.


“Free advice is worth every penny you pay for”.

“Failure teaches more than successes”

ps: Kman, one problem you may encounter with Audio DiffMaker is that it seems to run only on Windows XP. :frowning:

My esteemed Steve:

I have the author of the research paper and application on the other line. Shouldn’t we defer this doubt to him?