Music Player Network

Previous Thread
Next Thread
Print Thread
Dealing with Humming & Hiss Noise when sampling… #2975077 02/11/19 02:39 PM
Joined: Jul 2001
Posts: 567
D
dalpozlead Offline OP
Gold Member
OP Offline
Gold Member
D
Joined: Jul 2001
Posts: 567
I’ll start trying to test VST noise gates to control hiss when sampling pianos and other decaying patches where we can hear noises (mixer/computer interferences...) when the note is fading out.
Do you sample? Is it a common issue? How to deal? sick
Tks! cool

KC Island
Re: Dealing with Humming & Hiss Noise when sampling… [Re: dalpozlead] #2975086 02/11/19 03:12 PM
Joined: Mar 2012
Posts: 664
B
BbAltered Offline
Gold Member
Offline
Gold Member
B
Joined: Mar 2012
Posts: 664
So - how you ask the question goes a long way in determining how useful the answer is.

My response: yes; noise is rarely a problem in the sampling I have done; I deal by getting rid of unwanted noise.

You will get a lot more useful responses if you re-do your question to tell us 1) what exactly are you sampling (Ac. piano or DP); 2) how exactly are you sampling the instrument (describe the signal chain when sampling and for playback of the samples); 3) what steps have you taken to gain-stage your signal chain; 4) describe for us the conditions during which you hear the noise, and the conditions during which you hear no noise.


J.S. Bach Well Tempered Klavier
The collected works of Scott Joplin
Ray Charles Genius plus Soul
Charlie Parker Omnibook
Stevie Wonder Songs in the Key of Life
Weather Report Mr. Gone
Re: Dealing with Humming & Hiss Noise when sampling… [Re: BbAltered] #2975096 02/11/19 03:54 PM
Joined: Jul 2001
Posts: 567
D
dalpozlead Offline OP
Gold Member
OP Offline
Gold Member
D
Joined: Jul 2001
Posts: 567

Originally Posted By: BbAltered
So - how you ask the question goes a long way in determining how useful the answer is.

My response: yes; noise is rarely a problem in the sampling I have done; I deal by getting rid of unwanted noise.

You will get a lot more useful responses if you re-do your question to tell us 1) what exactly are you sampling (Ac. piano or DP); 2) how exactly are you sampling the instrument (describe the signal chain when sampling and for playback of the samples); 3) what steps have you taken to gain-stage your signal chain; 4) describe for us the conditions during which you hear the noise, and the conditions during which you hear no noise.

sure... more details will help :-)
I'm sampling pianos and synths from a Kurzweil K2661 and vintage Korgs...
The route is: Synth -> Mixer -> Scarlet 2i2 Interface -> PC (running SampleRobot)
Regarding your question # 3: I'm not familiar with those possibilities...
I hear the noise in every patch, specifically when sampling decaying sounds... so it’s more obvious when playing pianos.
Tks!

Re: Dealing with Humming & Hiss Noise when sampling… [Re: dalpozlead] #2975103 02/11/19 04:30 PM
Joined: Oct 2000
Posts: 8,141
burningbusch Offline
MP Hall of Fame Member
Offline
MP Hall of Fame Member
Joined: Oct 2000
Posts: 8,141
Noise is always an issue that needs to be addressed.

First, there is the noise (hum, buzz, etc.) introduced by your electrical system. Dimmers, seemingly anywhere, can produce an additional buzz, but the electrical system has its own noise level. I prefer, whenever possible, to just go off-grid, i.e. battery-powered laptop, bus-powered interface, and battery or phantom-powered DI.

I use software-based noise reduction, in particular, Izotope RX. This is applied after the sampling has taken place. Software noise reduction works by taking a sample of just the noise source (the hiss/buzz) and then applying that as a mask to the recorded material, leaving you the instrument sound less the noise. Izotope RX has evolved into more of an audio restoration product, something I have little call for personally, so I have stuck with an earlier version of the product that's just the noise reduction. There might be other products that produce similar results but I'm satisfied with RX. Hardware noise gates are crude devices that might be OK in live/real-time usage but are going to create more problems than what they solve when sampling.

Busch.

Re: Dealing with Humming & Hiss Noise when sampling… [Re: burningbusch] #2975142 02/11/19 07:09 PM
Joined: Feb 2011
Posts: 3,268
R
Reezekeys Offline
MP Hall of Fame Member
Offline
MP Hall of Fame Member
R
Joined: Feb 2011
Posts: 3,268
Sounds like you may need to work on the gain staging or cabling in your recording setup. A decent D/A and output circuitry on any modern synth should be relatively clean, even at lower levels. However, if what you're hearing is quantization noise caused by companding on an older D/A circuit (which is kinda what it sounds like from your description), you're just dealing with the actual sound of the instrument. Busch's suggestion will probably work the best for that.

Re: Dealing with Humming & Hiss Noise when sampling… [Re: Reezekeys] #2975170 02/11/19 09:14 PM
Joined: Oct 2000
Posts: 8,141
burningbusch Offline
MP Hall of Fame Member
Offline
MP Hall of Fame Member
Joined: Oct 2000
Posts: 8,141
A couple of areas where noise pops up and why it must be addressed.

1) If you're sampling at multiple velocities, the lowest/quietest layers will have a low signal-to-noise ratio. If you normalize these samples, the noise will be VERY apparent. Generally, you normalize or raise gain and then let the sample-playback engine control response velocity. So in this case, noise reduction is a necessity.

2) The background noise built into the samples accumulates as more notes are played. With a single note, the background noise might seem acceptable, but as more notes are played that background noise becomes more audible.

Busch.

Re: Dealing with Humming & Hiss Noise when sampling… [Re: burningbusch] #2975175 02/11/19 09:51 PM
Joined: Feb 2011
Posts: 3,268
R
Reezekeys Offline
MP Hall of Fame Member
Offline
MP Hall of Fame Member
R
Joined: Feb 2011
Posts: 3,268
Originally Posted By: burningbusch
2) The background noise built into the samples accumulates as more notes are played. With a single note, the background noise might seem acceptable, but as more notes are played that background noise becomes more audible.

Not only the background noise in the samples, but the quiescent noise from the analog electronics after the D/A. Coming out of the synth, that noise would stay relatively constant no matter how many notes were sounding. When playing back samples, each individual sample would contain this same amount of quiescent hiss, so sounding multiple notes causes the noise to add together and get much louder.

I once sampled a velocity sensitive patch from my old XP50. Before I started, I edited the patch to remove the vel. sensitivity, so even the quiet layers played back at full volume. This was to get all samples playing as far off the noise floor as possible and use as many bits per sample word as I could without having to apply a lot of gain going into the sampler. I recreated the dynamics in the instrument I used to play the samples back. (I'm pretty sure I'm not telling you anything you don't know, Busch! smile )

Re: Dealing with Humming & Hiss Noise when sampling… [Re: Reezekeys] #2975180 02/11/19 10:12 PM
Joined: Dec 2010
Posts: 1,242
Sundown Offline
Platinum Member
Offline
Platinum Member
Joined: Dec 2010
Posts: 1,242
Originally Posted By: Reezekeys
Sounds like you may need to work on the gain staging or cabling in your recording setup. A decent D/A and output circuitry on any modern synth should be relatively clean, even at lower levels.


^^^
Very well said.

If there is significant noise from a modern synth, I would suspect cabling or something else in the signal path. You might try eliminating the mixer. If you're going from a modern synth to an audio interface, you don't really need the extra gain stage and connections.

Busch is more advanced than me with most topics (so I'll say the following with the belief that I'll probably learn something new today), but I would actually discourage normalizing if you're capturing multi-samples. Assuming you're capturing 24 bits, you've got *a lot* of headroom for peaks, and I don't see much value in raising the gain via normalization if you're ultimately going to lower it again with the sample engine.

But if you want to capture the lower velocity samples with better S/N, than Reeze's solution would work well.


Sundown

Just Finished: Condensation; Two Button Press
Working on: The Jupiter Bluff; Fishing in Kingsbury
Main axes: Kawai MP11 and Kurz PC361
DAW Platform: Cubase
Re: Dealing with Humming & Hiss Noise when sampling… [Re: Sundown] #2975182 02/11/19 10:20 PM
Joined: Dec 2010
Posts: 1,242
Sundown Offline
Platinum Member
Offline
Platinum Member
Joined: Dec 2010
Posts: 1,242
I'll add one other somewhat unrelated tidbit... Sampling old synths can be a great way to breathe new life into them, as strange as that sounds.

My first keyboard was a Yamaha PSR-32. It was a 2-op preset-only FM board with no velocity sensitivity. I got rid of it many years ago, but before I did, I multi-sampled the "Cosmic" patch and the Vibes patch. I created patches with HALion 5 that have velocity controlled filters, envelopes, modulation routings, etc., and the samples sound fantastic. The "Cosmic" patch in particular is a killer comp sound now that it has velocity sensitivity and some animation.

My Roland D-20 has a couple of great sounds, but with early L/A synths, you couldn't filter the PCM partials. I'm going to sample the Recorder tone into HALion, and add filtering, modulation, etc. It will make a great lead for a track I'm working on.


Sundown

Just Finished: Condensation; Two Button Press
Working on: The Jupiter Bluff; Fishing in Kingsbury
Main axes: Kawai MP11 and Kurz PC361
DAW Platform: Cubase
Re: Dealing with Humming & Hiss Noise when sampling… [Re: Sundown] #2975188 02/11/19 11:08 PM
Joined: Oct 2000
Posts: 8,141
burningbusch Offline
MP Hall of Fame Member
Offline
MP Hall of Fame Member
Joined: Oct 2000
Posts: 8,141
Originally Posted By: Sundown

Busch is more advanced than me with most topics (so I'll say the following with the belief that I'll probably learn something new today), but I would actually discourage normalizing if you're capturing multi-samples. Assuming you're capturing 24 bits, you've got *a lot* of headroom for peaks, and I don't see much value in raising the gain via normalization if you're ultimately going to lower it again with the sample engine.


The reason you would do this is you don't know that you will ultimately be lowering it again with the sample engine. You could very well create a "mellow" version of the instrument that only uses the ppp thru p velocity layers (creating a very different sound than what could be achieved through filtering/EQ) or you might want to add the ppp layer to a pad for a completely different sound. In sound design, you want ALL layers to be available for any purpose. If you review the samples in your Korg, Yamaha, Roland, Kurzweil instrument you'll find all at/near peak amplitude. I have run into some software piano where they don't use this approach, but in general, it's the rule. As I deal mostly with hardware instruments, I take an approach common to them.

It's not unusual for the quietest layers to be down -25db or more. I find it much better to prepare those samples with deeper noise reduction and bringing up the gain vs. relying on the sample engine to compensate.

Busch.

Re: Dealing with Humming & Hiss Noise when sampling… [Re: burningbusch] #2975237 02/12/19 02:23 AM
Joined: Jul 2008
Posts: 13,297
J. Dan Online Content
Bannination Free Since '08
10k Club
Online Content
Bannination Free Since '08
10k Club
Joined: Jul 2008
Posts: 13,297
Nobody has mentioned the mixer. Why the extra piece of noise producing gear in the middle? I'd use a high quality passive DI between to provide isolation without adding noise. Use high quality, well shielded cables, shortest length possible, makemsure they don't run near power cables or transformers, and as Busch said, no dimmers or fluorescent lights.

From there, I bow to their always excellent advice.


Dan

Acoustic/Electric stringed instruments ranging from 4 to 230 strings, hammered, picked, fingered, slapped, and plucked. Analog and Digital Electronic instruments, reeds, and throat/mouth.
Re: Dealing with Humming & Hiss Noise when sampling… [Re: burningbusch] #2978732 03/05/19 11:29 AM
Joined: Jul 2001
Posts: 567
D
dalpozlead Offline OP
Gold Member
OP Offline
Gold Member
D
Joined: Jul 2001
Posts: 567
RESOLVED.
I got near zero noise connecting gear directly to the audio interface and reducing sampling volume to 40%.


Moderated by  Dave Bryce, Stephen Fortner 

Powered by UBB.threads™ PHP Forum Software 7.7.3