Reducing Noise From Casio MT-65?

[AfxTwn]

Hi, I have an old Casio MT-65 keyboard and use it in the majority of my recordings but it has a lot of noise/hiss unfortunately. I have listened to the keyboard's built-in speaker when I put my ear next to it and can hear the noise coming through, so it's there even when the keyboard's line out or headphone out isn't connected to my audio interface. I have tried plugging the power adapter into different and separate wall sockets (I normally use an extension cable or a myVolts USB to 7.5v cable connected to my computer for powering the Casio), but this doesn't make any difference. I have also tried a few different cables and this also doesn't solve the issue.

Does anyone know if perhaps some kind of DI box would solve this issue? I know a noise gate does to some extent. I have connected the keyboard to my Zoom G3X multi-effects pedal and used a noise gate on there before going out of that into my Focusrite Scarlett 18i8 audio interface and it does reduce or get rid of the noise but not entirely and with certain sounds from the keyboard, the gate can alter the sound too much and not let certain frequencies through.

The MT-65 has a single RCA line out (it's mono-only) and the output impedance is rated at 2.8 kohms at 1v. I am not sure if that is high impedance or not and whether a DI box will help.

Thanks for any suggestions.

[Arpangel]

My GoTo tool for this job is the Behringer Composer Pro, despite its low price, it has a really good gate, it doesn’t cut off sounds with long decays, it works really well.
Also, I just cut the top end, I don’t loose much worth talking about.
Nice keyboard!

[Mike Stranks]

If the noise is present at source as you indicate, then I don't think a DI box is going to solve the problem you perceive.

You clearly like the sound of the keyboard. Accept the noise during recording, but use good noise reduction software on the recorded track to clean it up. Something like iZotope RX Elements https://www.izotope.com/en/shop/rx-10-elements.html won't cost the earth if you wait until iZotope do one of their frequent sales/offers, and used carefully (not just presets, but identifying the noise in isolation) on the track will remove the noise so that the track sounds much cleaner.

So; redirect your intended spend from a DI box to RX Elements!

[AfxTwn]

Thanks very much for the quick replies. I will check out the Behringer Composer Pro, I see plenty of people are selling used models of the MDX2600 on eBay at pretty cheap prices but I forgot in the meantime, I do actually have a copy of RX10 Standard which I can use to eliminate or reduce the noise. I just thought it might be better to stop the noise at source before it goes into the DAW.

I do really like the keyboard and it suits my Frank Sidebottom-style perfectly, the noise can sound nice in a sort of cheesy and cheap kind of way but when I have recorded multiple tracks and the noise floor adds up, it get's a bit too much and can be distracting.

[muzines]

Have you thought of sampling the keyboard, denoising the samples, and then playing it via a plugin sampler?

All the character, all the sounds, none of the noise floor... these old keyboards are very easy to sample and loop.

[Hugh Robjohns]

I just thought it might be better to stop the noise at source before it goes into the DAW.

No, exactly the reverse is true.

A noise gate only acts after the fact, so it will always mute the noise late and open up again late, thereby damaging the signal in a way that's difficult or impossible to fix.

In contrast, if you record it as is into the DAW, you can then process after the recording to get the sound clean in exactly the way you want, which will give far better results. Not so good for real-time performances, but definitely the way to go for recorded mixes.

[forumuser931182]

If you’re lucky it could just be a dodgy output chip - feeding the speaker and maybe headphone out ( possibly also the mono out ).
You can find the service manual online ( free ). Could try testing a point in the circuit just after the D/A converter.
Then again maybe it’s just a “feature” of the Casio like wow and flutter on a cassette recorder.

[Tomás Mulcahy]

+1 on using RX10 in the DAW, no noise gate, and +1 on sampling it. That's what I do with mine and it makes them much more musical to use. Frees up studio space too. Although I do keep a Casio CT-6000 in the rack for occasional use of its barmy auto-accompaniment patterns. And it looks nice!

All of these Casios are a bit noisy, so without hearing a recording of yours it's impossible to tell if it has issues. The output circuit is pretty standard, would be easy to diagnose assuming you are used to electronics repair. I would not recommend simply replacing a chip. Test, test test.

[AfxTwn]

Thanks for all the replies. I have been doing some spectral de-noise and de-hum in RX10 using the learn function on the sections of audio with just the noise or buzzing in the background and then rendering it to the whole clip and this seems to do the trick while still maintaining most of the original audio/sound. Sometimes a couple of passes is needed but I like being able to apply it to a copy of the original audio within Reaper so I still have the original should anything go wrong.

I have created a couple of very short recordings of the kind of noise I can hear in the background. This first recording: https://www.dropbox.com/s/fxc8txi5ya7jp ... 1.wav?dl=0 has a C chord with the noise present and this other recording: https://www.dropbox.com/s/uoxzm1fc5llwl ... 2.wav?dl=0 has a drum loop playing with the noise.

I have normalized the clips so as to make it a little easier to hear the noise I am talking about but you may need to adjust your volume when the chord and drums kick in. You should be able to hear the background noise which is buzzing and some hiss. This can be more or less and is obviously made worse if I turn the gain up or use compression etc in the mixing stage.

I am pretty happy with the initial results I am getting from RX10 so hopefully that will sort the issue, in some instances I don't mind the background noise as it adds to the character. I don't know too much about electronics and don't own a multimeter so I am not sure if that's something I could take on myself should a repair be required. I can just about take it all apart and clean the necessary parts with contact cleaner before putting it back together again but by and large the keyboard works rather well and I am happy with it so don't want to go tinkering and potentially break something.

With regard to sampling, I have thought about this but I am not too sure how to get the best results and to tbh, I really enjoy using the keyboard itself, I like the tactile nature of using instruments and synths so I look forward to getting it out, plopping it on my desk and just having a mess about. Restricting myself to just using these old keyboards and nothing else has really made me enjoy making music without having so many distractions or spending hours programming things etc. I like the minimal-ness of my current setup. Plus I'm probably one of the few people that genuinely things it sounds amazing and loves the auto-accompaniment patterns and drums.

[Tomás Mulcahy]

That's a normal Casio noise floor.

[Tim Gillett]

... I do actually have a copy of RX10 Standard which I can use to eliminate or reduce the noise. I just thought it might be better to stop the noise at source before it goes into the DAW.

You're quite right. This is usually the best way, if it means avoiding hum ground loops or badly optimised gain staging. Does increasing the volume control at the Casio increase the note volume while not increasing the background noise (and not bringing on distortion)? If so, that's the solution or part of it.

The samples suggest a constant background noise only really audible when no notes are played. This is not too hard to deal with. The difficulty comes if you are also playing quieter chords or notes, or with longer decays than in the samples, where the background noise may become audible when it's no longer masked by the louder notes.

But there can be the misunderstanding that there must be absolutely no background noise underneath any musical performance. It's not necessary. The real world is full of background noise. The trick is to make sure the noise doesnt detract from the performance rather than see the need to squash any noise to death as an end in itself.

[AfxTwn]

That's a normal Casio noise floor.

Thanks for confirming, I thought that might be the case.

You're quite right. This is usually the best way, if it means avoiding hum ground loops or badly optimised gain staging. Does increasing the volume control at the Casio increase the note volume while not increasing the background noise (and not bringing on distortion)? If so, that's the solution or part of it.

The samples suggest a constant background noise only really audible when no notes are played. This is not too hard to deal with. The difficulty comes if you are also playing quieter chords or notes, or with longer decays than in the samples, where the background noise may become audible when it's no longer masked by the louder notes.

But there can be the misunderstanding that there must be absolutely no background noise underneath any musical performance. It's not necessary. The real world is full of background noise. The trick is to make sure the noise doesnt detract from the performance rather than see the need to squash any noise to death as an end in itself.

I did some more tests and the noise floor increases as I turn up the volume on the Casio itself. When I am not playing any notes, I can hear the noise unless I turn the volume down to under halfway, then as I increase the volume, the noise becomes louder. Playing notes does indeed cut out or at least reduce a lot of the noise but as soon as I leave go of the keys, the noise can be heard more prominently.

I also tried some recordings using both less gain on my audio interface input and less volume from the output of the keyboard itself. This does reduce the noise but obviously the recordings are then too quite and so either normalizing them or gain staging them later increases the noise in the background and I am back to square one.

I think as Tomas said, this is a pretty normal quirk of these old and cheap Casio keyboards and is something I can live with but as you pointed out Tim, the problem is when it becomes distracting which in some instances it can be. However, RX10 definitely helps with reducing the volume and noticeability of the noise without taking away from the overall sound, I can just use it a little and still leave behind some background noise which adds character without it dominating the sound or being annoying. I also found that Neutron 4's gate works well on the drums, particularly the multiband expander gate which can learn the noise profile.

On a similar note, how can I best reduce the background sound of headphone leakage when I am recording vocals? I have an Audio Technica AT2020 mic, again plugged into my Scarlett 18i8 1st gen and I set the gain knob to around the 2-3 o'clock position as that gives me enough gain with plenty of headroom. I tend to use my Beyerdynamic DT 770 Pro 250 oHm headphones for recording vocals but much prefer the sound of my Sennheiser HD 650's when mixing and general playback but obviously as these are open-back headphones, they just leak tonnes of sound which is no good for my mic.

So when recording vocals, I set my headphone level to about the 12 o'clock position on the Scarlett and enable monitoring in Reaper so I can hear the playback of the song and my voice (I may add a reverb or something in the FX on the vocals while recording as I prefer something to cover up my terrible singing and not have completely dry vocals as I sing). Anyway, when playing back the audio and mixing it later, I can hear the headphone-bleed in the background and have to go into RX10 to try to edit it out a little. A lot of it is covered up by the rest of the instruments when mixed in with the whole song but I was wondering how to go about reducing it further.

Is it just a case of turning down the gain on the mic and/or headphones? Sorry for asking another question and my amateur technique when it comes to recording.

[Drew Stephenson]

To reduce headphone leak when recording stick with the 770s for the tracking part. I find that I get next to zero leakage from mine, certainly not enough to worry about even using heavy parallel compression, so I wonder if it's a matter of fit around your skull? (My AKGs don't conform to my head as much as my 770s - maybe you're the reverse?)
It maybe that you just need to turn the volume down. Listening with just one earcup on means that you won't need as much 'you' in the monitoring (if at all) so could be worth a go. I tend to do that if I've had to crank the buffer up and latency is becoming a problem.

[Arpangel]

Part of the sound, and charm, of these things, is the noise, take it out and it looses character.
I’ve used the Composer gate many times on noisy keyboards, and it’s fast enough to work "well enough" but I only use it in extreme circumstances, when the noise is almost intolerable, or, the noise doesn’t fit in with the atmosphere of the track.
This concern to make everything as clean as possible can be counter productive.