limiting noise in rhodes when boosting highs

[tomdavids1]

hey guys,

i'm recording my rhodes, and found that when i boost the high frequencies after its been recorded, its brings out the bell like tone of the rhodes, but it also introduces some unwanted hiss.

is there a way i can make the rhodes as quiet as possible without installing an eq/or preamp into the rhodes?

thanks

[tomdavids1]

anyone?

[mvanmanen]

Are you taking the signal of the rhodes directly off the RCA jack (harp) using a DI?

[tomdavids1]

ive tried that, as well as taking it through the bass boost and volume controls. both with the same result when boosting the highs.

[tomdavids1]

heres the direct box im using. is this the correct kind of di?

[pnoboy]

The first thing to do is to figure out if the noise is coming from the Rhodes' output, from some kind of ground loop or EMI, or from the circuitry into which you're feeding the Rhodes' signal.  For example, if you take the signal from the  1/4" jack on the Rhodes' front panel, turn the volume control all the way down.  This connects the output of the Rhodes' to its common (ground).  Did the noise go away?  If not, the noise is not pickup noise from the Rhodes.  Next, unplug the cord that connects the Rhodes to your circuitry.  Did the noise go away?  If so, there's probably some kind of EMI in the area.  If not, maybe the noise is your circuitry.

[tomdavids1]

in turning down the volume pot, the noise goes away. telling me that the noise is coming from the harp.

the volume of the noise is ok, i just need the volume of the hammers hitting the tines to be louder.

any suggestions?

[The Real MC]

Had the same problem.  After trying many remedies, the solution that worked was the Countryman Type 10 DI plugged into the harp RCA jack.  It brought out the bell tone and reduced the hiss at the same time.

[tomdavids1]

wow! thanks for this man. may have to try the countryman

[tomdavids1]

wow! thanks for this man. may have to try the countryman

so i've gotta ask, how was the country man different than the other options you tried?

for me it just seems like the signal of the hammers hitting the tines is too weak, so i have to boost every stage of my signal chain resulting in added noise.

[tomdavids1]

Had the same problem.  After trying many remedies, the solution that worked was the Countryman Type 10 DI plugged into the harp RCA jack.  It brought out the bell tone and reduced the hiss at the same time.

so i've gotta ask, how was the country man different than the other options you tried?

for me it just seems like the signal of the hammers hitting the tines is too weak, so i have to boost every stage of my signal chain resulting in added noise.

[pnoboy]

Maybe your piano is too quiet.  Are the pickups an appropriate distance from the tines?  Could you possibly have a bad  electrical connection? A Rhodes piano should work well with any ordinary guitar amp in terms of its signal.  To put it another way, a guitar amp should have plenty of gain for a Rhodes.

[The Real MC]

Had the same problem.  After trying many remedies, the solution that worked was the Countryman Type 10 DI plugged into the harp RCA jack.  It brought out the bell tone and reduced the hiss at the same time.

so i've gotta ask, how was the country man different than the other options you tried?

Very high input impedance, 10M for the Countryman (with the switch at 0dB).

I have others around 1M input impedance that altered the tone and did not lower the noise.  Quality brands too, like Radial.  They didn't perform as well as the Countryman.

[pnoboy]

I am 99.99% sure that 10Meg input impedance is not the reason for the quietness of the Countryman DI.  All versions of the stage piano have input impedances at least 10X lower than 1Meg.  I have several amp heads, 1 tube and 3 SS, all with 1Meg input impedance, and all are dead quiet

[The Real MC]

I am 99.99% sure that 10Meg input impedance is not the reason for the quietness of the Countryman DI.

My empirical experiments are pretty strong evidence that is hard to dispute.  I don't know what is inside the Countryman other than ad copy claiming it has FET input stages, and the electronics are encased in epoxy. 

In fact I learned of the trick on Gearslutz from Mr. Jim Williams, an audio engineer with a long history of improving audio equipment for a few decades including some major name recording artists from the 1970s.

All versions of the stage piano have input impedances at least 10X lower than 1Meg.

No, you're confusing input impedance with output impedance.  Stage piano outputs have an output impedance.

I have several amp heads, 1 tube and 3 SS, all with 1Meg input impedance, and all are dead quiet.

As do I.  But guitar amps cut off at about 5K because guitars don't do any higher, so that 5K upper limit is filtering off the noise.

The big difference is the OP wants to record his Rhodes without a guitar amp and direct into a recorder, which is full bandwidth out to 20K where the noise will now be clearly heard.

In my experiments with various DIs and amps, the pickups on the Rhodes are very sensitive to loading from input impedances.  Another member here on EP forum graphed the frequency response of the pickups with varying input impedances and it is very revealing, pretty much mirrors my results.  Their tone changes, and the noise varies.  The pickups seem to be happy with an optimal input impedance.

[tomdavids1]

I am 99.99% sure that 10Meg input impedance is not the reason for the quietness of the Countryman DI.  All versions of the stage piano have input impedances at least 10X lower than 1Meg.  I have several amp heads, 1 tube and 3 SS, all with 1Meg input impedance, and all are dead quiet

hey there! would you mind sharing a clip of your rhodes with the highs boosted. Ive already thrown money at several boxes with no improvement. and before i spend money on the countryman it would be nice to hear yours! thanks

[pnoboy]

I am 99.99% sure that 10Meg input impedance is not the reason for the quietness of the Countryman DI.

My empirical experiments are pretty strong evidence that is hard to dispute.  I don't know what is inside the Countryman other than ad copy claiming it has FET input stages, and the electronics are encased in epoxy. 

In fact I learned of the trick on Gearslutz from Mr. Jim Williams, an audio engineer with a long history of improving audio equipment for a few decades including some major name recording artists from the 1970s.

All versions of the stage piano have input impedances at least 10X lower than 1Meg.

No, you're confusing input impedance with output impedance.  Stage piano outputs have an output impedance.

I have several amp heads, 1 tube and 3 SS, all with 1Meg input impedance, and all are dead quiet.

As do I.  But guitar amps cut off at about 5K because guitars don't do any higher, so that 5K upper limit is filtering off the noise.

The big difference is the OP wants to record his Rhodes without a guitar amp and direct into a recorder, which is full bandwidth out to 20K where the noise will now be clearly heard.

In my experiments with various DIs and amps, the pickups on the Rhodes are very sensitive to loading from input impedances.  Another member here on EP forum graphed the frequency response of the pickups with varying input impedances and it is very revealing, pretty much mirrors my results.  Their tone changes, and the noise varies.  The pickups seem to be happy with an optimal input impedance.

Actually, I'm not confusing input and output impedances.  The impedances of the suitcase pianos I referred to are the input impedances of the preamp circuitry they used, not the output impedance of the harp.  Sorry if my writing wasn't clear.  As for amp frequency response, all my amps are flat out to beyond 10 kHz, and not more than 2 or 3 dB down at 20 kHz.  A Rhodes piano doesn't put out much signal beyond 5 kHz or so, anyway.

In a double-blind test, properly set up, I claim that no one would be able to distinguish any tonal or noise difference between a preamp having an input impedance of 1 meg ohm and 10 meg ohm.  Also consider that the capacitance of the cable that connects the harp to the preamp of choice filters out some highs.  If we assume a cable capacitance of 200 pF, which would account for the cable from the harp to the front panel, and the front panel to the preamp, the impedance of that capacitance at 10 kHz is about 79 k.  That impedance, along with with the fairly low output impedance of the harp, makes it's just a bit hard for me to accept the idea that anyone could hear a difference between 1 and 10 meg ohm preamp input impedance.  I simply don't believe it.  There could be any number of reasons why the Countryman sounds better.

[The Real MC]

In a double-blind test, properly set up, I claim that no one would be able to distinguish any tonal or noise difference between a preamp having an input impedance of 1 meg ohm and 10 meg ohm.  Also consider that the capacitance of the cable that connects the harp to the preamp of choice filters out some highs.  If we assume a cable capacitance of 200 pF, which would account for the cable from the harp to the front panel, and the front panel to the preamp, the impedance of that capacitance at 10 kHz is about 79 k.  That impedance, along with with the fairly low output impedance of the harp, makes it's just a bit hard for me to accept the idea that anyone could hear a difference between 1 and 10 meg ohm preamp input impedance.  I simply don't believe it.  There could be any number of reasons why the Countryman sounds better.

Challenge accepted.  Countryman type 10 in 0dB mode (10M input Z) vs Radial J48 (1M input Z) using an 18in cable from the harp equipped with 1960s green coil pickups.

That was the setup I tested, and I guarantee you WILL hear a distinct difference in tone.

[pnoboy]

Let's grant that you can hear a difference, though a double-blind test would be necessary to remove all doubt.  The real question, however, is why.  There are many possibilities, and you claim that the difference is input impedance.  I checked, and the Radial device has an input impedance of 220 k--we were discussing the difference between 1 meg and 10 meg input impedance.  At 220 k, a very slight difference in tone might be possible to discern, but a blind test would be in order.  At input impedances of 1 meg or higher I claim no difference would be detectable.  However, I am a man of science, and if a valid double-blind test proved me wrong, I would be happy to admit it.  Let me say two things in closing-- one, that good double-blind tests are more difficult to set up and run than most people appreciate, and two, there are many, many things in the signal chain that have a significantly more profound effect on the sound of the Rhodes than the input impedance of almost any normal amp one would use.

Also, allow me to ask a question of you--when you were comparing the countryman to the Radial, were you taking the signal directly from the harp or from the piano's front panel?

[The Real MC]

Directly off the harp.

And I am a man of science too.

[tomdavids1]

Directly off the harp.

And I am a man of science too.

is the pissing contest over between you two?

can we finally actually HEAR your piano?

thanks

[pnoboy]

I never knew that technical discussions and debates were pi--ing contests.  The history of science would be nowhere without them, as responses to scientific papers and hypotheses, both critical and elucidative, are an essential part of the scientific method. 

[Tim Hodges]

Agreed, it's interesting seeing the different viewpoints.

[Tim W]

And to add yet more to this experiment...

I think we might be best able to compare the difference in sound due to input impedance if that is the ONLY thing being altered.  This is the main point of the argument- that the input impedance affects the sound- and that 1Meg makes a difference compared to 10Meg, so it should be the only thing that gets changed in a comparison study.

The Countryman should be used in both recordings but in one recording, a 1.1Meg resistor should be added in parallel with the input to the Countryman.  This will then provide almost a 1 Meg input impedance (if the Countryman's own impedance is close to exactly 10Meg as advertised) and the Rhodes should respond accordingly.

This insures that other things in the signal path are held constant and completely out of the question.  Subbing in a Radial DI changes a lot more than just the input impedance.

It will be hard to duplicate 2 passages exactly on the Rhodes since a human will be playing in both performances a little differently... so this alone might color perceptions of what sounds better.
Both samples must also be normalized so that their amplitude levels are very, very, closely matched- as it is normal for human's to always pick the sample that is a little louder as "sounding better".

This is a really tough experiment to pull off properly in a very well controlled manner!

Tim

[pnoboy]

Tim,

Yes, you have described exactly what to do, and also the difficulty of running the test.  I did a somewhat similar test myself, though I I can't claim it proved too much, because my ears were the only judge, and it was not even single-blind, let alone double blind.  In any case, noticing that the stage piano output is loaded by the 10 k volume pot, and noticing that the sound of the stage piano is quite different when taken directly from the harp, I made the following test.

I set up a little fixture by which I could quickly add resistors in parallel to the harp's output, and then connected the harp output directly to an amp with a 1 meg input impedance.  Here's what I found:

10k vs. 1 meg--pretty dramatic difference in tone.  the 10k resistor in combination with the inductance of the harp formed a low-pass filter that filtered out much of the hammer-impact sound, which made the piano sound duller.

20k vs 1 meg--also, an easily detectable difference in tone, but surprisingly less dramatic than 10k vs. 1 meg.

50k vs. 1 meg--I could still hear a difference, but it was small, and I wasn't at all sure that it was significant.

100k vs. 1 meg--any difference was very small, indeed.  I couldn't be sure that I could play anything on the piano repeatably enough to accurately discern any difference.

So, there you have it--I found a 100k input impedance sufficient to reveal the full sound of the harp.  Maybe a different pair of ears would have been more discerning, but at some point, small adjustments of the amp's tone controls would be far more significant than worrying about the ultimate in high-impedance preamps.  Also, I didn't notice any changes in noise, though the 10k resistor did cut down the output from the harp, which had to have a slight negative effect on SNR.

[AvionKeys]

So...what most everyone is getting at is that you're running up against the limitations of the passive/loaded knobs.
1M is plenty high for an input impedance really, anything higher than 500k and I don't think anyone could tell the difference.
The countryman probably has something else going on that is helping the single.

My personal advice, why pull your hair out, go direct from the harp to a DI with a decent impedance and use an EQ.
Otherwise you're in a real squeezing blood from a stone scenario.
Particularly in the studio it's all about what works best and simple in-the-box eq with a good notch should take care of 95% of your sound.

Good luck mate!

[pnoboy]

I agree.  If the tine-to-pickup spacing is set accurately and consistently, and the speaker system is one that works well with a Rhodes, then a slight bass rolloff and a small notch filter of maybe 2 dB set somewhere between 100 and 200 Hz depending on the amp and speakers, should do the trick.  I have owned 3 Rhodes pianos over the years, and all were dead quiet.  Whenever I encountered noise problems it was always a deficiency in the amp and not the Rhodes itself.  Some guitar amps are poorly designed, or much more commonly, reasonably well designed, but with cheap components.  The biggest problem is often the carbon-composition resistors, which are the noisiest resistors around, and yet very common in many guitar amps.