R/E/P Community

Please login or register.

Login with username, password and session length
Advanced search  

Pages: [1]   Go Down

Author Topic: Acceptable Self Noise Level for VF14-VF14M  (Read 6425 times)

hasbeen

  • Newbie
  • *
  • Offline Offline
  • Posts: 44
  • Real Full Name: Troy Douglas Sutler III
Acceptable Self Noise Level for VF14-VF14M
« on: January 18, 2013, 10:49:52 pm »

I have 2 VF14s with one being marked with an "M" for my U47. I have been testing the difference between the two lately and am considering recording some audio files shortly because I am not able to find any that show what an acceptable noise level is for these type tubes. And by self noise test, I mean that you can hear with your ears using headphones. Hiss. Self noise or whatever you call it. Is there any where to access audio files for reference where you can hear for instance the amount of hiss at X amount of gain from these tubes?

What is considered an acceptable self noise level for VF14 or VF14M?

Also, so I can do it technically, is there any Mac software for a Microphone Self Noise Test? Thanks.
Logged
Who Is John Galt?

klaus

  • Moderator
  • Hero Member
  • *****
  • Offline Offline
  • Posts: 1725
Re: Acceptable Self Noise Level for VF14-VF14M
« Reply #1 on: January 18, 2013, 11:26:40 pm »

I have been doing this for a living, for close to thirty years now, and have not needed a "self noise test" or sound file comparison to determine what is an acceptable level of tube noise. I make a bold claim: you don't need one either.

The reason: it is not easy to isolate whether your mic's output is too low (relative to the mic pre's gain and noise floor), or whether the mic is generating additional noise outside of its usual s/n. The comes a much bigger question: who is the judge of what constitutes an acceptable level of tube hiss? You and I are!

Here is what I do to determine whether a VF14 or VF14M is within spec and quiet enough for professional recording:

1. I make very sure that the tube gets its correct voltages from the power supply: if the supply is correctly calibrated to put out between 105VDC to 108 VDC, measured at pin five of the supply's Tuchel connector, that will translate to ca. 32VDC on the plate (if the tube is healthy!) and ca. 34VDC to 38VDC on the tube heater, measured at the tube, of course.

If the plate voltage does not come up beyond 30-31 VDC, you don't even have to test any further: the tube will forever be noisy, and cannot be fixed- it is shot (unless you understand the voodoo science of tube regeneration, which I sometimes apply, and which actually sometimes works, and sometimes not.)

2. If the measured voltages are all in the ball park, the next thing you need to figure out is whether the noise is actually originating in the tube. There can be noise emanating from a bad capsule (especially a deteriorated M7), or bad circuitry. so you need to isolate the problem: swap heads or use the suspect tube in a known good U47.

3. If it's clearly your tube that is the cause of the noise, you now need to figure out whether you truly have a tube out of spec, or whether you are highly allergic against hiss at certain frequencies in general.

Here is what I would do if I were you, and if I did not have a healthy, known quiet VF14 as reference:
I would talk at normal volume into the mic at a distance of ca. 10" with headphones on. If the background noise is bothersome at that level, I would not use that tube for music recording: noise on a track of quiet music or soft passages does an insidious thing, even when you don't hear it outright through the music: it masks the signal, and collapses the dynamics.

We can even boil it down further: a tube mic like a U47 is not inherently noisier than a transistor mic like a U87 at equivalent gain.
Logged
Klaus Heyne
German Masterworks®
www.GermanMasterworks.com

hasbeen

  • Newbie
  • *
  • Offline Offline
  • Posts: 44
  • Real Full Name: Troy Douglas Sutler III
Re: Acceptable Self Noise Level for VF14-VF14M
« Reply #2 on: January 19, 2013, 12:03:28 am »

Thanks for the quick response.

I am happy to report that both of my tubes sound beautiful for recording purposes. The self noise in this case is intentionally produced at high gain from the preamp for the very purpose of discerning the audible difference between the two tubes in question. In fact, at these increased levels I can say that my VF14 may even be more quiet than the "M" marked tube!

Also, the components in both the mic and power supply are in tip top condition and have been serviced less than a year ago.

So there is no standard  for acceptable self noise in a VF14? Thanks for the info! With me being the judge my tubes pass the test.

I would still like to know if there is any software for the purpose of measuring self noise in microphones though. Please chime with any suggestions. Thanks all.
Logged
Who Is John Galt?

Kai

  • Sr. Member
  • ****
  • Offline Offline
  • Posts: 316
Re: Acceptable Self Noise Level for VF14-VF14M
« Reply #3 on: January 20, 2013, 11:53:31 am »

...if there is any software for the purpose of measuring self noise in microphones though.
This is not a question of software, but of a proper measuring setup.
Because of the huge dynamic and low levels you have to measure you need a 20dB trick on a sine generator if you don't have access to high quality measuring equipment.

What you need for a home made measurement:
- A generator putting out 1KHz sine, be it hard or sofware, or a CD player with a CD, as source for the reference level.
You need two levels, 0dB and -20dB selectable on the generator in any way.
- If using anything else then a hardware sine generator: a simple voltage divider, e.g. a potentiometer to bring down the level of the generator to about 10mV.
- A good mic pre capable of a gain of at least 60dB.
- A RMS voltage meter that can measure low level AC voltages down to 10mV with some resolution (at least 0,5mV for a measuring error of 10% or 1dB).
- A weigthing filter, usually A-curve, be it software (e.g. EQ-emulated) or hardware.
- A capacitor 500-1000pf/100V.
- A capacitor 47pf/100V.
- Knowledge about the reference output of the used capsule.
This means: what voltage does a healthy M7 or K47 capsule produce at the standard reference sound level of 1Pa / 94 dBSL?
I guess something between 10-20mV.

What to do:
1.  Remove the head of the U47
2. Connect the generator to the input pin on th U47 via the 1000pF cap, dial in the M7 ref. level (say 10mV) on the generator, output measured at the generator side of the cap,
2b. Then switch the generator to the -20dB output level (= 1mV at the mic).
3. Dial in a healthy level around 5V on the mic pre's gain.
Measure the output of the mic amp through the A-filter.
Note down this level.
4. Replace the 1000pF cap with the 47pf, connect the other side of it to GND (mic housing), screen the thing a bit with aluminium foil.
5. Measure the output level on the mic pre now, still through the filter.
This noise level is the self noise of the mic circuit reduced by 20dB (because we switched the generator in step 2b).

Now we calculate:
Measured noise divided by the reference level is the S/N ratio reduced by 20dB:
e.g.:
0,007V / 5V = 0,0014
log 0,0014 = -2.85
20 x  -2,85 = -57.1dB
-57,1dB - 20dB = -77.1dB A-weighted S/N ratio

The formula is:
 S/N  ratio (in dB) = (20x  log (noise voltage / reference voltage)) -20

If you like to know the absolut self noise of the mic (without the capsule noise), add 94 dBSL:
94 dBSL - 77.1 dB = 16.9 dBSL self noise A-weighted.
This is around the official figure of a U87 (18dB-A-SL electric self noise).
There is no official statement for the U47, but to my experience there is no significant difference.

Regards
Kai
Logged

J.J. Blair

  • Full Member
  • ***
  • Offline Offline
  • Posts: 130
Re: Acceptable Self Noise Level for VF14-VF14M
« Reply #4 on: January 20, 2013, 09:54:20 pm »

U47s are not perfectly quiet microphones to begin, and I suspect that prior to digital, we were never really bothered by the noise floor that exists on almost all vintage tube mics.  It was always hidden by tape hiss.  I think that in the modern recording age, we are much more aware of these imperfections. 
Logged

hasbeen

  • Newbie
  • *
  • Offline Offline
  • Posts: 44
  • Real Full Name: Troy Douglas Sutler III
Re: Acceptable Self Noise Level for VF14-VF14M
« Reply #5 on: January 20, 2013, 10:59:29 pm »

U47s are not perfectly quiet microphones to begin, and I suspect that prior to digital, we were never really bothered by the noise floor that exists on almost all vintage tube mics.  It was always hidden by tape hiss.  I think that in the modern recording age, we are much more aware of these imperfections.

Good point. Even now, after the vocals are recorded I can hear bleed from the headphones more than any self noise.

I tell you what though. There are subtle differences in sound between the two VF14 tubes in my possession. Having not much experience I would guess this is the case with most of these tubes these days?

Gratefully with the luxury of being able to choose between the two tubes I can see using them both for different occasions. Almost like switching between an M7 and K47/49 for a change of scenery. Anyone else done this before? Thanks.
Logged
Who Is John Galt?

usattler

  • Newbie
  • *
  • Offline Offline
  • Posts: 49
  • Real Full Name: Uwe Sattler
Re: Acceptable Self Noise Level for VF14-VF14M
« Reply #6 on: January 21, 2013, 10:12:30 am »

On page 71 of the 75th anniversary book "Neumann - The Microphone Company" by Anselm Roessler the selfnoise for the U47 amplifier is quoted as 15 dB(A). With the sensitivity of this model listed as 2.5 mV/dyne/cm˛ = 2.5 mV/µbar = 25 mV/Pa, the alsolute (A-weighted) self noise voltage level of the amplifier is 2.8 µV.

These values agree with various Neumann user manuals and brochures from 1958 and 1959.

In some of the earliest official Neumann publications for this microphone different values for sensitivity and self noise can be found. For example, the 1949 'catalog' gives the U47 senstivity as -50 dBv/µbar (= 31.6 mV/Pa), with a 'lower sound pressure limit (self noise) of 24 dB, apparently unweighted. With these values the absolute noise level (unweighted) calculates to 10 µV.

A 1952 Telefunken U47 brochure lists slightly a different sensitivity (-49 dBv/µbar), but the same noise limit of 24 dB.

Consideration of the noise spectrum for typical vacuum tubes explains the 9 dB difference between the A-weighted and the earlier unweighted values.

Logged
Uwe Sattler
Technical Director - Neumann|USA (retired)

klaus

  • Moderator
  • Hero Member
  • *****
  • Offline Offline
  • Posts: 1725
Re: Acceptable Self Noise Level for VF14-VF14M
« Reply #7 on: January 21, 2013, 12:43:03 pm »

Duh! That book sits on my shelf seemingly to thicken it, rather than use it when appropriate.
Thanks, Uwe. Next  time I look there first.
Logged
Klaus Heyne
German Masterworks®
www.GermanMasterworks.com

Jim Williams

  • Hero Member
  • *****
  • Offline Offline
  • Posts: 579
Re: Acceptable Self Noise Level for VF14-VF14M
« Reply #8 on: January 21, 2013, 01:41:11 pm »

Noise spectrum also changes with devices. A set 20hz~20k hz noise measurement or S/N spec won't tell you 'the rest of the story'.

Some tubes have more shot noise, some more hiss. The only sure way to measure and compare individual tube samples is to use a quality analyzer like Audio Precision and do overlay plots.

If you do a noise vs frequency plot you will see some devices have less low frequency noise, some with more higher frequency noise. You will find these traces to not all be that close, there are variations to be measured when a spot s/n test may miss some of those problems.

Yes, tube mics will have more self noise, generally speaking. This is because some jfets are very low self noise, some as low as .7 nv/hz/sq. Follow those with more discrete .5~.7 nv noise BJT's and the mic's self noise can be made extremely low, like around 4~5 db. You will not find a tube with those noise specs.

I stopped using tube mics here for distance pickup or low level sources simply due to that self noise pickup. They have limitations.
Logged

klaus

  • Moderator
  • Hero Member
  • *****
  • Offline Offline
  • Posts: 1725
Re: Acceptable Self Noise Level for VF14-VF14M
« Reply #9 on: January 21, 2013, 04:37:34 pm »

I stopped using tube mics here for distance pickup or low level sources simply due to that self noise pickup. They have limitations.

That does not seem to deter the millions who keep buying recordings or reissues made in the 1950s, 60s and early 1970s, when there were nothing but tube mics-mostly very fine ones- used in recording studios. Listening to, say, an LP or CD of a 1960s Bill Evans Trio, I hear nothing but beautiful music.
Logged
Klaus Heyne
German Masterworks®
www.GermanMasterworks.com

usattler

  • Newbie
  • *
  • Offline Offline
  • Posts: 49
  • Real Full Name: Uwe Sattler
Re: Acceptable Self Noise Level for VF14-VF14M
« Reply #10 on: January 22, 2013, 09:57:57 am »

We may certainly agree on the obvious, namely that the audibility of noise is highly dependent on the spectral distribution of energy across the frequency range. We can further stipulate that the characteristic noise of vacuum tubes is different from that of solid state components, and FETs differ from bipolar transistors etc.

In order to provide comparisons and relevant measurements for what we actually hear, the concept of 'weighting' noise was introduced, and standards were agreed upon. As always however, a nice feature about standards is that there are so many to chose from! Judging by the default filters included in most professional audio analyzers, the audio industry appears to have accepted the 'A-weighting' and 'CCIR 468 weighting' (in its various flavors)...

It is with these considerations that I suggest the acceptable noise levels to strive for CAN be measured and should be are at or below the factory values cited in the earlier responses. You may chose to rely on your ears, but I am a strong defender of the proven concept of 'trust, but verify'.
Logged
Uwe Sattler
Technical Director - Neumann|USA (retired)

Jim Williams

  • Hero Member
  • *****
  • Offline Offline
  • Posts: 579
Re: Acceptable Self Noise Level for VF14-VF14M
« Reply #11 on: January 23, 2013, 11:45:17 am »

That does not seem to deter the millions who keep buying recordings or reissues made in the 1950s, 60s and early 1970s, when there were nothing but tube mics-mostly very fine ones- used in recording studios. Listening to, say, an LP or CD of a 1960s Bill Evans Trio, I hear nothing but beautiful music.

I used them too back then. The difference is then you had tape hiss to mask the mic noise, the mics were quieter than the tape format.
Now days with -120 dbu ADC's and -136 db EIN mic preamps, I'm much more careful about noise pickup. There are also some differences in performance abilities. Bill Evans was a cut above most of the current stuff like Bounce-cy and other lip syncers.

I avoid noise measuring tricks like A weighting. I have that in the Audio Precision, but I never use it. I prefer straight ahead comparisons.
Logged

Kai

  • Sr. Member
  • ****
  • Offline Offline
  • Posts: 316
Re: Acceptable Self Noise Level for VF14-VF14M
« Reply #12 on: January 24, 2013, 05:23:53 am »

... I avoid noise measuring tricks like A weighting...
Wheighting isn't a "trick".
I adapts the measurement to what is practically audible.
But- you have to choose the correct weighting for the situation.

E.g. the famous A-weighting curve is emulating the sensitivity of the ear for very low level signals. Low frequency noise that is inaudible in that case contribues less to the result.
Thus it fits perfectly to the meaurement of low level noise like from a mic.

Linear measurement would mostly compare LF noise on mic's, which you practically don't hear. It would be missleading.

Besides, it's always a good practice to listen to the noise for iregularitis like pops or cracks.
They show that the specimen is a bit "unhealthy" although these types of noises can often be found in mic's.
Short term noises are not covered with usual measurements.

Regards
Kai
Logged

David Satz

  • Full Member
  • ***
  • Online Online
  • Posts: 128
Re: Acceptable Self Noise Level for VF14-VF14M
« Reply #13 on: January 25, 2013, 09:01:01 am »

Just to repeat some fundamentals: The "A" weighting curve adapts the noise to the ear's response at moderately low levels--40 phon, which at 1 kHz = 40 dB SPL. To reflect the ear's response for noise levels that are some 20 - 30 dB lower, as in modern microphones and other circuitry, the curve would need to be somewhat different. The need for weighting is clear, but the use of "A" weighting instead of some other curve is a long-standing practical consensus, rather than being an ideal choice from a scientific standpoint.

In fact commercial considerations enter in to this to a greater extent than an engineer probably ought to feel comfortable with. Marketing people like to print "A"-weighted rms figures on spec sheets for microphones, for example, because they typically come out 8 to 12 dB lower than IEC quasi-peak figures. The latter give a much better approximation of the subjective amount of noise, but even serious microphone manufacturers are more or less forced by the market to print "A"-weighted rms noise specifications because all their competitors are doing so, and someone with a 6 dB quieter microphone could look 6 dB noisier than an inferior competitor.

And the issue of what used to be called "meter ballistics" is equally important (i.e. "rms" vs. "quasi-peak" response). A damped needle or its electronic equivalent will mainly indicate the smooth, continual component of noise while under-representing impulsive ("shot") noise, even though the latter can be much more significant audibly in a given case.

--best regards
Logged

Jim Williams

  • Hero Member
  • *****
  • Offline Offline
  • Posts: 579
Re: Acceptable Self Noise Level for VF14-VF14M
« Reply #14 on: January 25, 2013, 11:21:28 am »

Weighting curves are a trick, on the consumer. You could give them an honest spec, but some add several other tricks so one piece o crud can look as good as something that is better.

Besides A and other noise weighting algo's we add dbu, DBV and other 'reference' points to further confuse the consumer.

If you build top quality gear, your honest measurements will often beat those weighted specs presented by less stellar products. Print out the actual noise vs frequency plots and no one is left confused. Good luck seeing that info. Seems most post record industry collapse audio manufacturers are playing the old Beatle's song:

"Everyone's got something to hide except me and my monkey".

20 odd years ago I saw all the techical info I needed. These days I'm left to test the gear here myself to find out
"The rest of the story".

The average guy will have no idea what he is paying for these days.
Logged
Pages: [1]   Go Up