Noise Floor Testing Data for my Preamps....which Forum?

[quesne]

Hello All:

I have been running some noise floor tests on my various mic preamp channels....recording the noise floor signals, and measuring the levels on an apples-to-apples basis.

I would like to share the data I am gathering with other HR members to discuss my results and to improve my test protocols. The problem is that I'm not sure which forum would be the most appropriate. Can someone please give me some guidance on which forum I should post to?

Thanks

 

[TalismanRich]

I would think the Recording Techniques might be the right place. Comparing different equipment would be one of the things that you might employ when deciding how to do a recording, whether it be selecting a preamp, microphone, or mic technique.

For people that visit the forum regularly, it won't matter. I simply look at new posts, which means it will pop up on the list. I suspect that many people do the same.

 

[quesne]

I would think the Recording Techniques might be the right place. Comparing different equipment would be one of the things that you might employ when deciding how to do a recording, whether it be selecting a preamp, microphone, or mic technique.

For people that visit the forum regularly, it won't matter. I simply look at new posts, which means it will pop up on the list. I suspect that many people do the same.

Thanks Rich! Will do That sounds like a good place to start the conversation

 

[bouldersoundguy]

Or The Rack.

 

[ecc83]

Good morning Quesne, I don't know your technical knowledge level nor what equipment/software you have to hand so please excuse any "grannie's eggs" factor in this reply!

To get consistent and translatable (to other people) noise measurements it is important to have a specified MO. That means for mic pres say, a proper termination at the input. EU is mostly 200 Ohms and USA 150R but in practice it makes very little difference in my experience. The terminating R should be Metal Film and in an XLR plug for shielding.

Noise must of course be specified against a standard input level but since that would be around 1mV and balanced. That is hard for the home bod to generate accurately. In the past I have used 1kHz at 100mV (-20dBV) and as 'precision' attenuator as I could make. You need a known input voltage to find the exact gain since obviously more gain = more noise. You set the 100mV with a decent DMMeter.

Or, the quick and dirty way is to try to speak into the mic at a consistent level and fixed distance. Record at -20dBFS then swap mic for term' plug. Many would say this is a more "real world" way to specify noise? Or, you can have a noise generator, a speaker at say 500mm and feed it 1kHz and check the SPL with a meter (or phone?)

Then, how do you express the results? You could measure the noise output of your AI but as what? To be really accurate you need a true rms meter. Weighted or unweighted and which curve? Better I think to just give the DAW figure in dBFS. Some DAWs, Samplitude is one, give an actual numerical value as well as the dBFS 'bar'. For consistency across other people maybe stick to Audacity?

And lastly it is instructive to run the signal through Right Mark Audio Analyser. The RTA function gives you a graph that can show up unexpected hums and other nasties!

In practice I have found with dynamic mics which need the most low noise gain, modern AI pre amps are better than the ambient noise pickup in my room. Even at a very, subjectively silent 2am there is background noise above electronic noise. That result is for a MOTU M4, an NI KA6 and even a humble Behringer Xenyx mixer. The M4 has a slight edge on the rest.

Dave.

 

[quesne]

Good morning Quesne, I don't know your technical knowledge level nor what equipment/software you have to hand so please excuse any "grannie's eggs" factor in this reply!

To get consistent and translatable (to other people) noise measurements it is important to have a specified MO. That means for mic pres say, a proper termination at the input. EU is mostly 200 Ohms and USA 150R but in practice it makes very little difference in my experience. The terminating R should be Metal Film and in an XLR plug for shielding.

Noise must of course be specified against a standard input level but since that would be around 1mV and balanced. That is hard for the home bod to generate accurately. In the past I have used 1kHz at 100mV (-20dBV) and as 'precision' attenuator as I could make. You need a known input voltage to find the exact gain since obviously more gain = more noise. You set the 100mV with a decent DMMeter.

Or, the quick and dirty way is to try to speak into the mic at a consistent level and fixed distance. Record at -20dBFS then swap mic for term' plug. Many would say this is a more "real world" way to specify noise? Or, you can have a noise generator, a speaker at say 500mm and feed it 1kHz and check the SPL with a meter (or phone?)

Then, how do you express the results? You could measure the noise output of your AI but as what? To be really accurate you need a true rms meter. Weighted or unweighted and which curve? Better I think to just give the DAW figure in dBFS. Some DAWs, Samplitude is one, give an actual numerical value as well as the dBFS 'bar'. For consistency across other people maybe stick to Audacity?

And lastly it is instructive to run the signal through Right Mark Audio Analyser. The RTA function gives you a graph that can show up unexpected hums and other nasties!

In practice I have found with dynamic mics which need the most low noise gain, modern AI pre amps are better than the ambient noise pickup in my room. Even at a very, subjectively silent 2am there is background noise above electronic noise. That result is for a MOTU M4, an NI KA6 and even a humble Behringer Xenyx mixer. The M4 has a slight edge on the rest.

Dave.

Hi Dave:

Thanks very much for your responses both here and on my other posts. From a technical standpoint, I've have a lot of experience building and maintaining research instrumentation and circuitry, primarily for medical and scientific labs. Vacuum tube amplifier circuits, however, are not really in my wheelhouse

Your post motivates me, however, to use this opportunity to either build my own tone generator as a "standard" for these measurements, or to buy one. I see also that there are on-line "tone generators" (like this one - https://onlinetonegenerator.com). Do you think using one of these on-line tone generators is an adequate solution? Assuming that the signal chain (electronic or auditory) is always the same in all testing scenarios, of course.

 

[ecc83]

Bear with....

You don't need a third party app, see the screen shot from Audacity, a really useful audio recorder and editor and totally free.
I would also not use a tone for a reference signal it can give misleading results due to standing waves and 'phase' issues. Constant tones also drive you potty! Although you can't measure noise accurately without a true rms, wideband meter but you don't need to. All you need is to ensure that you use the same microphone* with an SPL meter at the same distance and set the same SPL. I suggest 93dB? Any guitar combo should put that out.
Use that setup for all the preamps and you will have a consistent input signal. You still can't get a gain figure but if pre A delivers XdB more signal than pre B you know A has XdB more gain and can correct for that in the noise figures.

If you want to learn more about valve circuits find Merlin Blencowe's book "Tube pre amp circuits for guitar amps". Don't fuss the "guitar" part, the book contains a wealth of knowledge about the design of valve pre amps from basic data sheets. Maths is very basic but I seem to think a bit of math won't faze you?

*If you wanted a pretty decent, "throw away" dynamic mic to dedicate to the task, the Behringer XM8500 is very cheap but there are even cheaper ones on Zon. BTW 440Hz is 773 mm. A distance where hard, reflective objects might lurk!

Dave.