PS Audio: Building the world’s best microphone preamplifier

[pozz]

I had lunch with Dave Blackmer's son and got the impression this pre-amp is based on the well known instrumentation amplifier approach with low noise bi-polar inputs. I wish the noise spec included a BW and source impedance.

Do you know of other commercial mic pres with similar specs? Outside of those meant for lab work (which will typically have single-ended connections, if I understand correctly).

 

[andreasmaaan]

Pro are following the same folklore as audiophile. they also do not like double blind test. There is nothing to hear there. Gearslutz,is,full of people raving about Masselec or SPL (add your favorite brand) eq or compressors. They are very close to each other’s and also very close to plugins. Btw the better the hardware is, the closer it is to plugins.

what I like with physical buttons is that it is easier to concentrate without looking at a screen. Sound wise there is almost no diff (but analogue hardware is a PITA to use wrt to daw+plugins).

end of rant.

what is the connection between Paul and Studer?

+1.

Have met so many people in the pro industry over the years who are deep down the rabbit hole.

 

[Blumlein 88]

I had lunch with Dave Blackmer's son and got the impression this pre-amp is based on the well known instrumentation amplifier approach with low noise bi-polar inputs. I wish the noise spec included a BW and source impedance.

Yes, kind of cheeky they left that out. Their EIN spec would have to have been with inputs shorted which doesn't tell us much. Also wish it listed maximum input voltage.

 

[Geert]

Do these guys actually have any experience as sound engineers or are they just pulling stuff out of their tail end? What's the point of building a state of the art preamp and then running everything through an antique mixing console? That Studer console doesn't even look suitable as a mix down console to me. Or they must go for the minimal EQ no effects approach. But then I wonder why the 40 channels, and why use an analog console if you're using a digital system for tracking anyway?

 

[dfuller]

I wanted to buy a good quality mic preamp in our local pro shop to put in front of my RME ADI-2 PRO fs as it has no mic input so I could use my Behringer ECM8000. The dealer more or less said that almost all expensive pro mic preamps are actually sound processors tailored for a certain "musical" sound, and mixing engineers chose them individually for specific tasks. His recommendation was to just get the RME mic preamp and be done with it. I didn't though because the smallest (QuadMic) had 4 channels and did cost some €600. I got the UMIK-1 instead.

Yup, you can think of a lot of them as distortion processors as well as mic preamps. Stuff like a Milennia HV-3C is incredibly, amazingly clean though.

What's the point of building a state of the art preamp and then running everything through an antique mixing console?

Probably because of how the studio is wired, they have to run through some part of the console (though this does speak of preposterously crappy studio setup...). Either that or they want the vibe from the sound of the line amps in the console.

But then I wonder why the 40 channels, and why use an analog console if you're using a digital system for tracking anyway?

It's one of two things. 1, a workflow thing where you actually want the analog front end. 2, they want the sound of the console. Old consoles aren't exactly the cleanest things ever and they do impart a noticeable color onto things.

 

[scott wurcer]

Do you know of other commercial mic pres with similar specs? Outside of those meant for lab work (which will typically have single-ended connections, if I understand correctly).

No idea I have never surveyed many. The basic Cohen circuit (though the basic idea was published in 1968) with a modern update could meet these specs in fact there are long threads around the web on this. http://leonaudio.biz/double.balanced.mic.amp.notes.pdf

One thread https://groupdiy.com/index.php?topic=48337.0

 

[Blumlein 88]

Do you know of other commercial mic pres with similar specs? Outside of those meant for lab work (which will typically have single-ended connections, if I understand correctly).

This one gets pretty close, but isn't exactly affordable.

https://en-de.neumann.com/v-402#technical-data

 

[AnalogSteph]

This one gets pretty close, but isn't exactly affordable.

https://en-de.neumann.com/v-402#technical-data

That's very good but I wouldn't exactly call it spectacular. It can output +26 dBu into 1 kOhm (close to +28 dBu unloaded), but the 0.00018% THD is given at +12 dBu out and +40 dB gain, and -129.4 dBu(A) of EIN for a 50 ohm source impedance is competent but in no way particularly exceptional (it should translate to about -128 dBu(A) with a more standard 150 ohm source impedance).

The very small difference in EIN between +60 dB and +40 dB does hint at a wide dynamic range... if I subtract -129.4 dB(A) from -129.2 dB(A), this gets me an EIN contribution of -142.7 dB(A) at 40 dB... or -102.7 dBu(A) in terms of output noise, or a maximum dynamic range of ~131 dB(A). (Well, minimum gain is 20 dB, so you only get to about 130 dB(A).) It can also handle a +28 dBu input with pad on, reducing minimum gain to 0 dB. This is why you buy one of these, alongside the extra functionality included (DI inputs, level meter, monitor headphone amplifier). It's basically a very good pro studio grade condenser (no surprise) mic amplifier.

In terms of raw performance as a mic preamp, I'd say the HV-3C should still be a bit better overall, but probably there's not terribly much in it in practice if you have a +24 dBu capable ADC. For the absolute best noise performance with low-impedance dynamic or ribbon microphones, you would order the "ribbon-optimized DC-coupled input" for the HV-3C or use a Cloudlifter with the V 402.

 

[Blumlein 88]

That's very good but I wouldn't exactly call it spectacular. It can output +26 dBu into 1 kOhm (close to +28 dBu unloaded), but the 0.00018% THD is given at +12 dBu out and +40 dB gain, and -129.4 dBu(A) of EIN for a 50 ohm source impedance is competent but in no way particularly exceptional (it should translate to about -128 dBu(A) with a more standard 150 ohm source impedance).

The very small difference in EIN between +60 dB and +40 dB does hint at a wide dynamic range... if I subtract -129.4 dB(A) from -129.2 dB(A), this gets me an EIN contribution of -142.7 dB(A) at 40 dB... or -102.7 dBu(A) in terms of output noise, or a maximum dynamic range of ~131 dB(A). (Well, minimum gain is 20 dB, so you only get to about 130 dB(A).) It can also handle a +28 dBu input with pad on, reducing minimum gain to 0 dB. This is why you buy one of these, alongside the extra functionality included (DI inputs, level meter, monitor headphone amplifier). It's basically a very good pro studio grade condenser (no surprise) mic amplifier.

In terms of raw performance as a mic preamp, I'd say the HV-3C should still be a bit better overall, but probably there's not terribly much in it in practice if you have a +24 dBu capable ADC. For the absolute best noise performance with low-impedance dynamic or ribbon microphones, you would order the "ribbon-optimized DC-coupled input" for the HV-3C or use a Cloudlifter with the V 402.

Saw commentary on SOTA mic preamps I agree with. It said the difference in Sota and other commonly available preamps is 2db of noise, 4-6 db difference in distortion. None of it makes much difference and in use makes effectively no difference. Room noise and microphone distortion swamp all of this, and usually microphone self noise swamps all of this. Microphone pres are a solved problem.

When you consider the advantages of small diaphragm condensers vs LDC's, and the inherent noise limitations it is even more so. Give up a few db of noise for other much more useful attributes to making great recordings.

And we have active ribbons now as well. I plan to get a couple in the near future. Using old school insensitive ribbons which are picky about impedance is akin to still running a turntable. An active ribbon preserves all the benefits of a ribbon without the drawbacks.

 

[Pluto]

What is the most important single parameter when specifying a mic. amplifier?

IMO, overload

In other words, what happens when the amp, is running at the highest gain and a totally unexpected loud burst arrives. Many amps. are specified to have a maximum output of +20dBm. The better ones go to about +24dBm and the top players make +28dBm but can they achieve that when running at a gain of 60dB or more? A great many cannot; the output stage may be perfectly capable of delivering +24dBm when the gain is set to a modest 30dB but wilts with increasing gain. I don't know specifically why this should be but I do know that many so-called professional mic. amps. are incapable of delivering the standards that live (i.e. unpredictable) conditions can impose.

The published specifications of most mic. amps. do not appear to deal with this aspect of the behaviour and, as long as this remains so, the choice of mic. amp. will remain something of a black art.

 

[Blumlein 88]

What is the most important single parameter when specifying a mic. amplifier?

IMO, overload

In other words, what happens when the amp, is running at the highest gain and a totally unexpected loud burst arrives. Many amps. are specified to have a maximum output of +20dBm. The better ones go to about +24dBm and the top players make +28dBm but can they achieve that when running at a gain of 60dB or more? A great many cannot; the output stage may be perfectly capable of delivering +24dBm when the gain is set to a modest 30dB but wilts with increasing gain. I don't know specifically why this should be but I do know that many so-called professional mic. amps. are incapable of delivering the standards that live (i.e. unpredictable) conditions can impose.

The published specifications of most mic. amps. do not appear to deal with this aspect of the behaviour and, as long as this remains so, the choice of mic. amp. will remain something of a black art.

Most mic preamps specify a max input level at minimum and maximum gain. This will show Max input and of course that goes down as you add gain. How else could it work?

 

[tvrgeek]

Viewpoint from the sideline as a consumer only:

Adding various distortions, linear and non-linear, is the prerogative of the musician and producer, executed by the recording engineer, so all for giving them choices in the tools they use. No best preamp, but one of many as a choice.

With regards to overload, I see that as a responsibility of the recording engineers to select the appropriate equipment and adjust it accordingly. RE demands great experience and skill. Unfortunately a few have not achieved it yet. A problem probably brought on by the ease one can produce release ready music from your basement with hundreds of dollars of equipment, not millions. So, everyone thinks they are an RE. Some will be and that is good. I always have found most first albums to be the most musical even if from the basement and of lower recording quality. If someone gets successful, they get over-produced. No thank you Phil Specter.

Live events are usually more of a problem of the lead guitarist ego cranking up his gain, vocalist not wanting to be lost, drummer playing harder until there is nothing but an unintelligent roar that makes me leave. Best thing most clubs could do is hire a professional and put the controls in their hands. A few lessons to singers on how to use a mic often helps.

I can see omitting critical specs from published equipment for consumer goods as how many would understand them, but in true professional equipment, no excuse unless you really have something to hide and then the lack of data should be a hint.

 

[Pluto]

How else could it work?

Some mic. amps. seem to handle everything beautifully, others go nasty no matter what you do.

I'm not sure the real mic amp world is as simple as you imply. I suspect some multi-stage mic amp designs (that use a combination of gain adjustment and inter-stage attenuation to adjust the overall gain) distort in ways not obviously predictable. Likewise, some consoles try to optimize their gain staging for noise performance rather than overload and this can end in tears.

If I had to stick my neck out I'd say that I've had a lot less hassle with mic amps with a nice chunky transformer in the front end than pure electronic balancing.

 

[sergeauckland]

What is the most important single parameter when specifying a mic. amplifier?

IMO, overload

In other words, what happens when the amp, is running at the highest gain and a totally unexpected loud burst arrives. Many amps. are specified to have a maximum output of +20dBm. The better ones go to about +24dBm and the top players make +28dBm but can they achieve that when running at a gain of 60dB or more? A great many cannot; the output stage may be perfectly capable of delivering +24dBm when the gain is set to a modest 30dB but wilts with increasing gain. I don't know specifically why this should be but I do know that many so-called professional mic. amps. are incapable of delivering the standards that live (i.e. unpredictable) conditions can impose.

The published specifications of most mic. amps. do not appear to deal with this aspect of the behaviour and, as long as this remains so, the choice of mic. amp. will remain something of a black art.

However, it doesn't help when the desk channel input can only handle +20dBu, and your mic amp is putting out +28! I don't know of many (any?) desk channels that have a passive balanced attenuator at the input.

Agree about the mic amp being a black art. Especially so if the recording engineer goes for a nice sound rather than ensuring low distortion and noise. I suppose that doesn't matter so much as the choice of microphones, positioning of microphones and choice of mic amps are all effectively 'subjective' choices.

 

[Blumlein 88]

Some mic. amps. seem to handle everything beautifully, others go nasty no matter what you do.

I'm not sure the real mic amp world is as simple as you imply. I suspect some multi-stage mic amp designs (that use a combination of gain adjustment and inter-stage attenuation to adjust the overall gain) distort in ways not obviously predictable. Likewise, some consoles try to optimize their gain staging for noise performance rather than overload and this can end in tears.

If I had to stick my neck out I'd say that I've had a lot less hassle with mic amps with a nice chunky transformer in the front end than pure electronic balancing.

Any such off behavior could be measured.

 

[Pluto]

Any such off behavior could be measured.

Of course it could, but that's not usually a luxury available to freelancers brought in to do a job with the kit someone else (often a production manager working to an accountant) has acquired.

 

[Pluto]

I don't know of many (any?) desk channels that have a passive balanced attenuator at the input.

ISTR the classic Neve front ends used a combination of gain adjustment and attenuation (driven by a multi-layered selector switch sometimes stiff enough to sprain a wrist). Neve got this right but several who copied their approach did not.

 

[Blumlein 88]

ISTR the classic Neve front ends used a combination of gain adjustment and attenuation (driven by a multi-layered selector switch sometimes stiff enough to sprain a wrist). Neve got this right but several who copied their approach did not.

That doesn't seem like it would be optimal for performance. I can see it having benefits in work flow and design of the console.

 

[scott wurcer]

This one gets pretty close, but isn't exactly affordable.

https://en-de.neumann.com/v-402#technical-data

After thinking about it the 1.6nV noise at 20dB gain is a tough one (Earthworks), it requires some pretty extreme resistor values in the feedback network if it is a Cohen circuit.