Topic: Why the heck do we use preamps anyway?

[hexyjones]

What kind of A/D device could/would accurately record just the mics voltage output - so later you could just replicate what ever analog setup - feed it the recorded mic signal...that way...you could use any preamp!

I'm guessing that the mic output voltage is so small that a regular A/D would not be used to potential - nor would it have sufficient resolution for such a low level signal...

My question is - would a much higher resolution A/D stage make recording with this approch feasible...?? Is that the issue...?

You would remove all electronics between the mic and the A/D...

It seems that the current approach to recording is based on -playback- ...ie recoring a signal as such that it can be played back later...

The approach I'm suggesting is record the signal so that it can be recreated and manipulated (in either an analog or digital environment   ) into a playback level signal...

It would seem like some sort of proprietary mega-super-resolution A/D | D/A device could do it...is there such a thing...? ( im sort of betting the answer is - "of course you idiot!")

Or not?

[utahtaper]

Ok, I'll bite, I got a few minutes.

What kind of A/D device could/would accurately record just the mics voltage output - so later you could just replicate what ever analog setup - feed it the recorded mic signal...that way...you could use any preamp!

I'm guessing that the mic output voltage is so small that a regular A/D would not be used to potential - nor would it have sufficient resolution for such a low level signal...

My question is - would a much higher resolution A/D stage make recording with this approch feasible...?? Is that the issue...?

You would remove all electronics between the mic and the A/D...

It seems that the current approach to recording is based on -playback- ...ie recoring a signal as such that it can be played back later...

The approach I'm suggesting is record the signal so that it can be recreated and manipulated (in either an analog or digital environment   ) into a playback level signal...

It would seem like some sort of proprietary mega-super-resolution A/D | D/A device could do it...is there such a thing...? ( im sort of betting the answer is - "of course you idiot!")

Or not?

You can use Hi to Low Z transformers to eliminate a pre if you like.
You can manipulate recordings in either analog or digital.
Are you just thinking out loud or is there a reason to all of this?
Does this have anything to do with your UA5 thread and your distain for the Oade mods?   

[Tim]

there's no rule that says you must run a preamp. Doug Oade, your favorite taper, has repeatedley said "there is no preamps like NO preamp". If that's the sound you're going for do what you like. Like UT said, some line transformers should give you enough gain.

[sickrick43]

Considering (click click pop) the guy (pop pop) is still (click pop) using a (pop pop pop) midiman CO2 (click POP) in his (pop pop click pop) signal chain (click CLICK) is this line (pop POP POP) of questioning (click CLICK) really RELEVANT?

R(click)ic(POP)k

I run a pre-amp a/d (V3) to provide phantom, clean gain, and digital conversion (and optical) in one small package, in a convenient, pleasing to my ears, quality way.  Why others do is up to them.  Putting the signal up closer to 0 utilizes more bits, gives more dynamic range (the difference between the quiet vs loud parts) and allows (with a high quality pre-A/D stage) the signal to be recorded at a level that requires less amplification on the playback end, thus allowing reasonable listening volumes on less than optimum playback systems with less added noise from amplification on a less than optimum playback system (or something like that).

[hexyjones]

Considering (click click pop) the guy (pop pop) is still (click pop) using a (pop pop pop) midiman CO2 (click POP) in his (pop pop click pop) signal chain (click CLICK) is this line (pop POP POP) of questioning (click CLICK) really RELEVANT?

R(click)ic(POP)k

Hey pal - I've never heard one click/blip out of my CO2 and I wont stop using until someone can convince me that it sucks.

Still waiting here...

http://www.taperssection.com/yabbse/index.php?topic=28597.0

I run a pre-amp a/d (V3) to provide phantom, clean gain, and digital conversion (and optical) in one small package, in a convenient, pleasing to my ears, quality way.  Why others do is up to them.  Putting the signal up closer to 0 utilizes more bits, gives more dynamic range...

The issue is a preamp...not phantom, not digital conversion...those are irrelevant to the question I asked...

...and allows (with a high quality pre-A/D stage) the signal to be recorded at a level that requires less amplification on the playback end, thus allowing reasonable listening volumes on less than optimum playback systems with less added noise from amplification on a less than optimum playback system (or something like that).

Right - see you do agree with me. The whole idea is you can play it back. - Forget about that for a second...We want to capture the mic signal/voltage without any intervention  - and so accurately that we can use that signal to run into ANY preamp to achieve the sound we are looking for...make sense...?

[Brian Skalinder]

My question is - would a much higher resolution A/D stage make recording with this approch feasible...?? Is that the issue...?

I think the issue is not the resolution of the ADC, but rather the signal-to-noise ratio.  Let's assume you have an uber-high-res ADC (say, 256-bit, 468kHz).  While capturing greater detail in the music, it would simply capture greater detail in the accompanying noise as well.

If you can get strong enough levels into the ADC without a preamp - to ensure utilizing maximum resolution of the ADC - then there's no reason to run a preamp.  I used to run mics > phantom power > line transformers > ADC.  The line transformers acted as my gain stage instead of a full-on preamp.  Not really what you're talking about, but...FWIW.  I've also heard of others running AKG 481s directly into a modSBM-1 because the mics provide a hot enough signal at loud rock shows to feed the modSBM-1s line-in stage directly, without a preamp.

[johnw]

Putting the signal up closer to 0 utilizes more bits, gives more dynamic range

To me this would be the #1 reason to use a preamp. I've been told that even after putting the signal closer to 0, I am still using only ~ 14.5 out of 16 bits in a 44.1/16 recording.

If you are really not wanting to use a preamp, maybe you should look into the Neumann Solution DS digital microphone pair - with the AD built in, there is no use for a pre or external AD. Field recording might be a problem with these though.

[jk labs]

My question is - would a much higher resolution A/D stage make recording with this approch feasible...?? Is that the issue...?

I think the issue is not the resolution of the ADC, but rather the signal-to-noise ratio.  Let's assume you have an uber-high-res ADC (say, 256-bit, 468kHz).  While capturing greater detail in the music, it would simply capture greater detail in the accompanying noise as well.

And even that is too optimistic! The ADC itself has a fixed noisefloor below which there are no details to capture: the "bit-determining" part of the ADC _itself_ yields random noise.  With a mic signal input say 20 dB above this noisefloor would result in a dynamic range of just that, 20 dB (+ some due to oversamplng etc). 

With mics putting out -40 dB at 94 dB SPL I do agree though that a separate mic-pre might not be needed if the line stage provides sufficient gain. The line stage within the sbm-1 provides 26 dB gain (20 + differential drive) ahead of the ADC chip.

Introducing a mic-pre here would be an issue of needing more gain and possibly also optimizing the signal path wrt to impedance/loading, preventing unfavorable settings of level controls etc.

Also, the ADC along with it's digital support circuitry, might not be the best neighborhood for doing amplification of analog signals.

so there are pros and cons

[hexyjones]

My question is - would a much higher resolution A/D stage make recording with this approch feasible...?? Is that the issue...?

I think the issue is not the resolution of the ADC, but rather the signal-to-noise ratio.  Let's assume you have an uber-high-res ADC (say, 256-bit, 468kHz).  While capturing greater detail in the music, it would simply capture greater detail in the accompanying noise as well.

If you can get strong enough levels into the ADC without a preamp - to ensure utilizing maximum resolution of the ADC - then there's no reason to run a preamp.  I used to run mics > phantom power > line transformers > ADC.  The line transformers acted as my gain stage instead of a full-on preamp.  Not really what you're talking about, but...FWIW.  I've also heard of others running AKG 481s directly into a modSBM-1 because the mics provide a hot enough signal at loud rock shows to feed the modSBM-1s line-in stage directly, without a preamp.

Wouldnt any noise inherent in the mic be amplified by the preamp? (assuming that's the "accompanying noise" you refer to) Wouldn't the inherent noise be a fixed ratio - no matter how much gain is added...

Also, wouldn't the uber A/D have a rediculously low niose floor?

[hexyjones]

Putting the signal up closer to 0 utilizes more bits, gives more dynamic range

To me this would be the #1 reason to use a preamp. I've been told that even after putting the signal closer to 0, I am still using only ~ 14.5 out of 16 bits in a 44.1/16 recording.

If you are really not wanting to use a preamp, maybe you should look into the Neumann Solution DS digital microphone pair - with the AD built in, there is no use for a pre or external AD. Field recording might be a problem with these though.

Im not really trying to debate the usefulness of a preamp...just wondering if there was a way to capture the signal right out of the mic and then use it as we please...

To clarify - I think I'm talking about a hypothetical AD and DA combo...

You would come home from a show...set up your UberA/D recorder...hit "play" and then you could run that signal into any mic preamp of your choice....you could produce both an SBM-1 version and a UA5 version from the same mic source...

The signal from the Uber AD>DA would be - in theory - exactly the same as if you were at the show with your mics...

make sense...?

[jpschust]

well you are still involving a mic preamp at some stage of the game if you are bringing it to line level then a/d and then d/a.  whether the gain comes internally from the mic, from an external source, or this uber ADDA you mention.

You would theoretically be introducing more noise into your signal by the way you mention it.  It seems you almost want a reamping technique for mics, but it doesnt quite work that way.

In theory your progression would go Mics > some sort of gain (which would technically be a pre) to bring it to line level for an A/D > D/A> a mic preamp > A/D to record. 

[hexyjones]

well you are still involving a mic preamp at some stage of the game if you are bringing it to line level then a/d and then d/a.  whether the gain comes internally from the mic, from an external source, or this uber ADDA you mention.

You would theoretically be introducing more noise into your signal by the way you mention it.  It seems you almost want a reamping technique for mics, but it doesnt quite work that way.

In theory your progression would go Mics > some sort of gain (which would technically be a pre) to bring it to line level for an A/D > D/A> a mic preamp > A/D to record. 

Nope - it would look like this...(again - I'm trying to avoid bringing the mic output to line level...the only reason that is done is for playback)

For recording:

Mics > Cables > Uber A/D equiped recorder

For post production:

Uber D/A playback > Preamp of your choice > A/D to 44.1 > CD

[Brian Skalinder]

Nope - it would look like this...(again - I'm trying to avoid bringing the mic output to line level...the only reason that is done is for playback)

For recording:

Mics > Cables > Uber A/D equiped recorder

For post production:

Uber D/A playback > Preamp of your choice > A/D to 44.1 > CD

Regardless, you're still amplifying the signal - it just happens that in your example it's happening in a different location/time.  And what's the point of performing A>D, then D>A, then A>D again?  Doesn't make much sense.

It must be fun to live in a fantasyland, Uber-corkScrew(ball). 

[hexyjones]

Nope - it would look like this...(again - I'm trying to avoid bringing the mic output to line level...the only reason that is done is for playback)

For recording:

Mics > Cables > Uber A/D equiped recorder

For post production:

Uber D/A playback > Preamp of your choice > A/D to 44.1 > CD

Regardless, you're still amplifying the signal - it just happens that in your example it's happening in a different location/time.  And what's the point of performing A>D, then D>A, then A>D again?  Doesn't make much sense.

It must be fun to live in a fantasyland, Uber-corkScrew(ball). 

Well - of course - we could have UA5 plug-ins and SBM-1 plug-ins so we would not have to perform the addtional D/A > preamp > A/D step...!!!

I'm just trying suggest a way we could capture the signal from our mics, and just do the preamp part later in post-production....be it analog or digital...

And dont just brush off the "different location and time..." part...that's BIG...

Say, 2 years from now, you luck into some sort of sweet high-end preamp...you think "Boy I wish I had had this thing at the 04 SCI NYE show!" - - Well...with my system - you can...just go back to your "mic master" and send the signal through your new pre...

[Brian Skalinder]

I'm just trying suggest a way we could capture the signal from our mics, and just do the preamp part later in post-production....be it analog or digital...

And dont just brush off the "different location and time..." part...that's BIG...

Yes, it is BIG.  Hence the reason to do it up front to maximize signal:noise.

Wouldnt any noise inherent in the mic be amplified by the preamp?

Yes.  But of greater importance is the inherent noise floor of the preamp and ADC.

I have neither the patience nor desire to educate you on the matter of device noise floor, signal-to-noise ratio, etc.  Spend some time on Google and you'll learn all you need to apply to your uber-gear fantasies.  Or if you're really lucky, JK Labs will chime in again.

[hexyjones]

Ok Ok Ok...forget about the reality of all this...I'm talking about a hypothetical device...assume it exists for a second...

Can you at least admit - that - if it were possible...it would be way cool and useful to our hobby...

[Ed.]

maybe, but i like the preamp i have now, and it kicks ass over my preamp in my stereo at home.

[Brian Skalinder]

Can you at least admit - that - if it were possible...it would be way cool and useful to our hobby...

If there were a way to capture the mic signal - without amplification - with very low noise at very high resolution and add gain later, sure, some people might find it cool.

:closes door on fantasyland:

Until then (or even if then), I'll enjoy the kick-ass gear I have now!   

[Todd R]

We want to capture the mic signal/voltage without any intervention  - and so accurately that we can use that signal to run into ANY preamp to achieve the sound we are looking for...make sense...?

No, to me it really does not make sense.  Perhaps in some completely theoretical world, but not in any real-world sense.  I can understand that you want to capture exactly what the mic is producing -- which is a very low level AC electrical current -- but there is no way to do this perfectly and exactly.  Any means by which to do this, whether capturing it on ferro-magnetic tape oxides or by digitizing it and storing it as a set of digital information, will add noise and anamolies to the original AC electrical signal coming from the mics.  Engineers can do their best to minimize this, but there will always be some level of noise in the system that they design.  The problem then is that the electrical signal coming off the mics is very small and on the order of the level of noise inherent in the system.  This is the noise floor that JKLabs alludes to.  It simply turns out that the best way to configure a system to most accurately capture that electrical signal coming from the mics is to add gain to the signal (as in the function of the mic preamp) so that the inherent noise of the system has less of an effect on the now higher level signal that is captured and digitized.  The process of adding gain adds some noise, but overall, the amount of noise and anomolies that the original mic signal is subject to is less than if no gain were added.

The bottom line is the systems we are using, including a mic preamp stage, are the best methods available to try to meet the theoretical pinnacle you're seeking -- an accurate capture of the electrical signal coming off the mics.  My guess is that the components simply are not available/able to be produced that can have a low enough noise floor that a low-level mic-produced electrical signal can be captured accurately without first applying gain to that signal (ie, using a mic preamp).  And even if such components could be made, they would need to be perfectly tuned to the sensitivity of your mics and the decibel level of the event you are trying to capture.  So even if they could be made, you'd need a different system to record Mule and a different system to record Yonder Mountain, and a different system to record un-amplified chamber music, and a different system....  Just not practical at all, and as I said, nothing that is able to be produced at this point anyway.

[jpschust]

im not sure that it has been elaborated as to why we add gain before our a/d processing.  by adding gain we give the a/d more to work with and on the best a/d chips with little room for error we get more accurate conversion from analog to digital.  that gain allows us a better quality result.

[hexyjones]

We want to capture the mic signal/voltage without any intervention  - and so accurately that we can use that signal to run into ANY preamp to achieve the sound we are looking for...make sense...?

No, to me it really does not make sense.  Perhaps in some completely theoretical world, but not in any real-world sense.  I can understand that you want to capture exactly what the mic is producing -- which is a very low level AC electrical current -- but there is no way to do this perfectly and exactly.  Any means by which to do this, whether capturing it on ferro-magnetic tape oxides or by digitizing it and storing it as a set of digital information, will add noise and anamolies to the original AC electrical signal coming from the mics.  Engineers can do their best to minimize this, but there will always be some level of noise in the system that they design.  The problem then is that the electrical signal coming off the mics is very small and on the order of the level of noise inherent in the system.  This is the noise floor that JKLabs alludes to.  It simply turns out that the best way to configure a system to most accurately capture that electrical signal coming from the mics is to add gain to the signal (as in the function of the mic preamp) so that the inherent noise of the system has less of an effect on the now higher level signal that is captured and digitized.  The process of adding gain adds some noise, but overall, the amount of noise and anomolies that the original mic signal is subject to is less than if no gain were added.

The bottom line is the systems we are using, including a mic preamp stage, are the best methods available to try to meet the theoretical pinnacle you're seeking -- an accurate capture of the electrical signal coming off the mics.  My guess is that the components simply are not available/able to be produced that can have a low enough noise floor that a low-level mic-produced electrical signal can be captured accurately without first applying gain to that signal (ie, using a mic preamp).  And even if such components could be made, they would need to be perfectly tuned to the sensitivity of your mics and the decibel level of the event you are trying to capture.  So even if they could be made, you'd need a different system to record Mule and a different system to record Yonder Mountain, and a different system to record un-amplified chamber music, and a different system....  Just not practical at all, and as I said, nothing that is able to be produced at this point anyway.

Thanks - this answer makes the most sense to me...it's beyond the current technology...but not implausible?

[lose humor attempt]Wait - - -you mean the world isn't completely theoretical?[/lose humor attempt]

[Brian Skalinder]

Thanks, Todd.  +T

[Lil Kim Jong-Il]

There is a physical limit to the sensitivity of a voltage differential detector implemented in silicon.  That sensitivity limit is a certain voltage increment.  The ADC operates by determining the integer multiple of that incremental sensitivity for a given sample of a signal.  

If the voltage increment that can be detected is very small compared to the sampled signal, then the ADC is said to be of high precision or have good resolution.  Such a system will have small relative quantization error.

If the input signal is small and only a few integer multiples of the ADC voltage resolution, then the ADC will have poor precision and the quantization error will be high realtive to the signal.

The point here is that in order to get a high resolution precise digital conversion, one must maximize the peak value of the integer multiples used to describe the signal.  

There are two ways to do this: 1) decrease the differential increment relative to the signal or 2) increase the signal relative to the differential increment.  Because the former option is constrained by the physical limitations of the device implementation, the common method is to scale the input signal to maximize the value of the integer multiple reported by the ADC. 

So to answer your original question, the reason we use preamps is to scale the input signal to allow a better resolution at the ADC.

For your uber-ADC to exist, you are going to have to overcome the physical limits of the detector resolution to get a high precision recording of a low level signal.  

[hexyjones]

There is a physical limit to the sensitivity of a voltage differential detector implemented in silicon.  That sensitivity limit is a certain voltage increment.  The ADC operates by determining the integer multiple of that incremental sensitivity for a given sample of a signal. 

If the voltage increment that can be detected is very small compared to the sampled signal, then the ADC is said to be of high precision or have good resolution.  Such a system will have small relative quantization error.

If the input signal is small and only a few integer multiples of the ADC voltage resolution, then the ADC will have poor precision and the quantization error will be high realtive to the signal.

The point here is that in order to get a high resolution precise digital conversion, one must maximize the peak value of the integer multiples used to describe the signal.   

There are two ways to do this: 1) decrease the differential increment relative to the signal or 2) increase the signal relative to the differential increment.  Because the former option is constrained by the physical limitations of the device implementation, the common method is to scale the input signal to maximize the value of the integer multiple reported by the ADC. 

So to answer your original question, the reason we use preamps is to scale the input signal to allow a better resolution at the ADC.

For your uber-ADC to exist, you are going to have to overcome the physical limits of the detector resolution to get a high precision recording of a low level signal. 

Extra thanks to you...that makes even more sense...

[jk labs]

Thanks - this answer makes the most sense to me...it's beyond the current technology...but not implausible?

This is a "difficult" discussion because you're discussing three very different issues at the same time:
1) levels & noisefloor
2) signal degradation
3) capturing acoustic energy to a representation suitable for storage.

The first is covered in this thread to some extent.

The second.. mmm..  well it's a can of worms anyways and it's xmas soon.

The third: YES there are alternatives in existance. Look for AES papers on digital, optical etc microphones to get a glimpse into what is possible.   

[BobW]

Anyone hearing much about the new digital mics ?

The workshop at AES

Neumann's Solution-D

Looks like Georg Neumann's leading the way......

[MattD]

I looked at the Solution-D almost 2 years ago - even before I picked up my SR77s. I think it required AC for the controller box, if I recall.

[hexyjones]

Anyone hearing much about the new digital mics ?

The workshop at AES

Neumann's Solution-D

Looks like Georg Neumann's leading the way......

Gosh - looks like I'm not as crazy as I sound...

[F.O.Bean]

Say, 2 years from now, you luck into some sort of sweet high-end preamp...you think "Boy I wish I had had this thing at the 04 SCI NYE show!" - - Well...with my system - you can...just go back to your "mic master" and send the signal through your new pre...

well that says it all, hes a cheesehead

[hexyjones]

Say, 2 years from now, you luck into some sort of sweet high-end preamp...you think "Boy I wish I had had this thing at the 04 SCI NYE show!" - - Well...with my system - you can...just go back to your "mic master" and send the signal through your new pre...

well that says it all, hes a cheesehead

Wrong - I've heard a few minutes of SCI - nothing more...dl'ed the T-giving show for friends...

...something about hula hoops?

[BobW]

Say, 2 years from now, you luck into some sort of sweet high-end preamp...you think "Boy I wish I had had this thing at the 04 SCI NYE show!" - - Well...with my system - you can...just go back to your "mic master" and send the signal through your new pre...

well that says it all, hes a cheesehead

That's it, Bean, !  I'm keeping my 9Vs       

[F.O.Bean]

Say, 2 years from now, you luck into some sort of sweet high-end preamp...you think "Boy I wish I had had this thing at the 04 SCI NYE show!" - - Well...with my system - you can...just go back to your "mic master" and send the signal through your new pre...

well that says it all, hes a cheesehead

That's it, Bean, ! I'm keeping my 9Vs

hey, thats not fair, i was just messin and thats the best i could think of

sorry really, ive seen and taped cheese, d00d was just making me mad and i was looking for ANY reason to try and burn him

[BobW]

 

[hexyjones]

hey, thats not fair, i was just messin and thats the best i could think of

sorry really, ive seen and taped cheese, d00d was just making me mad and i was looking for ANY reason to try and burn him

Keep looking!!!

...but beware...I use Bean-O!

[Tim]

Anyone hearing much about the new digital mics ?

The workshop at AES

Neumann's Solution-D

Looks like Georg Neumann's leading the way......

Gosh - looks like I'm not as crazy as I sound...

perhaps better reading comprehension skills... the neumann mic was mentioned by at least 2 other people in one of your other crazy threads.

[BC]

why use a preamp? The answer is simple, to be able to spend more money so we can think our tapes are better. Plus the source info in a txt file is much more interesting with a preamp (and more) in there. Which is more interesting?

1) AKG 481>DAP1

or

2) MK4>KC5>CMC6>Lunatec V3>Benchmark AD2k+>VXpocket V2>Sony Vaio C1VPK.