Preamps headroom and all that

[Oldradiohand]

On paper and in theory yes you spec a mic amp and off you go but it doesn't work like that in the real world.

 

Transients can cause overloads on headroom that may look OK on paper and I recall old Rupert Neve spec-ing his consoles to transmit audio up to 100k !

 

Why you may ask when we can only hear up to around 22k well it's all about the harmonic generation of sound and the more you can resolve at the front end the better it is throughout the audio chain.

 

When I worked at AMS Neve our digital consoles were spec-ed as 32 bit internally with an internal dynamic range of 1,500 db's, a lot you may initially think but when you bus together up to 512 inputs you can see why such spec is required.

 

We used to also get complaints that our EQ had "no sound" and was too clean but as most OP's had been used to hearing all the phase nasties of analogue it was just the shock of actually hearing it squeaky clean.

 

It's all down to taste and how manufacturers build their kit which is also done to a cost point so if you want the best you have to pay for it otherwise you have to live with that "grainy" (yamaha) sound or limited headroom of budget components.

 

22K? are you some kind of a bat? Hands up anyone who can genuinely hear over 15.....

[fincaman]

I am very surprised (and pleased) that no one has had a real go at me but I can only repeat if you build a preamp with the best ICs you can find as long as it has + & - 15 volt rails the headroom will be the same as any other mixer and will only give about 10v rms signal out which I still maintain is 10X what you need, but lets suppose you build that magic mixer with massive headroom which can handle say a 30 volt peak without clipping where are you going to send such a signal ? no power amp or recording system will handle it

[paulears]

Sitting here with my son - who is 21, in the same room. Very unscientific test. I played 22K through my macbook pro, grand son - 14 months old, asleep - woke up. 19K annoyed son, 14.5K annoyed me. That's an age spread 14mths/21/54!

[timsabre]

I am very surprised (and pleased) that no one has had a real go at me but I can only repeat if you build a preamp with the best ICs you can find as long as it has + & - 15 volt rails the headroom will be the same as any other mixer and will only give about 10v rms signal out which I still maintain is 10X what you need, but lets suppose you build that magic mixer with massive headroom which can handle say a 30 volt peak without clipping where are you going to send such a signal ? no power amp or recording system will handle it

 

Extra headroom gives you the opportunity to process the signal further down the line, for example with a compressor. Once it's clipped, you can't do that.

[mervaka]

22K? are you some kind of a bat? Hands up anyone who can genuinely hear over 15.....

~19KHz last time I checked, but that was from a function generator in a quiet room.

 

I am very surprised (and pleased) that no one has had a real go at me but I can only repeat if you build a preamp with the best ICs you can find as long as it has + & - 15 volt rails the headroom will be the same as any other mixer and will only give about 10v rms signal out which I still maintain is 10X what you need, but lets suppose you build that magic mixer with massive headroom which can handle say a 30 volt peak without clipping where are you going to send such a signal ? no power amp or recording system will handle it

Sadly, most of today's "engineers" aren't actually engineers. I doubt many could wire up even a simple opamp buffer circuit, let alone talk about things like filter stability, only reiterate things they've heard that back up an opinion they already hold. As for 'headroom', they are probably talking about a combination of dynamic range (which would also encompass noise floor) and the amount of which the mixer has set aside in its gain structure to allow for peaks in the audio signal.

[J Pearce]

I know a lot of engineers who could challenge that claim Mervaka.

Frequency response is fairly easy to get flat, phase response and transient response are much harder to get right, harmonic distortion worse in my experience.

[RichM]

 

Sadly, most of today's "engineers" aren't actually engineers. I doubt many could wire up even a simple opamp buffer circuit, let alone talk about things like filter stability

 

Who cares? I don't need to be able to do that anymore. Thankfully now I have access to reliable, high quality equipment that enables me to get on with what I'm paid to do - make a band sound as good as possible.

[GaryNattrass]

On paper and in theory yes you spec a mic amp and off you go but it doesn't work like that in the real world.

 

Transients can cause overloads on headroom that may look OK on paper and I recall old Rupert Neve spec-ing his consoles to transmit audio up to 100k !

 

Why you may ask when we can only hear up to around 22k well it's all about the harmonic generation of sound and the more you can resolve at the front end the better it is throughout the audio chain.

 

When I worked at AMS Neve our digital consoles were spec-ed as 32 bit internally with an internal dynamic range of 1,500 db's, a lot you may initially think but when you bus together up to 512 inputs you can see why such spec is required.

 

We used to also get complaints that our EQ had "no sound" and was too clean but as most OP's had been used to hearing all the phase nasties of analogue it was just the shock of actually hearing it squeaky clean.

 

It's all down to taste and how manufacturers build their kit which is also done to a cost point so if you want the best you have to pay for it otherwise you have to live with that "grainy" (yamaha) sound or limited headroom of budget components.

 

22K? are you some kind of a bat? Hands up anyone who can genuinely hear over 15.....

 

OK most people can't hear that high but I suppose I just quoted that as the general frequency response of 44.1k CD's but second and third harmonics can go way beyond that and if they are missing it can still affect tonality.

 

Interesting though I haven't had my ears checked since I was about 10 but I put up my tone osc and at 52 can still hear up to around 16k !

[dbuckley]

but lets suppose you build that magic mixer with massive headroom which can handle say a 30 volt peak without clipping where are you going to send such a signal ? no power amp or recording system will handle it

 

Get yourself down to the nearest "golden ears" music reproduction establishment, and ask to audition a Naim amplifier. Take a CD that tests the transient response of your system and ears. Report back.

[themadhippy]

AAAARGH you mentioned nain,pass me the 5 pin din,and hide that highly capacitive speaker cable

Sadly, most of today's "engineers" aren't actually engineers. I doubt many could wire up even a simple opamp buffer circuit,

pah if were going to do this lets use an ecc83

[dbuckley]

Yes, the dreaded NACA4. Mrs Buckley calls it liquorish. Mine had that about twenty years ago, but has run on ordinary speaker cable since then.

[MikeDennis]

I'm enjoying this thread! Though I think I might be a bit young to participate...

I do have a couple of opinions though - I've often thought that sound "engineer" is a very far-fetched title for my job. Like someone's signature on here - basically, I plug things in.

Also, it's interesting to read something by someone who is very experienced and knowledgable about electronics and electrical components and to read those revelations on pre-amps makes me feel like giving you a pat on the back (however inappropriate that would be). A little more learning and I'll be able to sneer all the harder at the boisterous, arrogant, stubborn touring techs when they try to school me whilst simultaneously abusing my equipment.(EDIT: sorry, probably better on a less intellectual thread)

[paulears]

I think that as I (we?) get older maybe we get less snobby, as we discover things that look, perform and sound good, but aren't on the internet horizon labelled 'Ok to Use'.

 

I'm trying to cram more stuff into my studio and started to remove the monitors from so mounts and spotted a sticker on the back with a date code on - and realised they're 25 year old! I've still got a Yamaha DSP-1, their first product using digital sampled environments to produce reverb. I lost the remote, years ago, and the thing has no controls, so it is locked on Munich Cathedral - but it sounds great!

 

The really silly thing is that in the big rack in the studio is a lot of old kit I don;t use, and a nice power amp for the monitors. I just discovered that this amp, turned on for ever, isn't actually connected to the monitors - they've been running off a Technics tuner-amp that because it wasn't rack mounted, had fell off it's rack tray and was laying on it's back in the bottom of the rack. God knows how long it had been there - I can't even remember messing the wiring around?

[Oldradiohand]

On paper and in theory yes you spec a mic amp and off you go but it doesn't work like that in the real world.

 

Transients can cause overloads on headroom that may look OK on paper and I recall old Rupert Neve spec-ing his consoles to transmit audio up to 100k !

 

Why you may ask when we can only hear up to around 22k well it's all about the harmonic generation of sound and the more you can resolve at the front end the better it is throughout the audio chain.

 

When I worked at AMS Neve our digital consoles were spec-ed as 32 bit internally with an internal dynamic range of 1,500 db's, a lot you may initially think but when you bus together up to 512 inputs you can see why such spec is required.

 

We used to also get complaints that our EQ had "no sound" and was too clean but as most OP's had been used to hearing all the phase nasties of analogue it was just the shock of actually hearing it squeaky clean.

 

It's all down to taste and how manufacturers build their kit which is also done to a cost point so if you want the best you have to pay for it otherwise you have to live with that "grainy" (yamaha) sound or limited headroom of budget components.

 

22K? are you some kind of a bat? Hands up anyone who can genuinely hear over 15.....

 

OK most people can't hear that high but I suppose I just quoted that as the general frequency response of 44.1k CD's but second and third harmonics can go way beyond that and if they are missing it can still affect tonality.

 

Interesting though I haven't had my ears checked since I was about 10 but I put up my tone osc and at 52 can still hear up to around 16k !

 

Having worked with a lot of 20 something radio trainees the post puberty upper limit seems to be between 16 and 18k - 16k at your age is good going. I've never been convinced by this "supersonic harmonic" concept - after all, 16k is about the fourth harmonic of the highest note on a piano. One of my ex colleagues had the theory that Neve's extreme frequency response circuitry sounded good was because, if it worked up to 30k it was working properly at 15k.

[fincaman]

22K? are you some kind of a bat? Hands up anyone who can genuinely hear over 15.....

~19KHz last time I checked, but that was from a function generator in a quiet room.

 

I am very surprised (and pleased) that no one has had a real go at me but I can only repeat if you build a preamp with the best ICs you can find as long as it has + & - 15 volt rails the headroom will be the same as any other mixer and will only give about 10v rms signal out which I still maintain is 10X what you need, but lets suppose you build that magic mixer with massive headroom which can handle say a 30 volt peak without clipping where are you going to send such a signal ? no power amp or recording system will handle it

Sadly, most of today's "engineers" aren't actually engineers. I doubt many could wire up even a simple opamp buffer circuit, let alone talk about things like filter stability, only reiterate things they've heard that back up an opinion they already hold. As for 'headroom', they are probably talking about a combination of dynamic range (which would also encompass noise floor) and the amount of which the mixer has set aside in its gain structure to allow for peaks in the audio signal.

 

I agree the word "engineer" is everywhere I have been finding and repairing faults down to component level all my life so much so that I can tell you the value of any resistor by sight without having to work it out by colour code. (I suppose some people might think that's sad) I have always believed that a good understanding of what is happening inside the mixer helps you get it right but you can't expect anyone who is not technical to understand that. I would bet that a lot of people who call themselves "engineers" wouldn't know how balanced inputs work and why they are used and that is very basic

 

On paper and in theory yes you spec a mic amp and off you go but it doesn't work like that in the real world.

 

Transients can cause overloads on headroom that may look OK on paper and I recall old Rupert Neve spec-ing his consoles to transmit audio up to 100k !

 

Why you may ask when we can only hear up to around 22k well it's all about the harmonic generation of sound and the more you can resolve at the front end the better it is throughout the audio chain.

 

When I worked at AMS Neve our digital consoles were spec-ed as 32 bit internally with an internal dynamic range of 1,500 db's, a lot you may initially think but when you bus together up to 512 inputs you can see why such spec is required.

 

We used to also get complaints that our EQ had "no sound" and was too clean but as most OP's had been used to hearing all the phase nasties of analogue it was just the shock of actually hearing it squeaky clean.

 

It's all down to taste and how manufacturers build their kit which is also done to a cost point so if you want the best you have to pay for it otherwise you have to live with that "grainy" (yamaha) sound or limited headroom of budget components.

 

22K? are you some kind of a bat? Hands up anyone who can genuinely hear over 15.....

 

OK most people can't hear that high but I suppose I just quoted that as the general frequency response of 44.1k CD's but second and third harmonics can go way beyond that and if they are missing it can still affect tonality.

 

Interesting though I haven't had my ears checked since I was about 10 but I put up my tone osc and at 52 can still hear up to around 16k !

 

Having worked with a lot of 20 something radio trainees the post puberty upper limit seems to be between 16 and 18k - 16k at your age is good going. I've never been convinced by this "supersonic harmonic" concept - after all, 16k is about the fourth harmonic of the highest note on a piano. One of my ex colleagues had the theory that Neve's extreme frequency response circuitry sounded good was because, if it worked up to 30k it was working properly at 15k.

 

This "supersonic harmonic" argument has been around for years the idea being that if you limit the upper bandwidth to say 22k this effects things that are lower down in the audible range. I did a test on a digital desk that had variable rolloff and it was very surprising how low you had to go before anyone could tell a difference. I saw some tests on the net of the X32 and it seems it is delibrately rolled off at 20HZ and 20KHZ.

 

but lets suppose you build that magic mixer with massive headroom which can handle say a 30 volt peak without clipping where are you going to send such a signal ? no power amp or recording system will handle it

 

Get yourself down to the nearest "golden ears" music reproduction establishment, and ask to audition a Naim amplifier. Take a CD that tests the transient response of your system and ears. Report back.

 

If the Naim amp you want me to listen to was 100 watts with sensitivity of say 1 volt it would have to be capable of delivering 88,200 watts on transients to handle the example I gave above, that equates to HT rails (undriven) of + and - 1300 volts

 

I doubt that that is the case although there are a few amps around that deliver higher power for a short time