How to reamp into preamps (repreamp?)

[Paul W]

I'm thinking of doing some controlled experiments on preamps and it seems like it ought to be easy (in principle) to play mic-level recorded signals into preamps for very controlled experiments.

I'm familiar with reamp boxes that generate instrument-level signals for guitar amps and pedals, but I don't recall hearing about re-preamping boxes that generate mic-level signals.

Is that a thing? How would I go about that?

Related question: are there any archives or other good sources of pristine recordings of mic-level signals (presumably through high-end "clean" preamps and converters), for such experiments?

It seems like there ought to be, and it ought to be comparatively easy to compare preamps (at least relative to things like compressors), running varied standard clips through them at different gain levels to see what "color" you get (or not). (And maybe with a few different output impedance levels from the repreamp box?)

 

[pmfalcman]

A passive DI box?
They're meant to convert hi-Z line output to low-Z mic level after all, right?
Unless the hi-Z to low-Z part causes issues it should do the trick.

 

[Khron]

They're meant to convert hi-Z line output to low-Z mic level

* instrument output

(Line-to-mic, especially if galvanic isolation is not necessary and/or the line-level source is balanced, can simply be achieved with a resistive pad / attenuator)

 

[emrr]

Line to mic pad. ‘U’ pad with 40-50dB loss. Don’t overthink it.

 

[Rob Flinn]

There are a few DI boxes that already do this.

For example this Hughes & Kettner one.
https://www.ebay.co.uk/itm/256715023977?_skw=di+box&itmmeta=01JCQYZ4504A990KA0X04E01R2&hash=item3bc5686669:g:AgYAAOSwJAtnNVTL&itmprp=enc:AQAJAAAA4HoV3kP08IDx+KZ9MfhVJKkAPXzDIyiXGUUET0pJkqlYvYzT6D5ST8vNhAyANeQi2n3k8/XxvshPDoxRN8AT+5D37uPqakMDNDZS8h85/I8--hW2xPa97GMKyGCIQrYcH/2p/CHpX8YWUu1wGZhDkRnAfyA9GXjH3GX4/8LREfHl51tJNuNf+YKg0ZmCIrAMuCHZ3L3pKP7pnrNh0GmdMZb8EwoFasZxLwMdEhsPzn3g7giS+bYKdmeheKJMoBpQAItJkIzw7wE+/NJozvuFxm/rq/fkqxymwiFGhwQop--n|tkp:Bk9SR8rC_P7lZA

or this MTR one

https://www.ebay.co.uk/itm/25671476...ztpQkENivQBgHBp7jj/WEC8A==|tkp:Bk9SR8rC_P7lZA

 

[pmfalcman]

* instrument output

I stand corrected.
I took a shortcut -> instrument level hotter than mic level = ~line level.

 

[kingkorg]

I'm thinking of doing some controlled experiments on preamps and it seems like it ought to be easy (in principle) to play mic-level recorded signals into preamps for very controlled experiments.

I'm familiar with reamp boxes that generate instrument-level signals for guitar amps and pedals, but I don't recall hearing about re-preamping boxes that generate mic-level signals.

Is that a thing? How would I go about that?

Related question: are there any archives or other good sources of pristine recordings of mic-level signals (presumably through high-end "clean" preamps and converters), for such experiments?

It seems like there ought to be, and it ought to be comparatively easy to compare preamps (at least relative to things like compressors), running varied standard clips through them at different gain levels to see what "color" you get (or not). (And maybe with a few different output impedance levels from the repreamp box?)

It depends on what you are trying to do.

In lots of cases "color" will depend on mic output stage>pre input stage interaction.

If you are emulating say Neumann TLM mic into a pre, then you can just plug ballanced output of your interface into whatever pre you are testing. TLM has low impedance output and is no different to whatever interface you have. You just varry the output level of your interface, why would you even need anything in between?

If you want to emulate sm57, then you have to emulate the output impedance and transformer performance. Sm57 will behave differently depending if pre is transformer coupled. This goes for most transformer coupled mics, and it's where lots of the "vibe" comes from.

My 1176 into La2a chain creates weird HF roll-off which i like. If i test both of the devices on their own, it is not present. The "magic" happens between 1176 output, and La2a input transformers.

Just done with k67 testing, and unity gain mics at least in theory, can put out 25Vpp, so i don't think you need attenuation att all, you just use output slider. Doug Ford confirmed this to be the case when Røde NTK circuit was designed and tested.

 

[kags]

I just did a shootout with some students of an SM 57 into a handful of different preamps, some with variable impedance (ISA, API512, Neve 1073, Warm wa12, summit, SUPRE, Chandler) and the 57 really sounds different into different pres/impedances.

 

[emrr]

Well sure, as soon as you connect anything other than just a mic to a preamp directly, you are no longer comparing mic/preamp interactions. You can still tell some things about amplifiers.

 

[Paul W]

It depends on what you are trying to do.

In lots of cases "color" will depend on mic output stage>pre input stage interaction.

I'm not sure what all I want to do, and that will depend on how hard various things are.

One of the things I'd like to be able to do is to replay signals through different preamps (or slightly more complicated chains) and then try to adjust the frequency response, and see if that alone makes things audibly much more similar.

Does it make sense to characterize a mic's output impedance as a single number, or is it actually a more complicated function of frequency, as with speakers, and does the shape of that function vary idiosyncratically, or is it regular in some way?

I can imagine building a box (with a couple of transformers in it?) to get a few different impedances, if two or three levels would adequately cover most cases (lowish, normal, high?) Or is even a first cut at the problem much more complicated than that?

My impression is that if you're not saturating transformers, they tend to act like parts of minimum phase EQ circuits, and I'm wondering if using EQ to match frequency response will also tend to correct phase, and give you considerably deeper nulls when null testing, as long as things are operating in their nearly linear ranges. (As well as letting you match levels better---you can't really match levels if you don't EQ match first. For example if something is scooped, you have to pick between matching the midrange or matching the bass & treble, and a null test will either be full of bass & treble, or full of midrange, depending on which you do.)

If that's right, it'd be interesting to see how deep a null you can really get, and what the part that doesn't null looks and sounds like.

I have an idea that some people would be surprised by how little their transformer-coupled circuits are usually doing for their sound, aside from acting as poorly-understood EQ circuits. Or maybe I'm the one that would be surprised, and it's not like that at all.

If you are emulating say Neumann TLM mic into a pre, then you can just plug ballanced output of your interface into whatever pre you are testing. TLM has low impedance output and is no different to whatever interface you have. You just varry the output level of your interface, why would you even need anything in between?

If you want to emulate sm57, then you have to emulate the output impedance and transformer performance. Sm57 will behave differently depending if pre is transformer coupled. This goes for most transformer coupled mics, and it's where lots of the "vibe" comes from.

My 1176 into La2a chain creates weird HF roll-off which i like. If i test both of the devices on their own, it is not present. The "magic" happens between 1176 output, and La2a input transformers.

That's an interesting example.

People often talk about the cool effects of vintage circuits when the circuits are simply in the chain but "not doing anything," like an EQ set to flat, or a compressor that's not actually compressing, and you're not driving things particularly hard either.

One thing I'd like to be able to test is how much of that coloration is just due to EQ effects, and doable with modern circuits and EQ once you understand it, and how much is other kinds of distortion.

 

[kags]

I'm not sure what all I want to do, and that will depend on how hard various things are.

One of the things I'd like to be able to do is to replay signals through different preamps (or slightly more complicated chains) and then try to adjust the frequency response, and see if that alone makes things audibly much more similar.

You could just use REW and run some frequency sweeps (using a signal of a few millivolts, but my guess is all your preamps will have very flat frequency responses until you reach saturation, and then you'll get some harmonic components.

I can imagine building a box (with a couple of transformers in it?) to get a few different impedances, if two or three levels would adequately cover most cases (lowish, normal, high?) Or is even a first cut at the problem much more complicated than that?

There are many threads here about that topic - here is an article with some download plans that people seem to like: https://www.sonarworks.com/blog/learn/transformers

That's an interesting example.

People often talk about the cool effects of vintage circuits when the circuits are simply in the chain but "not doing anything," like an EQ set to flat, or a compressor that's not actually compressing, and you're not driving things particularly hard either.

One thing I'd like to be able to test is how much of that coloration is just due to EQ effects, and doable with modern circuits and EQ once you understand it, and how much is other kinds of distortion.

You are spot on with the transformer (or other input/output circuit) thought. The "magic" added by running through EQ or Compressors without turning any knobs is mostly due to transformer coloration (or some other input/output matching circuit that adds harmonics).

Check out how many preamps and compressors these days allow you to flip a switch (or use a different set of output jacks) to swap out the iron. Products from Shadow Hills, Manley, Warm Audio, DIYRecordingEquipment, along with many others provide this feature.

Many companies have released various transformer color boxes, but seem to discontinue production rather quickly. My hunch is that there is little profit margin in manufacturing and selling affordable boxes, few are actually bought, and they are relatively easy to DIY with spare parts (for many of us).

Definitely look for color options, but the preamp amplifier circuit itself is probably not where you'll get the most bang for your buck...

 

[JeffTD]

Mostly an impedance issue thing. Something like a reverse reamp box does the trick. Avedis makes one specifically:

https://avedisaudio.com/product/lpz/

I personally find line amps to often impart more color on a given signal than simply changing mic pres would. Not all mic pres have a line-in, and this solves that problem. Yes, you are technically getting mic pre and line amp coloration this way. Who cares? Stacking amplifiers of any kind, provided you have impedance matched in a way that doesn’t skew your eq curve too much, usually sounds awesome and is part of what we miss in “modern digital recordings.”

 

[kags]

You could also buy/build a passive "summing mixer," like the one from diyrecordingequipment.com and get the 40dB level drop and impedance matching that the Avedis provides. The Avedis is very nice, but building one or even the kit from DIY saves at leat 50 bucks. Diyrecordingequipment uses this schematic (creative commons license).

​