1. Register NOW and become part of this fantastic knowledge base forum! This message will go away once you have registered.

How low can you go(on the levels when recording acoustic guitar)

Discussion in 'Guitars' started by elephant, Jun 1, 2012.

  1. elephant

    elephant Active Member

    I am micing an acoustic guitar with 2 condensor mics(not the greatest, but fair), using an Art Tps II 2 channel preamp, and a korg D888. In the past I felt I have recorded with the levels too loud, not peaking, but with the gain high enough that it creates a guitar tone that is far from what I desire. I have found at lower levels the tone is more natural and warmer. My question is how, low is too low? It looks like on the recording I did today the levels are mostly between -30db and -20db with a couple spots getting as loud as -13db and as quiet as -40. When mixing and mastering will I be able to get the levels up high enough to make for a satisfying listen for most people?
     
  2. bouldersound

    bouldersound Real guitars are for old people. Well-Known Member

    For 24 bit audio peaks of -13dBFS are probably just about right. You can have peaks down around -30dBFS and still be further above the digital noise floor than you can possible get with 16 bit audio. But be sure you're optimizing your analog gain structure.

    You can make it arbitrarily loud later in the process.

    If you got bad sound at higher levels it's more likely due to the analog signal chain. Digital doesn't change the sound until you clip while analog may get progressively grungier as you drive things harder.
     
  3. elephant

    elephant Active Member

    Thanks for the reply. "Optimizing your analog gain structure" What is this and how do I do it? I need to learn the jargon.
     
  4. RemyRAD

    RemyRAD Guest

    When recording anything, analog pre-amplification of the microphones is still utilized. Outputs from mixers going into a digital audio interface would still be part of the analog realm. This is where gain staging can make all the difference. And that's because regardless of what sample rate or bit depth one records at, the analog section will still be restricted to nothing more than 100-115 DB. And within that spectrum of gain, amplification stages can vary in their tonality. That's because, again structure is generally varied by what is known as " negative feedback ". This has nothing to do with the feedback one gets from an amplifier and PA system. So this jargon can become confusing to early entrants.

    When you change the gain trim of an operational amplifier pre-amplification section, lower gain settings means that greater negative feedback is utilized. Higher gain settings means that less negative feedback is utilized. This negative feedback can make an instrument sound more squeezed or more open. So it's somewhat of a juggling act to get things to sound just the way you want them to sound. Inexpensive units frequently have a fixed first stage gain setting so as to make things a little more goof proof. Older school designs allow for greater variations within a certain finite factor. With those devices, you'll frequently find a switch labeled " pad ". That usually indicates an older school design concept. The pad will include a purely resistive element before the actual pre-amplification. This allows one to compromise for a bit more noise to gain a more open sound even when something does not require the pad due to loud sound pressure levels. So a lot of this depends upon your equipment. I'm an old-school hag with graying hair and love handles. But at times, I can dye my hair and slip on a girdle. And then I'm young again. And I can use digital mixers. I can use cheap stuff that only kids can afford. Or even really expensive ultramodern stuff. Most of the time though, I enjoy my gray and look a little fatter, just like my audio. And that's gain structure. But sometimes you see old guys driving sports cars and young guys driving vans. Other times you see young guys riding crotch rockets and old guys in big fat Cadillacs. So sometimes it's the bloat that rocks the boat. And if you're Canadian, that's what it's all a boat. Otherwise for an American that's what it's all about. So I'm dealing with cheaper equipment, I frequently go for more noise to gain head room. With my good stuff, my old stuff, I can go balls to the wall and laugh all the way to +30 DBM. The cheap stuff can barely squeak out +18 DBU. So there is head room versus noise level. And when setting gain, those are your two opposing forces to deal with.

    It's a gain some people play.
    Mx. Remy Ann David
     
  5. bouldersound

    bouldersound Real guitars are for old people. Well-Known Member

    Basically you have a "window" that you have to fit your signal into at each stage of the signal chain. With digital that window is wide all the way to the top of the level scale. It doesn't change the sound much until it gets too high and then it changes suddenly. An analog window often has a sweet spot, a range where it sounds best. If you go too low you get too close to the noise floor. If you go higher there is progressively more distortion. You have to balance between the distortion and the noise floor to get the best sound. That's why analog metering is generally not focused on peak levels but on average levels. Digital metering is all about peaks because you have an absolute hard boundary between loud enough and too loud.

    For the most part correct analog levels will translate to digital levels of about -18dBFS to -12dBFS.
     
  6. RemyRAD

    RemyRAD Guest

    I've got my windows blacked out so I can sleep better. So I have no window. Neither does my Remote Truck have any windows.

    My windows of opportunities seem to be dimming.
    Mx. Remy Ann David
     
  7. DrGonz

    DrGonz Active Member

    As far as I understand negative feedback, aside from my days in high school, it is resistance added to an input signal in a looping manner. Lets say you feed the first stage of a preamp with the main positive value of an input signal and then at some point right after it passes that op amp or tube stage in the circuit you run the output of this signal through say a resistor. At that stage the signal then can be fed back with a trace or wire to the positive side of that input (jack). That signal is then fed again through that op amp or tube to create a third signal called negative feedback. The signal that was first sent to the op amp or such is outputted as the original signal but should be 180 degree out of phase. That inverted signal is then sent through something like a resistor or other component and then fed right back to the input cancelling the two inverted waves into one new wave. Since increase in amplitude occurs in the op amp or tube stage so does noise, ringing, harmonic and inter-modulation distortion. Negative feedback is used to minimize these alterations and to greatly improve performance.

    Is this correct Remy?

    Every time I read something from you Remy it is right in sync with something that I am working on... When I first came across negative feedback I was really confused as to what it did. I have this Valve Junior Epiphone amp that I am going to mod with a negative feed back loop. So, I just thought to throw out my most basic explanation of this topic to spew out my readings on this subject. I think everyone interested in electronics/audio/etc should understand this concept. Now we should also pay attention to Positive Feedback too, but not so much that it makes us think we know anything... lol. But after collecting these thoughts and adding them to this forum, I feel a bit more understanding of this concept.

    But would positive feedback be like an XLR input or is that still negative feedback??? Oh man this is gonna open a can of worms inside my brain. Back to the research... And to listen to Pink Floyd Waiting for the Worms...
     
  8. RemyRAD

    RemyRAD Guest

    Well, you have a close idea of what's going on but it's basically 180° out of phase to the correct concept, LOL.

    In an Operational Amplifier, whatever you are feeding it, generally goes into the positive or non-inverting input. There is a secondary input which is an inverting input. This allows one to utilize an operational amplifier in place of a input balancing transformer. A simple microphone preamp would have the balance output of the microphone feeding both the non-inverting & inverting input of the Op-Amp. This is how most balanced inputs are now designed so as to eliminate the heavy iron core transformer. While at the same time, coming from the noninverting output of the Op-Amp, a small amount is returned to the inverting input. That's why it's easy to understand the confusion when talking about negative feedback. It's actually positive feedback feeding the inverting input. By doing this, the Op-Amp starts to function differently. Most op amps are not utilized without some kind of feedback to the inverting input. Though, one may see specifications for an operational amplifier indicating its open loop gain value. The actual frequency response and linearity of the operational amplifier varies with the amount of negative feedback utilized. Running an operational amplifier at its full open loop gain value will frequently be limited in its frequency response bandwidth and will produce sizable distortion components since it's being told to function balls to the wall. By adding some negative feedback or positive feedback to the inverting input, it becomes more stable and linear in its operation, limits the amount of gain and extends its upper range frequency response capabilities. While at the same time, when you are taking the positive phase output to the negative phase input, it will change the tonality. Running your operational amplifier with the smallest amount of feedback going into the inverting input, it will be operating at a higher gain structure and produce a more " open sound ". When this feedback is increased to lower the overall amplification value, while it might improve the linearity of the device, lowering the amount of gain that is producing, the tonality will start to sound a little more " squeezed ". So in some instances, with older school design philosophies and concepts, one would put in a resistive pad between the source device feeding the input and the input of the operational amplifier. This pad would then stop excessive level from over driving the operational amplifiers gain structure where you could then lower the positive phase output, which would allow you to reduce the feedback to the phase inverting input allowing for more gain to produce a more open sounding tonality by allowing the operational amplifier to run in a more open loop style. So this trades off in reducing the operational amplifiers high frequency bandwidth limit from say, 100 kHz to a lower value such as 20-30 kHz. It also changes the operational amplifiers linearity value. In such a way so as to change its tonality. In mixers and preamps such as Mackie, the first stage microphone preamp operates at a fixed gain of only 20 DB. And because of that, no pad switch is utilized and keeps the first stage preamp much more consistent sounding, allowing one to feed a low-level source such as a microphone or even a higher level source such as a line level device without fear of overloading the first stage preamp. This is then followed by a second stage operational amplifier working strictly as a buffer amplifier of lesser value, where they gain trim can then be utilized as a lesser value gain device providing for a more consistent sound. Even the highly coveted NEVE 1073 utilized this concept. So Greg Mackie also utilized that concept where you will find no resistive pad switch necessary. That allows the microphone preamp to work at its optimum and most linear way as per the gain structure of the first stage preamp. This makes those mixers a little more proof proof in their operation. It makes them more stable and more consistent sounding. This amounts to the homogenization of the design concept. So then the second stage buffer amplifier while it has a trim control does not have to make the second stage operational amplifier work anywhere outside of its linearity since the gain trim control restricts the secondary buffer amplifier from becoming unstable, operating outside of its gain capabilities and restricts the amount of gain that can produce. The two stages operating in concert with each other allows for more than adequate necessary gain without pushing either operational amplifier beyond their operational design capabilities. So you get 20 DB at the first stage and 30-40 DB at the second stage and the two added together provides for a range of 50-60 DB of clean usable gain. And this mostly pertains to inexpensive IC chip design concepts. Though, as I stated, even the highly coveted NEVE 1073 microphone preamp which is all discrete transistors, also works in this manner. API and others including NEVE went the route of the pad switch relying upon its single stage operational amplifier. These particular units have basically a class A input section combined with a class A/B, output section for greater current drive output capabilities. And within that circuitry, the input sections which are class A, can be overdriven, similarly to the tube input design of a guitar amplifier so as to produce some richer, even order, second harmonic distortion artifacts which can be very musical sounding when it is driven slightly beyond its linearity. And then the output section will still have enough gusto in order to saturate the output transformer which provides yet a different tonality where the electromagnetic forces within the transformer are pushed beyond their capabilities. And the same could hold true for the microphone preamp and input transformer when it is pushed beyond its electromagnetic flux capabilities to the first stage, mostly class A preamp input of the operational amplifier. This is what makes those vintage pieces like the API & NEVE so cool sounding. It really gives you a broad range of tonal variations in which to work with. The NEVE 1073 and its summing/line driving output sibling the 1272 goes further by including a single Class A biased output transistor that can produce yet more even order, second harmonic distortion components. Tubes do it even better because they will " soft overload " compared to a transistors flat clip hard overload. But even there, the output transformer will saturate before that output transistor actually flat top clips.

    There have been some other design concepts for pre-amplification circuits known as feed forward. These are said to produce even less of that squeezed sounding tonality but are a little more involved to design correctly requiring additional circuitry which then also runs up the price yet further. But some people want that totally open sound and are willing to pay for it, dearly. Those are usually reserved for the more upper echelon financially flush folks willing to spend a lot more money for what they want. Those are the guys you see driving around in the really expensive cars and driving them to their really large yachts which cost them millions of dollars. But if you've got millions of dollars to spend on your audio, ain't no problem for those folks wanting the very best, of getting the cleanest, most open sounding equipment on the market. So money can buy you happiness when you have a nearly unlimited budget. Such is not the case for most of the average folks. So it really comes down to what is most practical and what the average person can afford. Thankfully, it's almost impossible to design a tube circuit that features its lush, class A biased circuitry to produce lovely even order, second harmonic overload saturation. But even there, not all tube circuits are created equally. It requires more than one tube of amplification to produce a good tube circuit. And in inexpensive tube preamps IC chips do most of the amplification while running into a single tube operating in a " starved plate " configuration without a proper plate voltage to create nothing more than a tube operating strictly in a saturated configuration for that soft overload tube sound. That's where that ART $50 tube microphone preamp and others whose tubes are not being utilized for any amplification whatsoever. This is where the wool can be pulled over everybody's eyes in thinking they have a real tube preamp. So that's what we could also refer to as creative fraud because in a real tube preamp multiple tubes with full plate voltage for amplification purposes are actually required. So in a guitar preamp, one can get away with a single dual triode of the 12 AX & similar variety.

    I am not an electrical engineer so this is an oversimplified explanation of these design concepts.
    Mx. Remy Ann David
     
  9. StephenByerley

    StephenByerley Active Member

    Wow! There have been some spectacular responses to this question. I might add that if in their current position you have to attenuate the gain below the zero point as it hits the board, then possibly you might have the mics either ) too close, or ) not in the sweet spot. It has been in my experience that turning down the gain alters the structure in a way that mimics compression. If it were me, I would back the mics up a bit, or move them around, or make certain that they are not phasing with each other, and try to record with the gain at the zero if possible. Why? The least amount of distortion exists at the zero mark, so if I can get a good sounding recording at zero throughout the signal path, then I'll most likely have the least amount of additional noise and distortion. If you can use the natural damping factor of the air to provide a natural compression, then the track will have an organic and pleasing sound.
    Just my 2 bits. Good luck!
     
  10. RemyRAD

    RemyRAD Guest

    Stephen, you are in Winter Park, Florida? At Full Sale? Is Jim Michaels still teaching video editing? If he is, tell my colleague from NBC-TV that Remy says hi and sends her best. I haven't talked to him in years. He was there last I heard a few years ago. I believe he, like myself accepted a buyout offer from NBC-TV to reduce their engineering staff. I remember Full Sale from 1979 when they were just a little 8 track studio called Bee Jay Recordings.Their 8 track studio was far less impressive than the one I redesigned & rebuilt for an international syndicated advertising agency I was working for in Fort Lauderdale. 10 years ago they were teaching their students that ribbon microphones were noisy and you shouldn't use them. WTF? I've found that some of the worst most appalling information I had ever heard. And one of their engineers was arguing that with me when I came in to coproduce overdubs at Omega Studios in Rockville, Maryland for a client of mine I had made a couple of other CDs for. I was going to record a female soprano and trumpet overdubs. I didn't think I would need to bring in my own ribbon microphones at the largest multi-million-dollar studio in the Washington DC area. My mistake. They didn't have a single ribbon microphone. It was pretty funny though because the day after, I was vindicated when I caught a GRP Records, Dave Grusin produced jazz extravaganza on PBS television. There in front of all the trumpets were ribbon microphones. So ever since that particular incident I've never thought much of Full Sale. And now look at all of the microphone manufacturers out there. You can hardly find one that doesn't make a ribbon microphone today. And they won't hire me to teach since I don't have a school sanctioned piece of paper. So I don't recommend them to anybody. I might if they were to hire me?

    I can teach rings around most
    Mx. Remy Ann David
     
  11. StephenByerley

    StephenByerley Active Member

    Hey Remy, nope, didn't go to FS and don't work there, but I know what you mean in that I've heard quite a bit of 'misguided' information coming from students. For balance, I also do know plenty of really solid engineers that teach good stuff, and you would not believe the facilities that are there now. They've taken over acres of land and have built incredible studios, and offer to their motivated students unprecedented opportunities. Plus, Gary is a nice guy and I can't imagine why he wouldn't have hired you... But I will tell you that I've asked nearly every kid "what is the resistance of three 8 ohm speakers wired in parallel" and in ten years, I've only received 1 correct answer. Sigh. And as for ribbon mics, either a) never heard Sinatra, or b) scared to admit they left the +48 on...
     
  12. RemyRAD

    RemyRAD Guest

    LMAO! Sinatra who? You mean ribbon microphones don't need +48? Isn't that gain? Or low frequencies?

    Many of these recording schools today have all become college accredited and offer actual bachelors & masters degree programs. Because of that, not having a piece of paper from another school, that I had to pay unreasonable money for, disqualifies me from teaching. It doesn't matter that I've taught some of the teachers. It doesn't matter that I have +41 years in the business, Grammy, Emmy & Soul Train Music Awards nominations and 20 years that NBC television. It only matters that you went to school to learn something that you never did for a living so you can teach it. My friends who have bachelors and Masters degrees in electrical engineering have consulted with me. I've designed and built numerous audio consoles from large 24 track desks to specialized broadcast consoles for a network owned and operated radio station. It's OK, if I really want to, just like my mother, a former Metropolitan Opera star, I could teach privately. One-on-one is a good way to go. That's how virtuoso musical students learn from virtuoso musicians.

    I wonder how I'll look in a blue vest working for Wal-Mart?
    Mx. Remy Ann David
     

Share This Page