Large vs small diaphragm condenser mics

[JanVigne]

"Experimentation is great! I don't think anyone is suggesting otherwise. My only suggestion was to try to get more levels than he has. It's also reasonable to try to leave a little room so you don't accidently run over, and also to allow you to easily add a few db of EQ if needed.

But it's just a suggestion, if someone wants to try to nail 0db perfectly, then by all means, go for it. In my experience, -6db is a nice middle ground. It works well, it allows for the occaisonal louder-than expected note, I've had it requested by mastering engineers, etc. But in recording, there are no rules."




"I'm sure its been mentioned, but the reasoning is that with most recording rigs now being digital 24 bit systems, theres enough headroom and low enough noise floors to run slightly lower levels as they can be boosted after recording, whereas as soon as signal goes above 0dbfs, it will clip and cannot be recovered.

In practice, setting the gain so that you're right on the limit could mean that one sudden increase in volume while recording could end up ruining an otherwise good take."




If you mess up a home recording, what should you do with it? Throw it away, hit "delete all", take a hammer to the recorder, swallow the SD card! Something other than worry about whether you hit an unexpected blip. It happens. It's not the end of the world unless you forgot to hit "record" during your recent interview with Elvis. If you burn the cookies, what do you do? Give them to the neighbor kids and make another batch. Move on, learn from what you've done and make concessions which will prevent the same mistake.

So, the way I see it, the additional system noise of boosting the levels isn't bothersome to you guys? You're more worried someone will mess up a home recording of themself?

You'll never learn the limits and the qualities of your system unless you experiment. I don't advocate blowing things up but messing up one take is not the end of the world. If the system overloads when you set the levels "this" high, set the levels a bit lower and try again.

Meters are virtually never true indicators of what's going into the system anyway. Manufacturers build in tolerances which tend to safeguard the average joe who feels the meters should be in the red at all times. Running the meters up to flash red on the highest peaks - like those we are likely to find in the solo guitar performance we have here; no deep bass, no ultra high frequencies, moderate transients and a dynamic range of about 15dB - is very likely still inputting realistically at less than 0dB.


But you'll never know unless you experiment. It's just that simple. Learn your equipment. You say experimentation is great, then you provide rules which should not be broken. Tesla would be left scratching his head.


All this about the low noise floor on digital systems leads me to believe we have a philosophical divide between folks who believe bits is bits and all digital is perfect and someone - me - who disagrees with that reasoning. A few dB of EQ is still not going to harm most digital recordings of a solo guitar. Where does it exist? Not where the recorder is most likely to overload. Are you not planning on some EQ before you hit "record"? You should be if you've experimented with your system and you already have a sound in your head that you're trying to get out of your system. If you haven't planned for EQ, then you're still just experimenting, right? Why are you throwing up all these obstacles to just trying something? If you have learned your system and have an idea of the sound you want out of the system, then before you ever pluck a string, you probably know whether you're going to be adding "a few dB" of EQ. You guys all sound like you have strict rules for just how far from the soundhole the mic should be placed and, if you violate that rule, you've committed some horrible crime.


Experiment - but experiment with a destination in mind - is all I'm saying. If you're sticking microphones in places no one else sticks mics, then there's probably a reason why no one uses those locations. But go ahead and try them. You can always hit "delete" after you discover why no one uses those locations.



Arguments form authority aside, I have no recordings to post, rick. I got out of audio awhile back, haven't recorded anything in several years and don't have a clue how I would post a recording. Over the last few years I've been trying to divest myself from the piles of unused equipment sitting in my closets - anyone want a Revox reel to reel that hasn't run in a couple of decades? I'm now down to a Tascam DR-05 and a Zoom that won't synch to my computer. Even it has not seen much duty lately as I've become rather lazy about recording myself while I'm just noodling with a song. I might be able to find the SD card with my recording of the 12 piece tuba band but, even if I did, like I said, posting to the web is not something I want to bother with.

The idea is to learn your system and its abilities. That's all.

[Luke W]

Originally Posted by JanVigne

So, the way I see it, the additional system noise of boosting the levels isn't bothersome to you guys?

I don't think people are saying that, people are generally saying that its better to record a bit lower and be able to boost later, than to clip and not be able to fix it, which is right.

This doesn't mean to say that people absolutely should record at a lower level (because they shouldn't really!), as you say, anywhere up to around -4 is a good level to aim for as it provides a strong enough signal to be well above system noise, but allows enough for peaks not to clip on the way in.

With even budget equipment these days though, you'd have to be recording at silly low levels and using a large amount of gain afterwards to run into problems with noise.

[ombudsman]

Originally Posted by Luke W

This doesn't mean to say that people absolutely should record at a lower level (because they shouldn't really!), as you say, anywhere up to around -4 is a good level to aim for as it provides a strong enough signal to be well above system noise, but allows enough for peaks not to clip on the way in.

Well, I'll say it. I shoot for -18 dbfs with peaks no higher than -12. There is no advantage to going any higher than that in 24 bit recording. The disadvantage, even if you never get any overs, is that you then have little headroom for your plugins (which have headroom needs in some cases beyond what is evident from their output level).

Noise in a good modern system used correctly, is really all about the source signal (and specifically, more often than not, the room). However much distance you have from the peaks to the noise floor is an attribute of the analog input signal that is not going to change at one tracking level on the A to D vs. another. It's not like analog recording where you had a relatively high noise floor in the recording medium itself, making it critical to give yourself as much distance from that as possible with big peaks.

And if you just want to do that anyway, you still can, but that's a gain staging choice rather than a tracking level issue. For example I run my preamps hot and probably push about a +25 dbu clean analog signal level on the outputs on hotter signals like drums. Then I turn down that output, either using a ladder attenuator or by using the line amp circuit on my interface, until the peaks are at -12 dbfs.

http://www.soundonsound.com/sos/sep1...s/qa0910-1.htm

[Joseph Hanna]

Originally Posted by JanVigne

Meters are virtually never true indicators of what's going into the system anyway. Manufacturers build in tolerances which tend to safeguard the average joe who feels the meters should be in the red at all times. Running the meters up to flash red on the highest peaks - like those we are likely to find in the solo guitar performance we have here; no deep bass, no ultra high frequencies, moderate transients and a dynamic range of about 15dB - is very likely still inputting realistically at less than 0dB.

I'm speechless....where are you getting this information? My job, my existence (and more and more each day) depends on me managing audio down to a 1/10th of a dB. The FCC has shifted it's rules and regs in the last several years but I can tell you I'd be out of work (in a single miscue) if your above statement was even remotely accurate. It's not. The Dolby spec now requires a loudness meter that will be rejected audio for air if it drops 1/10th of a dB below -23 rms. By the way that -23 dBfs which most of us here use. You appear to be assigning dB scales to dBfa conversations. The Pro Tools meters as well as the Logic meters as well as the Dorrough meters are all excessively accurate (as they should be) There is absolute no built in dummy room and to suggest so is way off kilter.

Originally Posted by JanVigne

Not where the recorder is most likely to overload. Are you not planning on some EQ before you hit "record"? You should be if you've experimented with your system and you already have a sound in your head that you're trying to get out of your system.

Suggesting one shoot for input levels of -6 dBfs is crazy. Industry standard (although it fluctuates depending of delivery specs) is -10 dBfs peak, which is actually up from the old standard of -20 dBfs. You already exceeded most delivery spec's at that point. You also run the risk (as many here have mentioned) of crossing the 0 dBfs war line. Further you've boxed yourself in a corner (by leaving only 4 dBfs of headroom). Digital eq's by their very nature clip easily, even if there's no additive eq applied. So now you've got to manage what has become a volatile track that at an angels breath is gonna distort and with zero room to fix. Recording it again (as you have suggested) is often not possible and offers no pragmatic gains even if succesful. I know of no professional engineer (and I work with many of them) that would ever suggest to a hobbiest they tempt faith with the 0 dBfs line. The digital scale is a completely different beast than an analog scale and it "reports" an entirely different story, both sonically and level wise. You seem to be talking as if they're the same thing. They're not.

[ombudsman]

Originally Posted by JanVigne

I knew this would come down to a argument of authority, which, in case you don't know, is a logical fallacy, guys;
http://www.fallacyfiles.org/authorit.html

http://www.wordiq.com/definition/Arg...from_authority

To be clear, the logical fallacy of argument from authority is a fallacy to the extent that the entire argument is nothing more than "this person (in some sort of authority) says so", or that they are not really an authority.

It does not, however, serve to disqualify references from qualified professionals who saying why a certain method is good or bad, nor arguments that are substantive and specific.

I don't think the issue is abstract freedom of the individual to make mistakes. (I wasn't aware that it was at risk, or that getting advice from me is somehow more oppressive than getting advice from you, when we're both basically suggesting different target numbers for the same parameter.)

Far be it for me to discourage experimentation, but on the other hand time and patience are zero sum games and there are plenty of things to try in recording (like mic placement) that, unlike recording too hot at the convertor, can actually be fruitful.

[JanVigne]

To be clear, working with professional engineers does not give you the right to state what they believe. To be even more lucid, what you believe is not what all "professional engineers" believe. To imply that only "unprofessional engineers" would think other than you is a logical fallacy and is just as rude as implying anyone who does not believe what you believe should be dismissed as a lunatic.

To be absolutely clear, none of this is of any assistance to the op, who can determine for himself where to set levels based upon a simple experiment.

So far, no one has told me one situation in which the op could not delete a bad recording and try again.

No one has stated the permanent damage done by experimenting with levels.

That leads me to believe this entire debate is nothing more than posturing on your part. What would happen if the op tried recording above -6dB? He might find out your advice wasn't perfect?

A bit of compression will keep the system from overloading should the input exceed safe levels. You guys have never used compression? A small amount of overload does what for a music signal? A little more sustain possibly?

Why not allow the op to find out the result for himself rather than insulting me? He has ears. He has a brain. Let him use both.


The only rule I know of in recording is, there are no hard and fast rules. What you did yesterday is probably not what you will do today. Experiment to find out what works for you today. Experience will guide you if you simply listen and think for yourself.

[ombudsman]

Originally Posted by JanVigne

No one has stated the permanent damage done by experimenting with levels.

I can fill that in for you. In 1992 I got a DAT machine and started using it to document performances. I didn't know much about digital recording back then, so at first I had the levels too high.

Now in my archive I've got 4 or 5 DATs that are unlistenable due to clipping. These are shows by bands and in some cases people who aren't around any more, including a number of tunes that I have no other recordings of or where the other recordings are lesser performances. It's almost worse having them around with terrible distortion than not having them at all.

If I had simply recorded at lower levels, they would be fine. They're already digital, and I transferred my DATs to my computer using an AES interface to avoid a new round of conversion; raising the level would be as simple as turning up the fader in my DAW, which adds no noise whatsoever. I've done this on dozens of other digital recordings.

You're trying to have it both ways. Either you're for "experimentation", or you're for higher levels (as you already said). You can't be both of those things at the same time while defining people who use a different level as being anti experimentation; we have an opinion already just like you do. I've experimented with different recording levels; didn't really mean to, in that particular case, I just didn't know any better; and then I learned from it.

Originally Posted by JanVigne

A small amount of overload does what for a music signal? A little more sustain possibly?

Statements like this are why you get the sort of response you are getting. You sound like you have some familiarity with some kinds of analog signals and are assuming digital works the same way. It doesn't. Not even all analog signals work the same way. In a clean amplifier circuit with a lot of negative feedback - like most modern mic preamps - a "small amount of overload" sounds horrible. It does not add nice sustain or gradually add musically useful harmonics in the way that a tube guitar amp or a low/no negative feedback amplifier circuit does.

In a digital recording system, there is no exceeding the clipping point like you can in analog. It's a hard limit. Your wave forms are flat topped; whatever information was above the clipping point is gone forever.

Don't take my word for it; experiment by recording some overs, and then use your waveform editor to zoom in and look at the shape of the tops of the clipped peaks. Note where the 0 line is. Listen to them, compared to smaller peaks that don't clip. All can be revealed with these easy steps.

Originally Posted by JanVigne

A bit of compression will keep the system from overloading should the input exceed safe levels. You guys have never used compression?

Setting a compressor or peak limiter to avoid clipping without undesired coloration of the sound is a much more complex task than simply recording at levels that will not clip. It does not make much sense to me to propose that as a solution when the problem goes away by turning one knob or fader, and absolutely nothing of value is lost by doing so.

I use analog compressors while tracking sometimes. What are the chances that someone who records too hot is even going to own one, much less know how to use it ? And in case it isn't obvious, a plugin compressor will not work for this because they would necessarily have to be after the A to D convertor at which point it is too late to avoid overs.

[sdelsolray]

Originally Posted by JanVigne

...
To be absolutely clear, none of this is of any assistance to the op....

Agreed.....

[Cocobolo Kid]

Originally Posted by JanVigne

To be absolutely clear, none of this is of any assistance to the op, who can determine for himself where to set levels based upon a simple experiment.

Originally Posted by sdelsolray

Agreed.....

Agreed too, but it sure is good reading. I'm settled in with some pepperoni pizza and Pepsi (today I'm only consuming things that start with "p" - the "q" day tomorrow is going to be tough; quiche and quesadillas again) until the admins call a halt to this fun. Maybe the admins won't notice since it's in the "record" section.

John

[rick-slo]

Well, seems how the topic has drifted here are some links

http://www.massivemastering.com/blog...ing_Levels.php

http://therecordingrevolution.com/20...rding-too-hot/

http://www.markdannrecording.com/Mar...nal_Sound.html

[KevWind]

Originally Posted by rick-slo

Well, seems how the topic has drifted here are some links

http://www.massivemastering.com/blog...ing_Levels.php

http://therecordingrevolution.com/20...rding-too-hot/

http://www.markdannrecording.com/Mar...nal_Sound.html

thanks for the links a great read

[JanVigne]

Quote:
Originally Posted by JanVigne View Post
To be absolutely clear, none of this is of any assistance to the op, who can determine for himself where to set levels based upon a simple experiment.

Quote:
Originally Posted by sdelsolray View Post
Agreed.....




AGREED!!! All this "honor" stuff and not a word about mic's and mic placement. What a waste.

[Cocobolo Kid]

Originally Posted by rick-slo

Well, seems how the topic has drifted here are some links

http://www.massivemastering.com/blog...ing_Levels.php

http://therecordingrevolution.com/20...rding-too-hot/

http://www.markdannrecording.com/Mar...nal_Sound.html

Rick,

Thanks for the links.

John

[Ty Ford]

Originally Posted by ljguitar

Hi Doug...

A question (not a challenge)…

Your CM6-MK41 is a 20mm which you are obviously calling small diaphragm. I find the tone from 20mm mics versus an equivilent 10-12mm mic more robust with adequate detail in the high end for me (which may have to do with 65 year old ears). Some of the very small diaphragm mics seem a bit sterile for my ears.

Larry,

It would be wrong to put a cmc641 into a group of "all SD mics." I have heard a lot of mics over the years, SD, LD, MD, dynamic and ribbon (and others). The cmc641 offers a special sound I've not heard in any other SD mic. The intricacy of the midrange and frequency balance are exceptional.

I didn't "find" them until just over ten years ago. Wish I had found them sooner. Find a place to rent one for a week and put it through its paces.

Regards,

Ty Ford

[RP]

Originally Posted by Cocobolo Kid

Maybe the admins won't notice since it's in the "record" section.

John

Wrong! AGF policy says that discussion and disagreement should be done with mutual respect, not personal attacks....

BE NICE! The Acoustic Forum is a happy place. Mean people are not welcome here. If you can't say something nice, don't say anything at all. Treat others with respect. If you bumped into each other in the guitar store, you'd probably be cordial. We ask that you be the same way here. The people on the forum are your neighbors, and share similar love for music and guitars.
We respect your opinions and you are welcome to express your opinions in a thoughtful and respectful manner. At times we all disagree, but it's the manner in which we disagree that is important. Slander, harassment, belittlement, hatefulness, racism, sexism, threats, degrading comments, insults & flames will not be allowed.