Topic: Is levels peaking at 0 to high?

[admkrk]

I have the limiters turned off in all my recorders, or at least I think they are. Some bands play at a more or less steady sound level and having the recording peaking around 0 is fine. Some bands on the other hand can be averaging around -20 and still end up full of clips. That is the bit that frustrates me.

[fireonshakedwnstreet]

DSatz, yes it would be the latter. Average level would be around -16dbfs with peaks around -6.

[DSatz]

fireonshakedwnstreet and SMsound, thanks for the replies.

I've been quite puzzled about where the other interpretation of "headroom" came from, where people think that it's a more or less wide zone below 0 dBFS that should be kept clear altogether. I think that it might come from the very dishonest polemics against digital audio that many audiophile magazines waged in the 1980s; they used to debunk (or so they pretended) 16-bit PCM's claim to have greater dynamic range than analog recording, in part by subtracting 10, 12 or more dB from the top, claiming that this was the "headroom region" and shouldn't be counted. (Of course, though, they counted every last dB of headroom for analog recording, despite the sharply increased distortion that occurs in that region.)

SMsound, there's no one norm for limiter behavior, and a circuit that starts to restrain the recorded levels as low as -7 dB is a bit of a "soft knee" in my book, between being a limiter and a compressor. I'm more used to "harder" limiters that kick in just below the limit of the medium, and that hold the line there. In any case I agree that anything more than the smallest amount of limiting will change the character of sound, and I don't want any non-emergency limiting in any live recordings that I make. It's better to under-record by 6 - 8 dB than to over-record by 3 or 4, especially if the system doesn't overload gracefully (some do and some don't). Also, with stereo recording it's very important that any limiters be "ganged" between channels. The limiting action, if it occurs, shouldn't cause the stereo image to shift left or right; that's more distracting than momentary "clean" overloads.

--Classical and opera are what I mainly record as well, but even with a 24-bit recorder and quiet preamps, I'm pretty sure that -18 dBFS for one's "all-evening" peaks is lower than optimal, especially if your recordings are ever played back in analog.

But circumstances do vary. It's impossible to calculate accurately with so many unknowns. Preamp noise depends greatly on gain settings, for example. The published noise specs for nearly all preamps are obtained at their maximum gain settings (50 - 60 dB or even greater), where the equivalent input noise is nearly always at its lowest by a significant amount. In concert recording I usually run preamps at ca. 30 - 35 dB gain. Around 20 years ago I measured the noise of all my preamps at ca. 30 dB gain; their relative rankings came out very, very differently from full gain. One of the quietest preamps at 30 dB gain, for example, was the noisiest one at full gain by a considerable margin.

And the noise spectra of condenser microphones vs. preamps are typically quite different, as is the prevalence of "shot" (impulse) noise vs. smooth, continuous-sounding noise. All in all the total, effective, potentially audible noise that you'd get from any given combination of microphone + preamp ... as I said, it's tricky, and back-of-the-envelope arithmetic isn't up to the task. So I stick with the accustomed practice of maximizing the signal levels at each step along the way while taking care to avoid overload. Nothing fancy, just basic principles.

-18 dB, incidentally, seems to come from Sony's old recommended practice for the PCM-1600, -1610 and -1630, which was to set 0 VU = -18 dBFS on steady tone. (I worked quite intensively with those processors in various East Coast recording and LP mastering studios back in the day.) But that recommendation was based on realizing that VU meters are "syllabic" rather than peak-responding, and that people quite properly let them go up to +3 VU for momentary signal peaks. The result with actual program material is that signal peaks on the digital side would commonly reach about -6 to -3 dBFS. Back then, dither wasn't widely understood and with many recorders (including the professional Sony processors that everyone used for CD mastering, in their default settings), distortion and "granular noise" increased quite audibly at lower signal levels. There was never any intention of leaving a wide zone unused near the top of the range!

All told, to me it seems very likely that there would be some s/n benefit if you raised your levels at least somewhat, especially if you consider the noise of your recorder's (or any media player's) analog output circuitry.

--best regards

[SMsound]

David,
I always learn something from your posts.

Something related I have been wondering about: Schoeps recently released some videos that state that with their mics, you maximize signal-to-noise ratio right around 20dB of gain. How should we think about reconciling the Schoeps-optimal 20dB of gain with the recorder-and-source optimal amount of gain that puts you closer to 0dB? (can't seem to find the video, but maybe this one: https://www.youtube.com/watch?v=aSAZN-B9DpU&t=2535s)

Re: limiter threshold, I don't see the limiter threshold on my MixPre published online (I can probably find it in settings somewhere), but the Sound Devices 6-series literature states that the limiter kicks in 'a few dB shy of clipping'. I have observed that the orange limiter lights go on when the meters say I have more than a few dB of headroom...maybe there were a couple samples that went loud and didn't move the meters but alerted the limiter.

I have also noticed that aiming for -18 dB on a Centrance MixerFace resulted in unusably noisy recordings, but on my MixPre-6 (24 bit) I have ambient noise floor above MixPre noise floor when I aim for -18, and this leaves maybe (?) 15dB of headroom in case the soprano gets excited (happens a lot). In practice, I should clarify that I usually ask the soprano to sing me one of the forte passages and put that at -18, but they often don't sing the (louder) cadenza during setup (you can only sing so many cadenzas per day...)...the result is probably more like a -12dB peak level, but just for a moment at the end of the song, and the rest at -18. I did have a show over the summer where I got more aggressive and aimed for -12dB during setup, and a soprano banged the limiters on her last note.

Your 'linked channel limiter' idea makes a lot of sense and I will do that from now on.

[DSatz]

I really wish that you could find the specific statement. Hannes is a down-to-earth guy and totally knows what he's talking about--incidentally he's an active performing musician in his off hours, and a very good one, I think (see https://www.youtube.com/user/nicaroschea). But the statement the way you paraphrased it, in and of itself, honestly doesn't make sense to me.

The preamp gain setting that gives the best s/n ratio depends on several variables, not the least of which is the peak amplitude of the program material that you're recording at the location where your mikes are. Different microphones have different sensitivities and equivalent noise levels. And as I said, different preamps have distinctly different equivalent input noise levels at different gain levels, and at one gain level preamp "A" may be quieter than preamp "B" while at a different gain level the reverse may very well be true--and even that will depend on the source impedance of the microphone (in the tests that I spoke of, I was using a test head connected to a Schoeps CMC microphone amplifier, with the preamp's phantom powering turned on). There's also the input noise and overload point of your recorder's (or a/d converter's) analog input electronics to consider.

So I don't see how there's any way that any one setting, such as 20 dB, can be considered optimal even generally. For my typical setups, that's definitely not enough gain.

--best regards

[SMsound]

I really wish that you could find the specific statement.
So I don't see how there's any way that any one setting, such as 20 dB, can be considered optimal even generally. For my typical setups, that's definitely not enough gain.

I think this is the video/statement I am remembering:
https://youtu.be/2st7KzoEHlo?si=Adw6ObxceFTEvOED&t=1729

They show a plot around 28:50 of dynamic range vs. gain where dynamic range peaks around 15-20dB of gain.

[Gutbucket]

Following along. Mr Satz, can you clarify a couple lines in your post above for me?  You said-

"..especially if your recordings are ever played back in analog." [and later..] "All told, to me it seems very likely that there would be some s/n benefit if you raised your levels at least somewhat, especially if you consider the noise of your recorder's (or any media player's) analog output circuitry."

Its the statements after "especially" that caught my eye.  Does that assume analog transfer of the raw recording from the recorder? Or that a digitally transferred recording will not have levels adjusted in the process of producing the released recording?  If so, I see how what you describe can be the case.  However, most tapers are digitally transferring the raw recording to computer and adjusting levels in editing software as part of the process of producing the finished output they and others will listen to.  That process typically includes increasing signal level to remove excess recording headroom that no longer serves a useful purpose in the released version, and instead becomes a potential detriment for the reason you describe.  When that is the case, I fail to see how analog playback or analog output circuitry noise would effect a recording originally made with excess headroom differently than one made with optimal recording levels.  I understand how the first may potentially have a lower s/n due to a higher noise floor, but that would be baked into the recording and independent of playback.

Just making sure I'm understanding you correctly and not missing something there.

Thanks for the discussion - in particular your mention of the different noise spectra of condenser microphones in comparison to preamps and the noise characteristics of preamps at various gain settings.

[DSatz]

SMsound, the topic from about 24:00 to about 30:00 in the recorded talk is dynamic range, which takes into account both the noise floor and the maximum SPL of the signal chain. He's very concerned to compare apples with apples: the digital microphone with the analog microphone PLUS the preamp that you have to use with it.

As he says (ca. 26:30), the CMD 42 isn't limited by having to pump out high voltages (more electrical power) at higher SPLs, so its maximum SPL is a few dB higher than that of the CMC 6*. But when you bring an analog mike preamp into the picture (ca. 27:15), the clipping point of its output becomes a much more serious limiting factor when you get near the maximum SPL of the microphone. In the slide at the point that you mentioned, he's saying (around 28:20) that you would need to limit the gain on your preamp to about 15 - 20 dB in order to avoid this overload at the maximum SPL.

He's definitely not saying that this is an optimal setting--on the contrary! His whole point is that analog preamps typically have lower input noise than the microphone's output noise only when set to higher gain levels than that, such as 30 - 35 dB. (This was shown in earlier slides, ca. 25:15 - 26:00.) Thus there exists no one gain setting for a typical analog preamp that both accommodates the maximum SPL of the microphone without clipping, and simultaneously offers the lowest noise for the quietest sounds that the microphone can pick up. If you need to record both at the very highest and the very lowest SPLs without touching any gain settings in between, the digital microphone offers a definite advantage as compared with the analog microphone--given that typical analog mike preamps can't put out 10, 20, 30 or more Volts (not that you really would want such levels to occur in practice).

But then he goes on to say two things. (ca. 28:55) "With suitable operation of the preamp, I can obtain performance from the analog system that is exactly as good, or nearly as good, as that of the digital microphone. But I must operate it in a suitable way; I must really set exactly the gain level that suits my application exactly." Secondly, (ca. 29:50) combination analog mike preamps and A/D converters exist that use multiple analog gain stages and gain ranging--and he says that such equipment can produce results that are very nearly equivalent to digital microphones.

--best regards

* note from me: This assumes a standard 1 kOhm load on the CMC 6. It can put out higher voltages if the load is lighter.

[noahbickart]

This is one the best ts threads in a while. Thanks everyone, especially David.

For “what we do,” that is, PA concert taping, 20db is likely about right much of the time.

To that end, we tapers should perform Dsatz’s test on the usual taper preamp suspects, at 20db.

[Gutbucket]

"Thus there exists no one gain setting for a typical analog preamp that both accommodates the maximum SPL of the microphone without clipping, and that simultaneously offers the lowest-noise performance for the quietest sounds that the microphone can pick up."

This brings into question how all the recent 32bit-floating-point recorders operate, each of which have integrated analog preamps of some sort or another which must accommodate a wide range of standard microphones of various sensitivities and output levels. In light of the above statement, accommodating such a range of diverse inputs would seem to be a considerably more complex challenge than the analog>ADC hand-off inside a digital microphone, where the output parameters of the specific microphone capsule and the ADC in the amplifier body can be closely matched.

"Secondly, (ca. 29:50) combination analog mike preamps and A/D converters exist that use multiple analog gain stages and gain ranging--and he says that such equipment can produce results that are very nearly equivalent to digital microphones."

I suppose this hints at is how its done.  Up until this point I've conceptually thought of those recorders as auto-ranging across multiple ADCs after a traditional preamp front-end stage, but had not considered that the auto-ranging is likely occurring ahead of multiple preamp stages, each specifically matched to its own ADC to optimize noise performance.

[wforwumbo]

Just piping that given I've been deep in the guts of a number of code bases that include metering, without going into detail and to respect individual companies' IP I wholly reinforce everything that DSatz says.

There's plenty of unknowns in implementation and too many standards (some for audio, some for film, some for television, some for game design) that all conflict with each other. I'm of the opinion that we are most likely to make a clean tape by setting levels conservatively at the preamp, and then raising levels digitally in post.

To reinforce a thought that keeps popping up in this thread, I'll comment that I frequently run my Omnis - which happen to be DSatz's former mk2 pair - with 20 dB of gain at the preamp, then I raise levels in post. It almost always makes a clean tape I'm proud of hearing without clipping, and I don't miss the added dynamic range at the lower end of the scale from setting levels so conservatively.

One additional monkey wrench I'll throw into this, is that I expect that all other things being equal (mic body, preamp, and digital gain at the bit bucket), if I were to run my mk4 cardioids vs my mk2 Omnis pointed at the same source I would expect that from the difference in frequency response the preamp would see less energy at its input for the cardioid. I suspect the best way to compensate for that is simply increasing gain, but I think it's foolish to run things through the mathematics for "proper level setting and gain compensation" instead of relying on our guts, intuition, and experience as recordists. And that's coming from someone who's known to be overly dependent on the scientific method, to a fault.

An additional note, to reinforce something DSatz has said above: "So I stick with the accustomed practice of maximizing the signal levels at each step along the way while taking care to avoid overload. Nothing fancy, just basic principles." This mirrors what analog electronics theory tells us, and is commonly referred to as "gain staging". The platonic ideal of electrical energy theory is to maximize power transfer from one section to the next, with as little distortion as possible. The theoretical ideal of any amplifier is to have infinite gain, infinite input impedance, and zero output impedance. In practice that's not exactly realizable (and actually not necessarily useful); so every analog design engineer worth their weight in gold that I know of, tunes amplifiers to match impedance to sections before/after it to maximize power transfer between sections, and tunes gain such that distortion can at least be measured and expected. In practice for the rest of us, what DSatz says above is what I think the best path forward is.

[Gutbucket]

It almost always makes a clean tape I'm proud of hearing without clipping, and I don't miss the added dynamic range at the lower end of the scale from setting levels so conservatively.

Typo?  More conservative levels will tend to reduce total available dynamic range.

This certainly bears repeating- "So I stick with the accustomed practice of maximizing the signal levels at each step along the way while taking care to avoid overload. Nothing fancy, just basic principles."

[rocksuitcase]

An additional note, to reinforce something DSatz has said above: "So I stick with the accustomed practice of maximizing the signal levels at each step along the way while taking care to avoid overload. Nothing fancy, just basic principles." This mirrors what analog electronics theory tells us, and is commonly referred to as "gain staging". The platonic ideal of electrical energy theory is to maximize power transfer from one section to the next, with as little distortion as possible. The theoretical ideal of any amplifier is to have infinite gain, infinite input impedance, and zero output impedance. In practice that's not exactly realizable (and actually not necessarily useful); so every analog design engineer worth their weight in gold that I know of, tunes amplifiers to match impedance to sections before/after it to maximize power transfer between sections, and tunes gain such that distortion can at least be measured and expected. In practice for the rest of us, what DSatz says above is what I think the best path forward is.

basic principles    YES!
on the topic of gain staging, not electronics as discussed above, but gear. my very first recording rig was Beyer M201c Hypercardioids and a Sony TC-D5M (1982). My mentor took the specs pf the Sony D5, focusing on its' input impedance. Then looked at the output impedance of several microphones we were choosing from. (Sennheiser 421, Sennheiser 441, Shure SM58's, Beyer M88, Beyer M201, Beyer M160). He advised us to buy the Beyer M201's because their output impedance was almost an exact match of the Sony's inputs. He told us that combo of mics/deck should almost never distort until we reached the max input SPL of the Beyers. (may have been 126dB???). I notice listening back that we RARELY had input distortion on those M201 tapes.
These were the stock pre-amps of the TC-D5M. Three years later, in 1985, we had the OADEs customize the inputs on three of our D5's.
Most of us are aware, but in summary for those who are referencing TS for technical reasons: levels can peak for a number of reasons, sometime overloading pre-amps and creating distortion. "Paying attention" to all gain stages in a recording chain is important. Utilizing most typical audience taper style pre-amplifiers gain at "medium-20-25dB" is probably ideal, as mentioned, one can increase gain at the recorder or mixer after that point in the signal chain. Microphones and pre-amps have various input and output impedance which should be analyzed before matching specific microphones with specific pre-amps.

[capnhook]

Re: Is levels peaking at 0 to(sic) high?







No

[Gutbucket]

Kevin bringing it back to ground level. Right. Less Than Zero is an Elvis Costello song.

Higher than zero is too high.
^
Print that on a T-shirt!