Topic: Sony PCM M10 with Preamp level balance

[F.O.Bean]

levels? I run line in to the M10, using the built-in limiter and CRUSH my levels.

You can get the red over lights flashing on, but margin of -1dB if you set it carefully. In fact, if you only briefly light the OVER indicator on the LCD, the resultant signal will sometimes not even hit -0dB(full scale!)

Anywhere from 4-6 is pretty safe on the M10's knob. If you're running the knob between 1.5 and 2, I would worry that you could brickwall your inputs, but the thing is rated to take 2.0 volts on the line inputs!!!

Why oh why do you run a limiter? Its just a cheap compressor I would NEVER recommend doing that, EVER

[Sloan Simpson]

I always run the limiter on my M10, but in my case it's just for insurance in case of accidental overs.

[acidjack]

I always run the limiter on my M10, but in my case it's just for insurance in case of accidental overs.

Maybe we should rehash what the M10 limiter actually does.. (I'm not totally sure).

Is it a hard limiter, like the function in a DAW, where it simply truncates the signal above a certain threshold?

Or (as I suspect), does it act more as a compressor/limiter, which starts to compress the signal as you approach the peak threshold?

I know you tape all the time, so I'm assuming you'd have noticed by now if it was screwing up the sound of your stuff.  But this is why I've always been loathe to run one, and why I thought the conventional wisdom here has always been not to do so.

[George]

I figured it that way too acidjack.  I just set my rec level to 5 and slowly work up or down at the beginning of the show.  My tinybox is set to medium and this is with the B3's.

[Sloan Simpson]

I always run the limiter on my M10, but in my case it's just for insurance in case of accidental overs.

Maybe we should rehash what the M10 limiter actually does.. (I'm not totally sure).

Is it a hard limiter, like the function in a DAW, where it simply truncates the signal above a certain threshold?

Or (as I suspect), does it act more as a compressor/limiter, which starts to compress the signal as you approach the peak threshold?

I know you tape all the time, so I'm assuming you'd have noticed by now if it was screwing up the sound of your stuff.  But this is why I've always been loathe to run one, and why I thought the conventional wisdom here has always been not to do so.

Yeah I don't know what it's really doing but I've never noticed any adverse effect from it. I'm curious to know where the threshold is. I generally don't run the levels hot enough for it to engage (I'm assuming), just like the idea of a safety.

[stevetoney]

I don't know if the M10 does hard or soft limiting, but I remember from a thread that I read a long time ago that the limiter on the m10 functions real well and is quite transparent.  I don't ever use it though because I always run with my levels peaking between -12db and -6db.

Another reason to like the m10.

[Ghost of sml42]

i could've sworn i read in another thread that unity gain on the m10 was around the dial set to 6.

i performed some tests on my m10 when i first got it & found unity gain to be a hair under 6. so i try to run my m10 around there and let the preamp do the rest (dpa406x > mma6k > m10, for what it's worth)

[fmaderjr]

Anywhere from 4-6 is pretty safe on the M10's knob. If you're running the knob between 1.5 and 2, I would worry that you could brickwall your inputs, but the thing is rated to take 2.0 volts on the line inputs!!!

The M10 is much better than most recorders in this regard. Guysonic's tests showed that as long as the knob is at or above 1.0 with the levels under 0 dB you won't get distortion. An informal test I made seemed to back this up. Definitely no worries if the knob is at or above 1.5. (Note the D50 is not as good in this regard-the knob must be at or above 2.5).

[DSatz]

The problem described in the original post affects everyone who ever uses an analog preamp or mixer to feed the analog line inputs of any piece of audio equipment. It isn't a simple problem, it doesn't lend itself to "cookbook" solutions, and some of the overly simplified answers that have been given in this thread are rather likely to be wrong, I think.

In reality, to figure out the best set of gain settings (assuming that you want to avoid a "trial and error" approach) would require knowing quite a few specifics of the recorder, the preamp or mixer, the microphones, the noise levels in the recording venue, and the maximum sound pressure levels to be expected. When all these things are properly considered, it may turn out that there's a fairly wide range of settings for preamp gain and recorder gain that would be as good as it gets and as good as each other. In other cases there may be only a narrow range of optimal settings--or it may be that only one precise set of gain settings is optimal for that combination of equipment and recording conditions. Without all the facts, there's no way to predict accurately. And those facts can be hard to come by. So let's be reasonably cautious with our suggested answers, OK?

One thing that's for sure: The "unity gain" setting on a recorder is completely arbitrary, and has no direct bearing on optimal gain settings. On the other hand "unity gain" settings are unlikely to be extreme, and extreme gain settings (in either direction) are generally best avoided. So if you're stuck trying to solve this type of problem by trial and error, the "unity gain" setting on the recorder won't usually be a bad place to begin your trials. But there's no reason to expect that it will ultimately turn out to be the best setting, either.

--best regards

[DSatz]

BinauralBrisbane, maybe I should try to explain how I work this out when it's my problem. The mikes and preamp weren't specified in your original post, but their specifics have a lot to do with this, which I'll get to (I hope).

For me the #1 constraint is overload. With some equipment a mild overload may sound OK, but I prefer to treat overload as something to avoid absolutely. So one of the first things I want to know about a preamp or recorder is, at what signal voltage do its inputs overload? The key fact here is that the inputs of nearly all preamps and recorders have a maximum voltage that they can handle irrespective of their control settings. If you send in any higher voltage, the inputs will overload and distortion will occur, even if you have the gain controls turned down to the very lowest setting that passes a signal. This is because there almost always is at least one stage of active electronics that the signals have to pass through before they reach the gain control--and that circuitry always has some finite overload limit. The gain control can't help because of where it sits in the overall circuit. The only way to avoid this would be to put the gain control--a passive device--directly at the inputs. For technical reasons this approach is less than optimal from the standpoint of noise and sometimes frequency response, so it is hardly ever used.

My microphones can easily handle the highest sound pressure levels that ever occur in the kind of recording I do. The inputs of my preamps can handle the signal levels which my microphones would put out if their maximum sound pressure level was reached. So with the equipment choices I've made, I don't ever have to worry about preamp input overload. I recommend that people consider that issue when choosing microphones, preamps and/or recorders. The worst case occurs when people try to use professional condenser microphones with the microphone inputs of consumer recording equipment. Those inputs are generally designed for use with dynamic or a certain class of (mostly Asian-made) electret condenser microphones, while professional condenser microphones have typical output voltages ten or more times higher (and many people here record music that gets loud sometimes--I hope I'm not giving away any secrets there!). There are also problems with microphone powering and balanced vs. unbalanced connections.

Those are serious issues, and solving the problems can be quite a headache at times--especially the unbalancing of signals from balanced microphones, which may need to be done differently for each different type of microphone that you use; there's no such thing as a "balanced to unbalanced" adapter that works for all types of microphones, other than a good transformer--but good transformers are expensive and not very small. So a lot of people use mike preamps and feed their outputs into the line inputs of their consumer or "prosumer" recorders, whether that improves the signal-to-noise ratio of their recordings or not; the main thing is, it tidies up the whole set of connection issues and preserves relative sanity in that way.

What about the recorder's line inputs? They probably have an overload limit as well. There's a method that some people use here, which I think makes a good deal of sense: They send in a test signal at a level just below the input overload limit, then turn down the gain control on the recorder so that the meter reads 0 dB (full scale). Then they note where the recorder's gain control is set. That setting is the lowest setting that can be used. If you have to turn the control down any farther to avoid hitting 0 dB on the meter, the input voltage must logically be higher than the input overload point--so you'd be recording distortion even though the meters aren't reaching 0 dB.

While this definitely doesn't answer the whole question, I hope it provides enough to chew on for a while. We can maybe talk about the remaining parts of the question another time.

--best regards

[-ned-]

Maybe we should rehash what the M10 limiter actually does.. (I'm not totally sure).

Is it a hard limiter, like the function in a DAW, where it simply truncates the signal above a certain threshold?

Or (as I suspect), does it act more as a compressor/limiter, which starts to compress the signal as you approach the peak threshold?

I know you tape all the time, so I'm assuming you'd have noticed by now if it was screwing up the sound of your stuff.  But this is why I've always been loathe to run one, and why I thought the conventional wisdom here has always been not to do so.

Hi all, first post...

My understanding is that it's a digital limiter that records a second version of the audio into the buffer at -12bB lower level and then replaces any peaks that go over 0dBFS with the lower version's peaks. Only effective for peaks that don't go over +12bBFS. After that you just get distortion. So it's not really a limiter in the traditional sense more like a post processing fix done through tricky copy and paste algorithm. Pretty cool, but it must eat up processor and power resources.

Personally I wish the M10 had an analog limiter that simply brick-walled peaks that go over 0dBFS, but I guess if one's doing their job right with setting optimal levels it shouldn't matter which type you have on your recorder, you'll simply have enough headroom that you won't peak out (I know... that's in fantasy land were everyone and everything respects your level settings and don't go over )

Speaking of levels... I think I got the idea on how to calibrate the line-in levels: test tone at +4dBu into line-in then set your recorder level to the desired headroom level, -20dBFS, -18dBFS, -12dBFS. My question is how can one go about calibrating through the mic input in the same manner? Say for a set of binaural mics running straight into the mic input of the recorder?

[yousef]

My understanding is that it's a digital limiter that records a second version of the audio into the buffer at -12bB lower level and then replaces any peaks that go over 0dBFS with the lower version's peaks. Only effective for peaks that don't go over +12bBFS. After that you just get distortion. So it's not really a limiter in the traditional sense more like a post processing fix done through tricky copy and paste algorithm. Pretty cool, but it must eat up processor and power resources.

That's how the D50 works but I'm pretty sure that the M10 just has a traditional auto gain control function.

[-ned-]

Please don't take this as arguing, I got my information from a review on the m10 that stated the limiter functioned in a similar way as the m50, so I pretty much took it on face value.

I did some checking around this forum and found this thread >> http://taperssection.com/index.php?topic=133323.0 >> where the general consensus was that the m10 did not share the same limiter as the m50 or d1. Though reading through the posts one member described a peculiar side effect on his d1 exhibited when using the limiter and recording things that went outside the limiter's effective range...

The Operating Instructions Manual states that the limiter cannot prevent clipping “when audio over 20dB is input.” This is supposed to result in clipping. In my experience clipping under such conditions is extremely rare. Instead, what the limiter does when its capacity is exceeded is to create a “hole-in-the sound.” This has a split second lag time: the sound starts to peak, and then disappears for a second.

This got me thinking that if the m10 exhibited a similar 'hole' in the clipped peak when overloading the limiter then it might suggest that it does share a similar style of limiter as the m50 and d1 recorders... so I did a quick test with my m10.

With the mic gain on low and rec. level at 8 I snapped my fingers a few inches away from the right side mic to get the clip indicator red. Opening the file in my audio editor and zooming in I could see that the transient peak did indeed have a sort of hole in the manner described by Noam and his d1 experience. I've attached a screen capture showing the visual effect when viewing the waveform.



I don't know if this is definitive proof that the m10's limiter is the same as it's more expensive counterparts, nor am I sure what is going on with this side effect. If Noam is still active on the forum he could probably verify if this is the same effect he had on his d1 recordings, or maybe other members with a m50 or d1 could try a similar test to confirm... that is if you cared to get all geeky to do a test that is probably not that relevant. 

Though it's not a deal breaker by any stretch I still wish it had an analog limiter, but hey, I can always run a preamp with an analog limiter into the m10.

[yousef]

Interesting.

I'm not sure if that waveform looks like what Noam seems to be describing: rather than the sound "disappearing for a second", that looks to me like a clipped waveform that has been lowered in volume after the clipping has occurred - ie similar to what you might see if someone had reduced the volume of a clipped area in an audio editor. Or, of course, an AGC kicking in to reduce the level of a signal that had already brickwalled at an earlier stage.

I'm not sure what this tells us about the limiter in the M10 though 

I always thought that the D50's limiter was a very elegant solution (in theory that is, not sure if I've ever heard it in practice) and I'd be very pleased if the M10 did the same thing. I suspect Sony would have made it explicit in their product literature if it did though.