Some thoughts on the RH-1 gain structure

[ozpeter]

First up - this is not a RH-1 vs Zoom H2 thread! It's intended to get a better understanding (by me, maybe by you) of how to get the best from the RH-1 using lessons learned from the Zoom H2.

I shall make some statements as if they are facts, but I stand open to correction on facts, or others may want to express dissenting opinions without being able to offer proof!

The Zoom H2 has a three position mic input sensitivity switch (physical, but possibly controlling software) - High, Medium, Low.

The RH-1 has a two-position mic sensitivity switch, via a menu (in other words, clearly software controlled).

In both devices it would appear that this is the only control of the actual mic preamplification.

After the signal leaves the mic preamp, for both devices one can assume that it is passed to an analogue to digital converter, and either in that chip or in a DSP chip later in the the signal chain, there is provision for the digitized audio to be increased or decreased in level before it is passed to the storage media. In the case of the Zoom H2, this processing is determined by a control having a range of 0 - 127. In the case of the RH-1, the control has a range of 0 - 30.

It seems logical that there is no particular point in amplifying the signal within either device by means of the DSP chip if you are going to load the resulting file into a PC, where you could simply normalise it if too low, thus avoiding distortion in the digital domain with the benefit of hindsight. You could think of this in the same terms as with digital zoom in a camera - you might as well do the same zooming thing by cropping in Photoshop (etc), as the result will be the same, and you can change your mind about the crop.

It seems widely accepted that in the case of the Zoom H2, the level at which the signal is, in essence, passed straight through, is 100. If you need to reduce the level below this figure, then you're getting distortion within the mic preamp before the AD conversion, and lowering the level below 100 will simply give you a quieter distorted file. You should instead reduce the mic sensitivity from high to medium or from medium to low.

There seems to be a bit of disquiet in the Zoom fraternity about this, as it seems to imply that recording level control is too crude, but I'm not so sure that it's actually any different, in principle, from the way the RH-1 (and other Hi-MD recorders) works.

In the case of the RH-1, I'm not so sure that the "magic figure" - from 1 to 30 - is so clear. In the case of using the digital input (which the H2 does not have), it's known that the default value of 23 results in a pass-through (though still subject to resampling). In the case of the analog input, some say that 13 is the figure you should not go below, to avoid distortion at the mic preamp - personally I'd go for 15 or even 16 (the latter figure seems to be correct for standard line-in level too).

So internally, does the RH-1 have two different levels at which the digitized signal is not subject to gain increase or decrease - say 23 for digital or 15 for mic input? Or perhaps 23 is the magic figure, but the A to D converter is set up in such a way that the digital recording of a mic input (at a "flat" level of 23) would clip well before the analog input would?

Why does this matter? Well, perhaps it doesn't, so long as we know that going below (say) 15 for a mic input means the mic input is already distorted - so long as the extra processing, digitally, of the audio to reduce the level to 15 compared with the (possibly) unprocessed level of 23 doesn't degrade the signal (compared to processing later in a PC). Perhaps if you need to lower the level of your mic recording below 23, you should be using a less sensitive mic or otherwise reduce the input level to the recorder (advice which normally would apply is you'd gone all the way down to 15 or even 13).

Lastly I can't help noticing that mathematically, 23/30 is much the same as 100/127. So if the RH-1's "23" is the same as the Zoom H2's "100", but the latter has three choices of input sensitivity rather than the RH-1's two, it's we Hi-MD users who are not being given much control over recording level, really, compared to Zoom users.

Discuss! Disagree! Yawn! In a couple of weeks I should be able to come back with a careful and hopefully impartial comparison of the H2's actual performance vs the RH-1 - the H2 is rumoured to have finally got here to Australia a couple of days ago.

[Strungup]

just that the Quality of the Mic pre's in the RH1 are SOOOOOO much nicer . I own both The H2 and The RH1

So then the quality of the Audio in the computer when normalized or Raised will be Very noticable. Just two cents worth .

The thing about the H2 is the Limiter , I find myself leaving the Mic ain on Mid setting , The limiter ON type 3 , and raising the Recording levels up as high as I can without clipping or straining the limiter , the result is a good balance of audio level . Just not the same Clarity and Dynamic range of the RH1

Edited September 23, 2007 by Guitarfxr

[ozpeter]

It's very hard to be sure, but it seems generally believed that the Zoom H2 limiter, like the 0 - 127 level control, is also in the digital domain. What's your feeling? So if you use it in the H2, that's handy for giving an immediate punch to the sound when replaying it in the H2 right away, but if downloading to a PC, you could just as well limit it in your DAW software afterwards - and maybe fiddle with the settings without having to commit to something that later on you might wish you'd done otherwise.

Again, I'd go back to the camera analogy - most digital cameras can do stuff like sepia effect in the camera, but if you use it that way, later on you might think "damn, that shot would have been better in normal colour and now I don't have the choice" - as compared with taking the photo in full colour and deciding later to use Photoshop to add the sepia effect.

I think that's the thing that I'm getting at - should all processing with digital recorders of this type (RH-1 and H2 and anything with little control over analog input) be done only in postproduction, even levels, where possible? Given the dynamic range available, is getting the level up the last few dB in the recorder (by internal digital amplification) really necessary?

[Strungup]

well dont think of the Limiter as a Compressor .It isnt to Add Punch , it is to LIMIT the top part of the signal so you can raise the rec levels to be able to capture to quiet and Loud with a fairer balance , rather than no Limit and having to DROP levels and lose the quieter details.

But yes it is in the digital or software realm , I would much prefer a nice Tube limiter , but it just wouldnt fit in the ZOOM

Edited September 23, 2007 by Guitarfxr

[Nolonemo]

well dont think of the Limiter as a Compressor .It isnt to Add Punch , it is to LIMIT the top part of the signal so you can raise the rec levels to be able to capture to quiet and Loud with a fairer balance , rather than no Limit and having to DROP levels and lose the quieter details.

Guitarfxr, could you clarify as to your technique. Are you saying that with the H2 you use the limiter so you can use M rather than L without clipping, or that you use the limiter so you can record at, say, 118 instead of 100?

If the latter, it seems from what ozpeter is saying that you wouldn't actually be capturing any quieter details because you'd just be boosting what you had already captured.

Bear with me, I'm new to audio and trying to get a handle on this stuff. Thanks

[Strungup]

think of this,

The quieter the sound to be captured , the higher the gain has to be , or the closer the mic has to be to the object

The Louder a sound is, the Lower the gain has to be , or the farther away the mic has to be to the object

So , a Loud sound is forcing you to lower the gain , thereby losing the Quiet part that you ALSO wish to capture .

by judiscous use , of limit functions you can Tame the loud sound , so as to raise the gain a little ( it requires balance for a natural sound)

and then you get the quiet parts ... with more clarity .

[ozpeter]

That's true of analog gain control and analog limiting indeed - but once the signal has been converted from analog to digital, it should normally be left unprocessed ready for non-destructive post-processing in a audio editing program. Limiting on the H2 is almost certainly in the digital domain, and is therefore destructive of the original data. It's amplification positive or negative) not gain control.

Let's look at it this way (and this applies to Hi-MD as well) -

If you set the variable level control, (1-127 or 1-30, whatever), which operates in the digital domain after conversion from analog data, to the point at which it neither increases nor decreases the level after A to D conversion, and the meter shows that the level is too high, you should reduce the analog input gain switch or control from high to low (Hi-MD) or from high to medium or medium to low (H2).

Now you've done that and the audio is no longer metering too high. In the case of the H2 (I'm not sure what the figures are for Hi-MD hardware) the input level sensitivity switch operates in 10dB steps. So say the incoming audio was originally going 1dB over the top. You've reduced the input sensitivity by 10dB so it's now going to be peaking at -9dB. If you want it to peak at say -3dB, your only option (with H2 or Hi-MD) is to increase the gain digitally - in this case, if my maths is correct, you'll be multiplying the sample value of each sample by 2 to get that 6dB gain. Or, you could leave it as it is, record the data stream to the disc or card unprocessed, and in your PC later on, perform that multiplication (digital amplification), and you'll achieve exactly the same result, bit for bit. That original recording at -9dB might look a bit undercooked but with input gain control only switchable in two steps (Hi-MD) or three steps (H2), there's nothing you can actually do about it that will make any difference.

(If you are using something like a hard disc recorder with built in mic preamps, you'll have proper rotary input trims so you can set a more exact level of the audio going into the AD converter. And you won't find any control for gain changing in the digital domain when recording, because it's simply not required. You might well find it available on playback but that's another whole issue).

Now consider limiting in the digital domain. Taking that same example, where your signal is currently peaking at -9dB with no digital amplification being used. Using the 1-30 or 1-127 control, you could digitally amplify the signal by 10dB so that it was peaking at +1 - which is in the case of the Hi-MD machine would give you digital distortion. In the case of the H2, you could turn on the limiter, so that at the points where +1 would have been hit, it will instantly turn down the amplification so that it doesn't go over. Trouble is, that's messed with the original dynamics of the signal, and it's therefore destructive. And again, if you put the unprocessed file into your PC instead, you could use your DAW software to do precisely the same thing non-destructively, saving to a copy.

But - you wouldn't need to, would you? Once in the DAW, the program can examine the level of the whole file (with the benefit of hindsight, as it were), find the highest peak (in this example, still back at -9dB) and normalise it by 9dB to exactly 0dB peak with no limiting being necessary, and the full dynamic range of the digital data is preserved. Or, if you want to compress it for effect, you can do that in the DAW, save the result under a different filename, and change your mind after. But if you limit in the recorder, you can't unlimit it afterwards. You are stuck with what you've done.

About the ONLY time you should amplify a recording digitally in the Hi-MD recorder, or amplify and/or limit it in the H2, is if you want to do something like posting the file on the net right out of the machine and you can't post-process it in a PC first. It's exactly the same as with digital camera - you'd never use digital zoom unless you wanted to sent the photo straight from the camera ready-cropped, rather than photoshopping it later.

Sorry for length but these are matters where a lot of confusion seems to arise.

[Strungup]

I hate to correct you , but you are forgetting two VERY crucial things .

#1 The lower the level of the quiet Signals , the closer to the Noise floor they are .

#2 When Raisng those levels , also raises the Noise , Either in the Digital or Analouge realm .

The Idea of the Limiter is to raise those levels BEFORE they are attached to a specific noise level ........ It isnt just the processing , there is an ART to capturing sound , and there are tricks to it .

But to each his own , I have been around Sound for a LOOOONG time (Live, Studio, Venue, Public Address, Recording , Processing , ............)

Edited September 27, 2007 by Guitarfxr

[ozpeter]

I'm afraid no amount of art will overcome the science here! Noise doesn't come into it, as by the time the H2 limiter gets to operate on the signal, the analog noise is already there from the preamp, and there's the noise from the AD conversion process. The lmiter cannot somehow "raise those levels before they are attached to a specific noise level" - they are already attached. In fact it's effectively worsening the noise level because you'd normally use a limiter to raise the level of the quiet bits and turn down the level of the loud bits. It can only reduce noise in those parts of the signal chain following the limiter, not those before it, by optimising the level being fed to the next part of the audio chain. In this case the next part is the recording medium which in itself cannot add noise.

(An expander, or a noise gate, can give the effect of reducing noise as an expander lowers levels when little signal is present, and a noise gate cuts off the signal altogether when only noise is present. But they are conceptually opposite to a compressor or limiter).

Here's the signal flow involved in the case of the H2 -

Recording: Analog signal > AD converter > Limiter > Recording media

Playback : Recording media > DA converter > Analog signal

Because the recording media is completely transparent to the process - after error correction, what goes in is precisely what comes out, with no added noise - then the above signal path is precisely the same as....

Recording: Analog signal > AD converter > Recording media

Playback : Recording media > Limiter > DA converter > Analog signal

The H2 can't apply the limiter on playback in that way, only in recording, but if it could, there would be absolutely no difference in the outcome.

What you can do, though, is this -

Recording: Analog signal > AD converter > Recording media

Playback : Recording media > PC DAW > DAW's Limiter > DA converter > Analog signal

which again would be exactly the same, except that the DAW limiter will be different in operation from the H2 limiter - it'll be either built into the DAW or it will be one of the dozens of VST limiter effects available. Now if you feel that the limiter on the H2 is better than the limiter in your DAW software (unlikely), or you don't want to use the DAW for postproduction but you're simply interested in playing it back straight from the H2, then the H2 limiter has a purpose. Otherwise, it's unnecessarily messing destructively with your audio, because you can't change your mind about what you've done afterwards.

And my original point was, the same can be said of the digital level controls on the RH-1 - as far as I can see they do nothing that you can't do in postproduction with less chance of getting it wrong.

[Strungup]

Oz , there used to be a Post on here but I had deleted it along with a buttload of other stuff,

The file was of Taiko Drums , The audience , and the The leader of the Taiko group , was basically lost because the Recording levels had to be set so low ,so as to keep from blowing the vu meters . Now If I would have had a limiter on the RH1 at that time , I could have captured a MUCH broader Dynamic Range , And would have had a Perfect recording .

A Limiter doesnt Compress , Compressing is a Sonically visible Artifact that is UNdesireable unless it is judisously used.

Limiter only Puts a Very fast Cap on the Rec level peaks , so that you can Raise the gain up a little to Get a Broader Dynamic range to the recording .

You cannot bring back , WHAT YOU DONT HAVE . If you dont capture it , it isnt there. There Has to be a Certain Level of workable audio , before you can do ANYTHING with it in your DAW .

Limiting is a Balancing act , .... getting just enough so you dont notice it , but getting more of the Audio you want , so as to be able to work with it later .

Obviously you have a little knowledge about electronics and audio pathways , and Todays Digital recording enviornments are not quite the same thing as the 48 channel , and 56 channel analouge boards I am use to . But Certain principles DO remain the same .

With Analouge you can Push a Much hotter signal , and instead of clipping you get Saturation , both are forms of distortion , But the Saturation can be used to your advantage, Clipping just is what it is ,Bad.

On the original recording , you want to get as much audio as you can with out Clipping , with Analouge You push the signal as hot as you can without the saturation becoming audiologically visible .

To get that audio in the Digital realm , you NEED limiting capability , without it , you have no punch , and reduced clarity , and dynamic range .

do you have the H2 yet ? If you do , go to a Loud Concert , record two files One with Limiting , and one without , On the H2 you can adjust the Rec levels while recording and if you dont mind the brief POP you can switch the mic gain switch also ( Dont recommend it , but for experiments sake)

also Ask any decent sound guy , about running live sound , without a Limiter , , ......The first thing he /she will do is Laugh at you .

You run a live show withouth Limiters you will blow your amps up . Literally . Thery will start Clipping with the first good bass slap or Drum kik , and 1/4 of the way into the show Poof , thermal shutdown .

No Limit , No Volume . Acoustic show fine no problem , Unless you want to get louder . Small venues fine no problem , but anything with Large dynamic peaks , needs a little more control to get it , and get it right .

[ozpeter]

Yeah, I recorded "TaikOz" (I think that's the name of the Australian Taiko group) some years back - and indeed they were loud in parts and close to inaudible in others. Probably the biggest dynamic range one will ever encounter. I entirely agree with you about the need for limiting and /or compression in that context in the analogue side of the system before the audio is fed to the AD converters. Trouble is with the H2 that the limiter is after the AD converter, too late in the audio chain to operate the way you describe.

Do I have one? Heh, I wish! I ordered one from a supplier across the country about 6 weeks ago or more at a good price, and apparently they arrived in the distributor's warehouse in the last few days. Meanwhile a company about five minutes from here are offering them for almost the same price and I could've popped down and picked one up, once they get out of that warehouse and into the hands of the retailers.

Once I get one I'll report on my experiences based on hands-on, rather than on reading the manual and reviews on the net!

Cheers!

[Nolonemo]

Yeah, I recorded "TaikOz" (I think that's the name of the Australian Taiko group) some years back - and indeed they were loud in parts and close to inaudible in others. Probably the biggest dynamic range one will ever encounter. I entirely agree with you about the need for limiting and /or compression in that context in the analogue side of the system before the audio is fed to the AD converters. Trouble is with the H2 that the limiter is after the AD converter, too late in the audio chain to operate the way you describe.

Do I have one? Heh, I wish! I ordered one from a supplier across the country about 6 weeks ago or more at a good price, and apparently they arrived in the distributor's warehouse in the last few days. Meanwhile a company about five minutes from here are offering them for almost the same price and I could've popped down and picked one up, once they get out of that warehouse and into the hands of the retailers.

Once I get one I'll report on my experiences based on hands-on, rather than on reading the manual and reviews on the net!

Cheers!

The effect OzPeter is talking about, I think, is illustrated in this thread on the 0-127 adjustment on the H4 changing the level of signals that are clipping, but because it's happening after the A-D conversion, it's just adjusting the gain of the clipped signals:

http://www.2090.org/zoom/bbs/viewtopic.php...8447d4e56fc9e86

(see screenshot halfway down the page)

Seems to me, and I believe this is what OzPeter is saying, that IF the limiter acts on the signal after the A-D conversion and not before then all it's doing is knocking an already clipped signal down to 0 dB, which really accomplishes nothing, since it doesn't avoid the actual problem, which is the clipping of the peak. Of course, when a limiter kicks in before the A-D conversion, that's not what is happening, and clipping is avoided.

Of course, it's easy to test what the H2 is doing: set the L-M-H switch so that the signal is clipping, and record. Egage the limiter and re-record the same sample. I suspect that both samples will show clipping, the difference being that the second sample will be limited to 0dB at the peaks.

Edited September 28, 2007 by Nolonemo

[Strungup]

The effect OzPeter is talking about, I think, is illustrated in this thread on the 0-127 adjustment on the H4 changing the level of signals that are clipping, but because it's happening after the A-D conversion, it's just adjusting the gain of the clipped signals:

http://www.2090.org/zoom/bbs/viewtopic.php...8447d4e56fc9e86

(see screenshot halfway down the page)

Seems to me, and I believe this is what OzPeter is saying, that IF the limiter acts on the signal after the A-D conversion and not before then all it's doing is knocking an already clipped signal down to 0 dB, which really accomplishes nothing, since it doesn't avoid the actual problem, which is the clipping of the peak. Of course, when a limiter kicks in before the A-D conversion, that's not what is happening, and clipping is avoided.

Of course, it's easy to test what the H2 is doing: set the L-M-H switch so that the signal is clipping, and record. Egage the limiter and re-record the same sample. I suspect that both samples will show clipping, the difference being that the second sample will be limited to 0dB at the peaks.

actually I own the H2 , and the limiter prevents clipping quite well , you cant push it to hard but , I get an extra 4 db easily .

What it doesnt prevent is distortion in the mic pre , if the mic pre is being overloaded you have distortion , I have noticed this , it isnt digital clipping , it is Mic input overload. I change the mic gain setting , then raise the levels back up . The Digital clipping happens at the write stage , where the recording actually happens , the A/D converter can go higher than 0 db . It happens after the limiter , there is a Definate difference . I should have used the H2 yesterday .

There was an Opera at my kids school , A Professional troup came to give a demo and the where doing Verdi . The Dynamics in the room were out of control . I had the RH1 , I was sitting at the back of the auditorium , The entire school in front , The Stage had no Amplification .

So the Pianissimos were quite low , and when the tenor hit the Mezzo forte's the Levels go thru the roof , ....Lets just say the jog dial on the 40elk was getting a workout .

I have experimented with the H2's limiter at home , using different distances , and power of stroke on my acoustic . My Guitar is pretty big sounding so I can easily peg the H2 . using the Limiter I can set the H2 a foot in front of me and Bang on my guitar , with mic gain on high I hear preamp distortion ,or breakup but not Clipping , two very distincly seperate sounds.

I put the mic gain to Mid and adjust levels I can give the guitar a workout , The only problem in the Proximity Effect , the H2 is Bassy at close range , way to much ,even with the Low cut in . So I prefer to use the RH1 ,and the AT822 .

Edited September 28, 2007 by Guitarfxr

[Nolonemo]

actually I own the H2 , and the limiter prevents clipping quite well , you cant push it to hard but , I get an extra 4 db easily .

What it doesnt prevent is distortion in the mic pre , if the mic pre is being overloaded you have distortion , I have noticed this , it isnt digital clipping , it is Mic input overload. I change the mic gain setting , then raise the levels back up . The Digital clipping happens at the write stage , where the recording actually happens , the A/D converter can go higher than 0 db . It happens after the limiter , there is a Definate difference . I should have used the H2 yesterday .

Ahh, thanks for the info on that, very interesting....

I also have an AT822. I recall you mentioned the impedance mismatch with the H2 in another post. Do you think I would get better results with an impedance matching transformer, or should I just go for a battery-powered preamp and run the mic into the line in (a considerably more expensive solution).

Thanks.

Edited September 29, 2007 by Nolonemo

[Strungup]

Ahh, thanks for the info on that, very interesting....

I also have an AT822. I recall you mentioned the impedance mismatch with the H2 in another post. Do you think I would get better results with an impedance matching transformer, or should I just go for a battery-powered preamp and run the mic into the line in (a considerably more expensive solution).

Thanks.

Line in Most definately , I was actually liking the line in on the H2 , it is fairly clean , the Mic input just doesnt cut it . But I have used to line in for an MD that I wanted to convert , So there is and option for Net MD's or files that the RH1 cannot Upload , is Use the Line in on the H2 and record to the H2 then Upload in an Audio Editor freindly format .

As far as a Preamp the Rolls MX34 would be a favorable option ,I want one , but it isnt available here .

Edited September 29, 2007 by Guitarfxr

[Nolonemo]

Line in Most definately , I was actually liking the line in on the H2 , it is fairly clean , the Mic input just doesnt cut it .

....

As far as a Preamp the Rolls MX34 would be a favorable option ,I want one , but it isnt available here .

The Rolls looks nice, and the price is right. I notice the input impedence is "600 Ohms or greater XLR," is that close enough to the AT822 (200 Ohms) to be ok with that mic? Also the specs for the Rolls at B&H give its output at 50 Ohms while the H2 line in is spec'd at 10K Ohms in the H2 manual. Problem there?

Thanks for your help on this.

[Strungup]

The Rolls looks nice, and the price is right. I notice the input impedence is "600 Ohms or greater XLR," is that close enough to the AT822 (200 Ohms) to be ok with that mic? Also the specs for the Rolls at B&H give its output at 50 Ohms while the H2 line in is spec'd at 10K Ohms in the H2 manual. Problem there?

Thanks for your help on this.

You would have to use XLR adapters for the AT822 , I believe the MX34 onlys has XLR connects (and the arent Nuetrik) but at any Guitar /Music Store in the PA section there will be a bunch of Adapters and Gizmos ,

The AT 822 is 250ohms oer side , so it might not be perfectly clean but it will be pretty good .

I have two cables with 1/4" Female -> XLR for my AT822 they are about 1 1/2 feet long . Mono Phone jack - XLR pins 1&3 ground Pin 2 hot

Edited September 29, 2007 by Guitarfxr

[Nolonemo]

You would have to use XLR adapters for the AT822 , I believe the MX34 onlys has XLR connects (and the arent Nuetrik) but at any Guitar /Music Store in the PA section there will be a bunch of Adapters and Gizmos ,

The AT 822 is 250ohms oer side , so it might not be perfectly clean but it will be pretty good .

I have two cables with 1/4" Female -> XLR for my AT822 they are about 1 1/2 feet long . Mono Phone jack - XLR pins 1&3 ground Pin 2 hot

Thanks for the XLR-phono pinout.

What about the impedance mismatch between the Rolls and the H2 (the specs for the Rolls at B&H give its output at 50 Ohms while the H2 line in is spec'd at 10K Ohms in the H2 manual) would that be a problem?

[Strungup]

Thanks for the XLR-phono pinout.

What about the impedance mismatch between the Rolls and the H2 (the specs for the Rolls at B&H give its output at 50 Ohms while the H2 line in is spec'd at 10K Ohms in the H2 manual) would that be a problem?

no that shouldnt be , I think that 50 might be a typo . Line out should be 50 kohms so I will bet it is a Typo